Please note that this
page might take at least half a minute to load because of its length.
Unfortunately,
Internet Explorer 11 will no longer play the videos I have
embedded below. As far as I can tell they play as intended in other Browsers.
However, if you have
Privacy Badger [PB] installed, they won't play in Google Chrome unless you
disable PB for this site.
[Having said that,
I have just discovered that they will play in IE11 if you have upgraded to
Windows 10. I
have no reason to believe otherwise of Windows 11. It looks like the problem was with Windows 7 and earlier
versions of that operating system. ]
If you are using Internet Explorer 10 (or later), you might find some of the
links I have used won't work properly unless you switch to 'Compatibility View'
(in the Tools Menu); for IE11 select 'Compatibility View Settings' and add this site (anti-dialectics.co.uk). Microsoft's browser,
Edge, automatically
renders these links compatible; Windows 10 does likewise. I haven't checked yet, but I assume
that this is also the
case with Windows 11.
However, if you are using Windows 10, IE11 and Edge unfortunately appear to colour these links
somewhat erratically. They are meant to be mid-blue, but those two browsers
render them intermittently light blue, yellow, purple and even red!
Firefox and Chrome reproduce them correctly.
Several browsers also appear
to underline
these links erratically. Many are underscored boldly in black, others more
lightly in blue! They are all meant to be the latter.
Finally, if you are viewing this
with Mozilla Firefox, you might not be able to read all the symbols I have
used
-- Mozilla often replaces them with an "º'.
There are no such problems with Chrome, Edge, or Internet Explorer as far as I can
determine.
As is the case with all my
work, nothing here should be read as an attack
either on Historical Materialism [HM] -- a theory I fully accept --, or,
indeed,
revolutionary socialism. I remain as committed to the self-emancipation of the
working class and the dictatorship of the proletariat as I was when I first became a revolutionary
over thirty-five years ago.
The
difference between
Dialectical Materialism [DM] and HM, as I see it, is explained
here.
Several readers have complained about the number
of links I have added to these Essays because they say it makes them very
difficult to read. Of course, DM-supporters can hardly lodge that complaint
since they believe everything is interconnected, and that must surely apply even to
Essays that attempt to debunk that
very idea. However, to those who find such links do make these Essays
difficult to read I say this: ignore them -- unless you want to access
further supporting evidence and argument for a particular point, or a certain
topic fires your interest.
Others wonder why I have linked to familiar
subjects and issues that are part of common knowledge (such as the names of
recent Presidents of the
USA, UK Prime Ministers, the names of rivers and mountains, the titles of
popular films, or certain words
that are in common usage). I have done so for the following reason: my Essays
are read all over the world and by people from all 'walks of life', so I can't
assume that topics which are part of common knowledge in 'the west' are equally
well-known across the planet -- or, indeed, by those who haven't had the benefit
of the sort of education that is generally available in the 'advanced economies',
or any at
all. Many of my readers also struggle with English, so any help I can give them
I will continue to provide.
Finally on this specific topic, several of the aforementioned links
connect to
web-pages that regularly change their
URLs, or which vanish from the
Internet altogether. While I try to update them when it becomes apparent
that they have changed or have disappeared I can't possibly keep on top of
this all the time. I would greatly appreciate it, therefore, if readers
informed me
of any dead links they happen to notice.
In general, links to 'Haloscan'
no longer seem to work, so readers needn't tell me about them! Links to
RevForum, RevLeft, Socialist Unity and The North Star also appear to have died.
~~~~~~oOo~~~~~~
This Essay seeks to
challenge a well established and dominant set of ideas about 'mind', language
and 'cognition' -- theories that are widely accepted by
philosophers, cognitive scientists, revolutionaries and other assorted Marxists.
I call this tradition the
Platonic-Christian-Cartesian Paradigm.
It is
important to add, however, that many of the conclusions drawn
below depend on much that has gone before in other Essays published at this site (particularly Essay Twelve
Part One),
as well as others yet to be published.
Moreover, the material
below is far from complete;
as I noted on the opening page of this site:
I am only publishing this material on the Internet
because several comrades whose opinions I respect urged me to do so back in 2005
-- even though
the work you see before you is less than half complete. Many of my ideas are still in
the formative stage and need
considerable
attention devoted to them to
mature.
I estimate this project will take another ten
or twenty years to complete before it is fit to publish, either here,
in its final form, or in hard copy.
At a later date, I will be
returning to this Essay to add material on Vygotsky and Chomsky, as well as a
handful others on the left who have written on this topic.
Even in
its incomplete state, the reader will find this Essay challenging the widely held
views
mentioned above -- i.e., those belonging to the Platonic-Christian-Cartesian
Paradigm. This family of theories has in one form or another dominated 'Western' thought since Ancient Greek times,
and that includes
ideas concerning the nature of 'mind', 'consciousness' and 'cognition' held by the vast majority of Dialectical Marxists.
~~~~~~oOo~~~~~~
It is important to
add that a good 50% of my case
against this area of DM has been relegated to the
End Notes. This has been done to allow the main body of the Essay to flow a little more
smoothly. This means that if readers want to appreciate fully my case against
this area of DM, they will need to
consult this material. In many
cases, I have qualified my comments (often adding much greater detail and
substantiating evidence), and I have even raised objections (some obvious, many not -- and,
indeed, some that will
have occurred to the reader) to my own arguments -- which I have then answered.
[I explain why I have adopted this tactic in
Essay One.]
If readers skip this material, then my answers to any
qualms or objections they might have will be missed, as will my expanded comments,
evidence
and clarifications. Since I have been
debating this theory with comrades for over 25 years, I have heard all the
objections there are! [Many of the more recent on-line debates have been listed here.]
It
is also worth adding that phrases like "ruling-class theory", "ruling-class view of reality",
"ruling-class ideology" (etc.) used at this site (in connection with
Traditional Philosophy and DM), aren't meant to
suggest that all or even most members of various ruling-classes
actually invented these ways of thinking or of
seeing the world (although some of them did -- for example,
Heraclitus,
Plato,
Cicero,
and
Marcus Aurelius).
They are intended to
highlight theories (or "ruling ideas") that are conducive to, or which rationalise, the
interests of the various ruling-classes history has inflicted on humanity, whoever invents them.
Up until
recently this dogmatic approach to knowledge had almost invariably been promoted by thinkers who
either relied on ruling-class patronage, or who, in one capacity or another, helped run
the system
for the elite.**
However, that will become the
central topic of Parts Two and Three of Essay Twelve (when they are published); until then, the reader is
directed
here,
here, and
here for
more
details.
[**Exactly
how this applies to DM will, of course, be explained in several other Essays
published at this site (especially
here,
here,
and here).
In addition to the three links in the previous paragraph, I have summarised the
argument (but this time tailored for absolute beginners!)
here.]
This Essay isn't
meant to be an academic study, merely an intervention in revolutionary theory.
In that case, unnecessary technicalities have
been omitted. For those who want more details, I have listed books and articles
in the End Notes that further elaborate on, or
which defend, the approach adopted at this site.
The reader must not, however, assume that I agree with everything contained in
these other works.
Throughout much
of this Essay I have blurred the distinction we should normally want to draw
between the meaning of a word and the
sense of a proposition.
A more pedantic deployment of this distinction wouldn't significantly alter many
of the conclusions reached in the
main body of this Essay, but it would merely stretch further the patience of the reader.
[I
have listed several different meanings of "meaning"
here, and
have outlined the rationale
behind the distinction between meaning and sense, here.]
Finally, I begin this Essay with a brief summary of some of the results of Essay Twelve. In that case, any
readers who find what I
have to say at the start somewhat dogmatic, controversial or unconvincing
should consult that Essay
for supporting argument and evidence. [Part
One of Essay Twelve has already been published; the unpublished material has,
however, been summarised
here.]
~~~~~~oOo~~~~~~
As of
November 2024, this Essay is just over 224,500 words long; a summary of some
of its main ideas will be posted at this site in the coming months.
The material presented
below does not represent my final view of any of the issues
raised; it is merely 'work in progress'.
Anyone using these links must remember that
they will be skipping past supporting argument and evidence set out in earlier
sections.
If your Firewall/Browser has a pop-up blocker, you will need to press the
"Ctrl" key at the same time or these and the other links here won't work!
I have adjusted the
font size used at this site to ensure that even those with impaired
vision can read what I have to say. However, if the text is still either too
big or too small for you, please adjust your browser settings!
In this Part
of Essay Thirteen I will be discussing in
much greater
detail several theories of language, 'consciousness' and 'cognition'
promoted both by Dialectical Marxists and Traditional Philosophers. In
the course of which, I will
focus
on the work of
Voloshinov
and Vygotsky
as well as others on the left who have written on these
and related topics. In addition, I will also examine the unfavourable attention certain
areas of
Wittgenstein's
work have attracted from a number of revolutionaries (which will augment what I have already
published
here). Finally, I will
also be criticising some of Chomsky's ideas in this area.
It is worth pointing from the start that
I won't be considering:
(i) The relation between language and
power;
(ii) The connection between gender and language;
(iii) Regional dialects;
(iv) The standardisation of written and
spoken language; and,
(v)
Examples of ideologically-compromised or politically offensive discourse -- i.e., racist,
sexist and reactionary speech.
That isn't because I think the above are
unimportant; far from it.
It is because several of them will be tackled in Part Seven of Essay Twelve (when it is finally published).
The rest have already been adequately addressed in books and articles written by
others on the far left. Since I don't disagree with the substantive points they make
on such issues, comment would clearly be superfluous in an Essay that is alreadyfar too long.
It will be established in Essay Twelve Part Seven
that a particular theory of language has dominated 'Western'
(and, indeed, 'Eastern') thought for over two thousand years. This approach sees the primary role of discourse (in fact, in many cases, its only
legitimate
role) as representational, and hence that it acts solely as a vehicle for thought
(or, perhaps, as an outer expression of 'inner thought'),
not as a means of communication. In fact, if discourse was
ever seen as a means of communication, it was often regarded as a vehicle for communicating thoughtsalready arrived at
independently of, and prior to, social interaction.
In fact, the
evidence shows that language was originally regarded (by
priests, theologians and philosophers, for example) as a gift of the 'gods', and
hence a
'hot line' which also allowed them to re-present their 'thoughts' to humanity.
Or,
to be more accurate,
which allowed them to be re-presented to a 'chosen' few (i.e., the
aforementioned priests, theologians and philosophers). This meant members of
this 'superior' social layer could 'process' all these 'divinely-sanctioned' thoughts on behalf of the masses,
ideas that were often expressed in obscure,
esoteric,
allegorical, poetic, figurative or highly technical language. These would then
present these
'profundities' to 'expectant humanity' as if
they had come from on high.
Indeed, as
Umberto Eco
points out (at least in relation to the 'western', Christian tradition):
"God spoke before all things,
and said, 'Let there be light.' In this way, he created both heaven and earth;
for with the utterance of the divine word, 'there was light'.... Thus Creation
itself arose through an act of speech; it is only by giving things their names
that he created them and gave them their
ontological
status.... In
Genesis..., the Lord
speaks to man for the first time.... We are not told in what language God spoke
to Adam. Tradition has pictured it as a sort of language of interior
illumination, in which God...expresses himself....
Clearly we are here
in the presence of a motif, common to other religions and mythologies -- that of
the
nomothete, the
name-giver, the creator of language." [Eco (1997), pp.7-8. Bold emphases
and links added; paragraphs merged.]
Language was therefore seen as a vehicle for the "inner illumination" of the 'soul'; a
hot-line to
'God'. Unsurprisingly, the theories concocted by countless generations of
ruling-class hacks turned out to be those that almost
invariably rationalised or 'justified' the status quo, class division,
inequality, exploitation and systematic oppression.
These
ancient 'intellectual' fantasies also implied that not only had the universe been called into existence by
the use of language,
but language
-- via the 'Word of God', the Logos -- now ran the entire show. And yet, the exclusive medium
in which much of this fairy-tale was expressed wasn't just any old language, and it
certainly wasn't the vernacular. It was a highly specialised language full of
freshly coined, jargonised expressions invented by this elite layer of theorists
so they could re-present
the 'divine' order and 'god's' thoughts to humanity. In relation to this,
ordinary discourse (that had grown out of and was based on the lives and
experience of ordinary working people) was declared completely inadequate. As the late
Professor Havelock pointed out (in connection with the
jargon concocted by Ancient Greek theorists):
"As long as preserved
communication remained oral, the environment could be described or explained
only in the guise of stories which represent it as the work of agents: that is
gods.
Hesiod takes the step of trying to unify those stories into one great
story, which becomes a cosmic theogony. A great series of matings and births of
gods is narrated to symbolise the present experience of the sky, earth, seas,
mountains, storms, rivers, and stars. His poem is the first attempt we have in a
style in which the resources of documentation have begun to intrude upon the
manner of an acoustic composition. But his account is still a narrative of
events, of 'beginnings,' that is, 'births,' as his critics the
Presocratics were to put it. From the standpoint of a sophisticated
philosophical language, such as was available to Aristotle, what was lacking
was a set of commonplace but abstract terms which by their interrelations could
describe the physical world conceptually; terms such as space, void, matter,
body, element, motion, immobility, change, permanence, substratum, quantity,
quality, dimension, unit, and the like. Aside altogether from the coinage of
abstract nouns, the conceptual task also required the elimination of verbs of
doing and acting and happening, one may even say, of living and dying, in favour
of a syntax which states permanent relationships between conceptual terms
systematically. For this purpose the required linguistic mechanism was furnished
by the timeless present of the verb to be -- the copula of analytic
statement.
"The history of early
philosophy is usually written under the assumption that this kind of vocabulary
was already available to the first Greek thinkers. The evidence of their own
language is that it was not. They had to initiate the process of inventing it....
Nevertheless, the
Presocratics could not invent such language by an act of novel creation. They
had to begin with what was available, namely, the vocabulary and syntax of
orally memorised speech, in particular the language of
Homer and
Hesiod. What they proceeded to do was to take the language of the mythos and
manipulate it, forcing its terms into fresh syntactical relationships which had
the constant effect of stretching and extending their application, giving them a
cosmic rather than a particular reference." [Havelock (1983), pp.13-14, 21.
Bold emphases added; quotation marks altered to conform with the conventions
adopted at this site. Spelling modified to agree with UK English. Links added;
several paragraphs merged.]
Subsequently, in the work of Plato,
Aristotle and
Plotinus,
for example, language became a medium that enabled the 'soul' to converse
with itself (via "inner speech"), which prompted these and subsequent
philosophers into concluding they had ready access to all those 'divine',
eternal verities,
but now derivedfrom
thought alone.1
As noted above, 'languageless thought' was regarded as the means by which the
'select few' could draw close to 'Being'/'God' -- an idea that then helped motivate the 'problem' of the relation between the
'Knower' and the 'Known',
which later re-surfaced as the main problematic of
German
Idealism. This was a class-compromised dogma that subsequently reappeared in an 'inverted' form in 'Materialist Dialectics'
where it became a key
component in addressing the alleged
relationship between 'Thought' and 'Being'.1a
In the work of early modern (and increasingly secular)
theorists, 'consciousness'
then came to refer to what supposedly took place on in an
inner, private arena where the bourgeois 'Mind'/'Soul'
--
operating now as a socially-isolated 'epistemological atom' --, could represent to itself,
not just these formerly 'divine truths', but any 'information' whatsoever (in
the form of 'impressions', 'images', 'ideas', 'concepts' or 'abstractions') that the senses
sent its way. In many cases the former ended up shaping the latter (as we
discovered in Essay Three Parts One
and Two, and Essay Twelve
Part One).
"Although the ancients
raised questions about our own knowledge of our perceptions and thought, and
introduced the idea of an inner sense, they had no word for consciousness and
they did not characterize the mind as the domain of consciousness. Aristotelians
conceived of the mind as the array of powers that distinguish humanity from the
rest of animate nature.... What is distinctive of humanity, and what
characterizes the mind, are the powers of the intellect -- of reason and of the
rational will. Knowledge of these powers is not obtained by 'consciousness' or
'introspection', but by observing their exercise in our engagement with the
world around us. The medievals followed suit. They too lacked a term for
consciousness, but they likewise indulged in reflection upon 'inner senses',
arguably -- in the wake of
Avicenna's distinguishing
five such senses -- to excess.
"Descartes's innovations with
regard to the uses in philosophy of the Latin 'conscientia' (which had not
hitherto signified consciousness at all) as well as the French 'la conscience',
were of capital importance. For it was he who introduced the novel use of the
term into the philosophical vocabulary. He invoked it in order to account for
the indubitable and infallible knowledge which he held we have of our Thoughts (cogitationes)
or Operations of the Mind. His reflections reshaped our conception of the mind
and redrew the boundaries of the mental. Thenceforth consciousness, as opposed
to intellect and sensitivity to reasons in thought, affection, intention and
action, was treated as the mark of the mental and the characteristic
of the mind.
"The expression 'conscius' and the French word 'conscient', and the
attendant conception of consciousness, caught on among his correspondents and
successors (Gassendi,
Arnauld,
La
Forge,
Malebranche). So too
'consciousness' and 'conscious' caught on among English philosophers, churchmen
and scientists (Stanley,
Tillotson,
Cumberland,
Cudworth
and
Boyle). But it is to
Locke
that we must turn
to find the most influential, fully fledged, philosophical conception of
consciousness that, with some variations, was to dominate reflection on the
nature of the human mind thenceforth. This conception was to come to its
baroque
culmination in
the writings of
Kant. In the Lockean
tradition, consciousness is an inner sense. Unlike outer sense, it is
indubitable and infallible. It is limited in its objects to the operations of
the mind. The objects of consciousness are private to each subject of experience
and thought. What one is thus conscious of in inner sense constitutes the
subjective foundation of empirical knowledge. Because consciousness is thus
confined to one's own mental operations, it was conceived to be equivalent to
self-consciousness -- understood as knowledge of how things are 'subjectively'
(privately, in foro interno ('inside the individual concerned' -- RL)) with
one's self.
"The ordinary use of the
English noun 'consciousness' and its cognates originates in the early
seventeenth century, a mere three or four decades prior to the Cartesian
introduction of a novel sense of 'conscius' and 'conscient' into philosophy in
the 1640s. So it evolved side by side with the philosophical use -- but, on the
whole, in fortunate independence of it. For the ordinary use developed, over the
next three centuries, into a valuable if specialized instrument in our toolkit
of cognitive concepts. By contrast, as we shall see, philosophical usage sank
deeper and deeper into quagmires of confusion and incoherence from which it has
not recovered to this day." [Hacker (2013a), pp.11-12. (See also the more
detailed comments on the history of this word: pp.15-19, as well as
this paper
by Hacker. (This links to a PDF.)) Italic emphases in the original; links added.]
"The term 'consciousness' is a
latecomer upon the stage of Western philosophy. The ancients had no such term.
Sunoida, like its Latin equivalent conscio, meant the same as 'I
know together with' or 'I am privy, with another, to the knowledge that'. If the
prefixes sun and cum functioned merely as intensifiers, then the
verbs meant simply 'I know well' or 'I am well aware that'. Although the
ancients did indeed raise questions about the nature of our knowledge of our own
perceptions and thought, and introduced the idea of an inner sense, they did not
characterize the mind as the domain of consciousness. Aristotelians conceived of
the mind as the array of powers that distinguish humanity from the rest of
animate nature. The powers of self-movement, of perception and sensation, and of
appetite, are shared with other animals. What is distinctive of humanity, and
what characterizes the mind, are the powers of the intellect -- of reason, and
of the rational will. Knowledge of these powers is not obtained by consciousness
or introspection, but by observation of their exercise in our engagement with
the world around us. The mediaevals followed suit. They likewise lacked any term
for consciousness, although they too indulged in reflections upon 'inner senses'
-- in the wake of Avicenna's distinguishing five such senses, arguably to
excess....
"The English word 'conscious' is
recorded by the OED [Oxford English Dictionary -- RL] as first occurring at the
beginning of the seventeenth century, when, like the Latin 'conscius', it
signified sharing knowledge with another or being witness to something. In its
early forms, it occurred in phrases such as 'being conscious to another' and
‘being conscious to something'. But sharing knowledge rapidly evolved into being
privy to unshared knowledge, either about others or about oneself. So 'to be
conscious to' quickly became a cousin to the much older expression 'to be aware
of'. The form 'to be conscious to' was slowly displaced by 'to be conscious of'.
'To be conscious of something', of course, signified a form of knowledge. So
like 'to know', 'to be conscious of something' is a
factive verb
-- one cannot be conscious of something that does not exist or is not the case.
Outside philosophy, there was no suggestion whatsoever that the objects of
consciousness, i.e. that of which one can be said to be conscious, are
restricted to one's own mental operations. One could be said to be conscious of
what one perceived, or of some feature of what one perceived, of
one's own or another's deeds -- both good and evil, of a pertinent fact (the
lateness of the hour, the merits of a case) and of one's own or another's
virtues or vices, and so forth. It was not until the middle of the nineteenth
century that 'consciousness' came to be used to signify wakefulness as opposed
to being unconscious. Thenceforth one could speak of losing and regaining
consciousness. The common or garden notions of self-consciousness, i.e. either
being excessively aware of one's appearance (a usage now lapsed) or being
embarrassingly aware that others are looking at one, is nineteenth-century
vintage. Being class conscious, money-conscious, or safety-conscious are
twentieth century coinage....
"The expression 'conscious' was
introduced into philosophy, almost inadvertently, by Descartes.
It does not appear in his work prior
to the Meditations(1641), and even there it occurs just once. In the
Third Meditation, it occurs not in relation to knowledge of one's 'thoughts' or
'operations of the mind', but in relation to awareness of the power to
perpetuate one's own existence (AT VII, 49; CSM II, 34). It was only under
pressure from objectors to this single remark that Descartes was forced, in his
'Replies to Objections', to elaborate his ideas on knowing our own 'thoughts'.
His developed position in the Principlesand late correspondence was
unstable. The expression and attendant conception, caught on among Descartes'
contemporaries and successors (Gassendi, Arnauld, La Forge) and among English
philosophers (Stanley, Tillotson, Cumberland and Cudworth). But it is to Locke,
almost fifty years later, that we must turn to find the most influential, fully
fledged, philosophical concept of consciousness that was to dominate
reflection on the nature of the human mind thenceforth. The attendant conception
was to come to its baroque culmination (or perhaps nadir of confusion) in the
writings of Kant and the post-Kantian German idealists.
"Descartes used the terms
conscientia, conscius, and conscio to signify a form of
knowledge, namely the alleged direct knowledge we have of what is passing in our
minds. What we are conscious of (which I shall call the 'objects of
consciousness') are Thoughts, a term which Descartes stretched to include
thinking (as ordinarily understood), sensing or perceiving (shorn of their
factive force), understanding, wanting, and imagining. Because he held thinking
to be the sole essential attribute of immaterial substances, he claimed that we
are thinking all the time, waking or sleeping. He also held that consciousness
of operations of the mind is indubitable and infallible. He argued that the mind
is, as it were, transparent. For, he wrote (AT VII, 214; CSM II, 150), it is
self-evident that one cannot have a thought and not be conscious of
it -- although the thoughts we have in sleep are immediately forgotten." [Hacker
(2012), pp.1-3. (This links to a PDF.) Italic emphases in the
original, links added. "AT" refers to one of the standard collections of
Descartes's work, edited by Charles Adam and Paul Tannery; "CSM" refers to the
more recent edition by John Cottingham, Robert Stoothoff, Dugald Murdoch and
Anthony Kenny. Even though the second passage of Hacker's repeats parts of the first, I have
quoted it since it adds extra important details.]
[I have said much more about the
Christian-Platonic-Cartesian Paradigm, making slightly different points in
Note 1.]
In general,
this family of theories held that such 'information' was processed by 'the mind' employing one or more of the
following:
(a) A set of 'innate' ideas;
(b) Privately applied rules
or 'habits of mind';
(c) A collection of (arbitrarily chosen) 'categories' or
'concepts', which were supposedly implanted in us by 'god', or the presence of
which was
necessitated by our psychological, 'logical', or, more recently, our genetic and
evolutionary make-up; and,
(d)
'Abstractions' that had been cobbled-together in an as-yet-to-be-explained
manner.
[Several
of the above were discussed in more detail in Essay Three Parts One and
Two (links a few paragraphs back).]
Hence, on this view, language was primarily regarded as
a means by which the inner
microcosm ('consciousness') could be put in the right intellectual order
so that it was capable of mirroring the outer macrocosm. Only then was
language allowed to function as a means of
communication. And, even then, language only served to provide outer expression to private
acts of 'intellection', cognition', or 'meaning'. 'Social meaning' was then constructed out of
these atomised
base units
-- supposedly
cobbled-together inside each individual, bourgeois skull. The social was thus an
expression of the individual,
not the other way round.
For
Rationalist
and Empiricist
Philosophers alike, in the end, truth was to be found by the individual
who examined the contents of her/his mind
-- the
difference between these two traditions now revolved around the stories their
respective ideologues told in order to turn each subjectively framed theory into an 'objective' account of
reality -- a 'reality' which, unsurprisingly, they now found rather hard to prove
actually exists!1b
Give or take a few
extra details and further complications, this is
largely how things remain to this day. The
dualism of Mind/World,
coupled with
Representationalist theories of knowledge and cognition have kept 'western'
thought permanently teetering on the edge of Idealism and Scepticism for more
than two millennia. This predicament isn't likely to alter this side of
massive social change.
[The reason for saying that is
set out in Essay Three
Part Two. Why it teeters on the
verge of Scepticism, at least as far as Dialectical Marxism is concerned, was
explained in Essay Ten
Part One.]
So,
outside the Marxist tradition, language
was seen
secondarily as a means of communication --, and even that was only so that the
private thoughts of each Social Atom might be shared with other similarly
placed Social Atoms.1c
This
dominant paradigm holds that
each 'mind' represents the world to itself
first -- perhaps constructing a private language to that end,
using "the light of reason", an inner "language
of thought", a "transformational
grammar" (now "unbounded
Merge"),
and/or a "Language
Acquisition Device" -- before it is able to convey its thoughts to other
like 'minds' trapped by the same
predicament. Indeed, only because of such inner goings on could human beings be said to have any thoughts at all
to convey to anyone else. 'Thought', on this view, wasn't a social phenomenon,
but a private, occult (hidden), and essentially individualised process or
device.1d
Which why we find that in most modern
forms of Cognitive
Theory the 'mind' is fragmented into a set of compartments, or 'processors',
each juggling with countless 'representations' --
the latter hived-off to
assorted 'modules',
now (metaphorically) seen as specialised,
deskilledpsychological subcontractors of some sort, the bourgeois social
division of labour now reproduced in the operation of the 'mental economy' at
work in each bourgeois cranium --, with
every such individual and her/his 'consciousness' reduced to the sum of these fragmented parts.2
To be sure, the view of the 'world'
that this approach
attributes to each one of us is no longer that which was intended by the
'gods', it is now that which has been
contrived by our genes. As if to cap it all, 'Evolutionary
Psychology' (henceforth, EP -- now the dominant intellectual force in this area)
of late projects the
origin of this inner bourgeois
individual (which we are all supposed to carry around in our heads) tens of thousands of years
back into the mists of time,
informing us that selfishness, individualism, male dominance, violence, the instinct to "truck and barter",
and much else besides, were
all hard-wired into our brains -- to such an extent that we would be foolish
even tothink about resisting them.2a0
Once
again, we see the status quo
under-pinned by a new set of ruling ideas, this time dressed up in the language
of Neo-Darwinism, Genetics and
Cognitive Science.
Each and everyone of us is thus
pictured as a perfectly selfish, social atom -- before we even begin to speak.
The bourgeois individual is indeed
alive and well, and living in a skull near you!
Worse still,
this particular set of ruling ideas aims to rule over all our other ideas --, it even
overshadows and dominates the doctrines invented by
erstwhile revolutionaries, as we have
seen, and will see again throughout this Essay.
Of course, as we also saw in Essay Three
Part Two, the problem here is that if
they were correct, each of these general
theories would be trapped in the private world of its inventor, with no
legitimate avenue of escape. Since no two theorists (or, indeed, human beings) can
possibly share the same ideas, communication -- given this view -- would be impossible.2a
Naturally, this only
undermines further the already shaky rationale that exists for adopting representationalism
in the first place.
The end result of all this is that
Marx and Engels's insight that language is the product of collective labour and
communal life -- and thus that its primary role lies in communication -- has never
seriously been considered, let alone adopted, even by those who claim to be
Marxists!
[Unfortunately, we saw this was true of
erstwhile Marxists in general
here; we
will witness further confirmation as this Essay unfolds.]
To that end, the ordinary language of the
working class has been distorted, depreciated and denigrated by ruling-class
hacks from ancient times onward as part of a class-motivated assault on the
vernacular.
The reason for this is plain
-- as Marx pointed out:
"The
philosophers have only
to dissolve their language into the ordinary language, from which it is
abstracted, in order to recognise it, as the distorted language of the actual
world,
and to realise that neither thoughts nor language in themselves form a
realm of their own, that they are only manifestations of actual life."
[Marx
and Engels (1970), p.118. Bold emphases added.]
As Essay Twelve Part Two will show (in the
meantime, see
here), the reason
for such disparagement is quite plain: it is impossible for anyone (let alone
Priests and Philosophers) to concoct metaphysical theories using only the vernacular.2b Hence, the vernacular had
to be declared limited at best, defective at worst, and a whole new complex and abstract terminology
was invented in its
place. This was done so that a 'hidden world' lying behind or beyond 'appearances', accessible to
'thought' alone, could be conjured into existence (as we saw
above). And, as we now know, this
approach was prosecuted in order to provide a
priori 'justification' for class division, oppression, inequality and state power.3
Representational theories still dominate
Philosophy, Psychology and Linguistics, so it isn't surprising to
see Marx's words amply confirmed in this regard, too:
"The ideas of the ruling class are in every epoch
the ruling ideas, i.e. the class which is the ruling material force of society,
is at the same time its ruling intellectual force. The class which has the means
of material production at its disposal, has control at the same time over the
means of mental production, so that thereby, generally speaking, the ideas of
those who lack the means of mental production are subject to it. The ruling
ideas are nothing more than the ideal expression of the dominant material
relationships, the dominant material relationships grasped as ideas; hence of
the relationships which make the one class the ruling one, therefore, the ideas
of its dominance. The individuals composing the ruling class possess among other
things consciousness, and therefore think. Insofar, therefore, as they rule as a
class and determine the extent and compass of an epoch, it is self-evident that
they do this in its whole range, hence among other things rule also as thinkers,
as producers of ideas, and regulate the production and distribution of the ideas
of their age: thus their ideas are the ruling ideas of the epoch...." [Marx
and Engels (1970), pp.64-65. Bold emphases added.]
However, this Essay is mainly concerned with
the ideas of those who at least give lip-service to the idea that language is a
social phenomenon and serves
primarily as a means of communication. To that end, I will begin with a brief look at how
certain Marxists have received the work of the single most important modern
champion of the social and communal approach to language:
Ludwig
Wittgenstein.
It was acknowledged in an
Additional Essay that there are
serious problems facing anyone who tries to combine Marx's and Wittgenstein's
ideas. Naturally, this doesn't mean that such a synthesis can't be achieved, but it does mean that if this is to happen it will require a
much more secure understanding of both thinkers than has hitherto been apparent.4
[TAR = The Algebra of
Revolution, or Rees (1998a); DM = Dialectical Materialism/Materialist,
depending on context.]
Having said that, there is still a high level of distrust of -- if not
resistance or open hostility shown toward --
Wittgenstein's ideas among
revolutionaries. This surfaces in TAR, for example, in the following passage:
"The social root of these
[postmodernist] ideas has been identified as the new middle class in retreat
from the values of the 1960s. But the narrower intellectual source of [such]
views is the intellectual climate in which postmodernist notions such as the
idea that '"reality" is a purely discursive phenomena, a product of various
codes, conventions, language games or signifying systems…'." [Rees (1998a),
p.297.]
While it is true that Rees is quoting
Christopher Norris here, his reference to "language games" is (intentionally or
not) clearly directed at Wittgenstein.5
Nevertheless, the puzzled reader might wonder
why there is no explicit mention in TAR of arguably the 20th
century's greatest philosopher, when numerous second-, and third-rate thinkers
receive inordinate attention. This in a book seeking to make the dialectic relevant!
That would be rather like, say, writing a history of modern Physics but forgetting to mention
Einstein,
Dirac or
Bohr.
As noted above,
revolutionaries in general have displayed a consistent level of hostility toward
Wittgenstein's ideas, a stance that hasn't always been matched by a serious attempt
to come to grips with his work -- or even summarise it accurately!
For example,
Cornforth [in Cornforth (1965)]
openly misrepresents Wittgenstein's work solely in order to rubbish it.
This is surprising since Cornforth had once been one of
Wittgenstein's close friends.
However, as is plain to anyone who bothers to
check, Cornforth confused parts of Wittgenstein's early work with that of
Russell
and
Carnap, asserting that he adopted a "verificationist" stance to "elementary
propositions" in the Tractatus, for example.
This interpretation muddles Russell's empiricist approach to such propositions
with the anti-metaphysical thrust of the Tractatus. Verificationism is completely foreign to that work.
The simple objects of the Tractatus aren't objects of possible experience, but
logical objects, as Wittgenstein himself clearly indicates. [Cf.,
Wittgenstein (1972), 2.01-2.0211,
2.023, 2.024-2.031, 4.1272. (This links to a PDF. The numbers refer to sections of the Tractatus.)]5a
Cornforth must have known this, which
perhaps explains why he offered no evidence to substantiate his wild allegations.
Little wonder, either, since there isn't any evidence; neither the word "verification", nor
any of its synonyms, occur in the Tractatus, and the entire idea is
completely at odds with Wittgenstein's own stated aims.6
Cornforth's depiction of
Wittgenstein's Tractatus is a catalogue of errors and
misrepresentations from beginning to end, to such an extent that it is doubtful whether
he actually read that
book! Or if he did, he plainly forgot much of what he had read before he pun pen
to misuse. In fact, it is abundantly clear that
Cornforth relied on second-, or
third-hand comments about the Tractatus, written by
Positivists (such as
Moritz Schlick), among others. In fact,
Cornforth only directly quotes the Tractatusonce in his five page 'summary' of it, and
even then this reference is brief and relates to the Preface alone!
Cornforth's discussion of
Wittgenstein's later work is, thankfully, less unreliable. Although he manages to
get one or two things right, he ends up confusing the method adopted in the Philosophical Investigations with that
found in Oxford
'Ordinary Language Philosophy'
(henceforth, OLP); that is, with the work of
Ryle,
Austin,
Warnock,
Strawson, Urmson and
Hampshire, etc. Beyond a few superficial
similarities, Wittgenstein's work bears no resemblance at all to "Oxford
Philosophy". [On this, see Cavell (1971a) and Dummett (1960).]
An equally inept attempt to come to grips
with Wittgenstein's work (and with OLP in general)
is to be found in Chapter Seven
of
Marcuse'sOne Dimensional Man. [Marcuse (1968).]
Unfortunately, Marcuse
made the mistake of referencing Ernest
Gellner's notorious
Words and Things [i.e., Gellner (1959)], which contains somewhat similar,
but lengthier, criticisms
of Wittgenstein and OLP (Marcuse (1968), note 2, p.141 -- i.e., note 136,
here). Gellner's execrable book won't be examined in this Essay; readers
interested in a thorough take-down of that scurrilous work, by someone who isn't
a fan of Wittgenstein or OLP, might like to check out Uschanov (2002), or,
perhaps, the longer version of that paper available
here.
[See also Dummett (1960).]6a0
Marcuse begins with this
hackneyed criticism of
both OLP and
Wittgenstein (in what follows, italic emphases are in the original, and
quotation marks have been altered to conform with the conventions adopted at
this site):
"Austin's contemptuous treatment of the alternatives to
the common usage of words, and his defamation of what we 'think up in our
armchairs of an afternoon'; Wittgenstein's assurance that philosophy 'leaves
everything as it is' -- such statements exhibit, to my mind, academic
sado-masochism, self-humiliation, and self-denunciation of the intellectual
whose labour does not issue in scientific, technical or like achievements. These
affirmations of modesty and dependence seem to recapture Hume's mood of
righteous contentment with the limitations of reason which, once recognized and
accepted, protect man from useless mental adventures but leave him perfectly
capable of orienting himself in the given environment. However, when Hume
debunked substances, he fought a powerful ideology, while his successors today
provide an intellectual justification for that which society has long since
accomplished -- namely, the defamation of alternative modes of thought which
contradict the established universe of discourse."
Added in a footnote:
"The proposition that philosophy leaves everything as it
is may be true in the context of Marx's Theses on Feuerbach (where it is
at the same time denied), or as self-characterization of neo-positivism, but as
a general proposition on philosophic thought it is incorrect." [Marcuse (1968),
pp.141-42. Quotation marks altered
to conform with the conventions adopted
at this site. Spelling altered to UK English. I have used the on-line text
here, and have corrected any typographical errors I managed to spot. The same is
true of the other passages from this book quoted below.]
I won't try to defend John Austin in this
Essay, but Marcuse clearly failed to notice that
when Wittgenstein said philosophy "leaves everything as it is" he was speaking of
the discipline as he practised it, not as it
has traditionally been pursued. Moreover, in view of the fact that Traditional Philosophy is little more than self-important hot air (on that, see Essay Twelve
Part One), except perhaps negatively,
it can't change anything, anyway.
Furthermore, Wittgenstein isn't advocating
"conformism", as Marcuse alleges. It is no more the role of philosophy to challenge
the status quo than it is the role of, say, basket weaving to challenge
advanced brain surgery. Alongside Marx
(who,
it is worth recalling,
had abandoned philosophy root and branch by the late 1840s and advised
others to do likewise),
Wittgenstein, again like Marx, would have argued that the point is in fact to change the world,
not build non-sensical
and
incoherent
philosophical theories about it. Change is the remit of
political action, science and technology, not philosophy (even if individual
philosophers might choose to involve themselves in the class struggle), as Wittgenstein
conceived it.
Here
is Wittgenstein:
"[W]hat is the use of studying
philosophy if all it does for you is to enable you to talk with some
plausibility about some abstruse questions of logic etc., if it does not
improve your thinking about the important questions of everyday life, if it does
not make you more conscientious than any...journalist in the use of the
DANGEROUS phrases such people use for their own ends." [Letter from Wittgenstein
to Malcolm, 16/11/1944, quoted in Malcolm (2001), p.93. Capitals in the
original.]
Here
is Marx:
"One has to 'leave philosophy aside'..., one has to leap out of it and
devote oneself like an ordinary man to the study of actuality...." [Marx
and Engels (1976), p.236. Bold
emphases alone added. Quotation marks altered to conform with the conventions
adopted at this site.]
"The philosophers have only interpreted the world, in various ways; the point is
to change it." [Theses
on Feuerbach.]
"It can be seen how subjectivism and objectivism, spiritualism and materialism,
activity and passivity, lose their antithetical character, and hence their
existence as such antitheses, only in the social condition; it can be seen how
the resolution of the theoretical antitheses themselves is possible
only in a practical way, only through the practical energy of man,
and how their resolution is for that reason by no means only a problem of
knowledge, but a real problem of life, a problem which philosophy
was unable to solve precisely because it treated it as a purely theoretical
problem." [Marx
(1975b), p.354.Bold emphasis added.]
"The
philosophers have only
to dissolve their language into the ordinary language, from which it is
abstracted, in order to recognise it, as the distorted language of the actual
world,
and to realise that neither thoughts nor language in themselves form a
realm of their own, that they are only manifestations of actual life."
[Marx
and Engels (1970), p.118. Bold emphasis added.]
Moreover, one only has to read the many
conversations that took place between Wittgenstein and those he gathered around
him to see that he wasn't a political quietist. Nor was he unsympathetic
to Marxism or, indeed, the gains made by the Russian Revolution. [On that, see
here.]
In
fact, Marcuse along with the vast majority of Wittgenstein critics (and, it is worth adding, many Wittgensteinians, too) misquote
or misinterpret him in this regard. Here is what
Wittgenstein actually said:
"Philosophy may in no way
interfere with the actual use of language; it can in the end only describe it.
For it cannot give any foundation either. It leaves everything as it is. It also
leaves mathematics as it is, and no mathematical discovery can advance it."
[Wittgenstein (1958), §124, page 49e.]
From this it is quite clear that the word
"everything" refers back to "the actual use of language". This is plain from the fact that he then goes
on to mention mathematics ("It also leaves mathematics as it is"),
which he wouldn't have added if "everything" were totally unqualified in
the way that many now suppose. So, philosophy leaves everything in language and
mathematics as they are, but (by default) nothing else. Whether or not one agrees with
Wittgenstein, this passage offers no support to those who characterise Wittgenstein as a
conservative.
Incidentally, the most recent translation of the above passage reads as follows:
(1) "Philosophy must not
interfere in any way with the actual use of language, so it can in the end only
describe it.
(2) "For it cannot justify it
either.
(3) "It leaves everything as it
is.
(4) "It also leaves mathematics
as it is, and no mathematical discovery can advance it." [Wittgenstein (2009), §124,
p.55e. (This links to a PDF.) Numbers added.]
If anything,
that is even clearer. As I have pointed out elsewhere at this
site:
Once more, it is clear that the word "everything",
in Line 3, is restricted to "the actual use of language", and that can be seen
from the next sentence. If "everything" were totally unrestricted, there would
be no point adding the extra caveat in Line 4. Why inform us that Philosophy
"also" leaves mathematics as it is if we have just been told it leaves
"everything" as it is? If the meaning of "everything" in Line 2 is
unrestricted, then mathematics can't be part of "everything", can it? So,
either Wittgenstein was a sloppy stylist (who wants or admit that?), or
he held rather odd beliefs about mathematics.
So, in Line 3, Wittgenstein is talking about "the
actual use of language", not
absolutely everything -- from politics to pottery, science to sociology,
quasars to quarks! The continual use of "it" should have alerted commentators to
this rather simple point (here coloured appropriately to assist the unconvinced
see the light):
(1) "Philosophy
must not interfere in any way with the actual use of
language, so
it can in the end only describe
it.
(2) "For it
cannot justify
it
either.
(3) "It
leaves
everything as it
is.
(4)
"It also leaves
mathematics as it is, and no
mathematical discovery can advance
it."
The red "it" clearly refers back to "philosophy",
but it is no less clear that the purple "it" refers back to "the actual use of
language", which means that "everything as it is" is correctly coloured purple,
too.
Furthermore,
we have on record the following interaction between Norman
Malcolm and Wittgenstein:
"[W]hat is the
use of studying philosophy if all it does for you is to enable you to talk with
some plausibility about some abstruse questions of logic etc., [and] if it does
not improve your thinking about the important questions of everyday life, if it
does not make you more conscientious than any...journalist in the use of the
DANGEROUS phrases such people use for their own ends." [Letter from Wittgenstein
to Malcolm, 16/11/1944, quoted in Malcolm (2001), p.93. Capitalisation in the
original; bold added.]
[Here Wittgenstein is clearly referring Malcolm to
his new conception of philosophy.]
This doesn't sound like the remark of a
'philosophical quietist', as he has often been portrayed by many academics and
activists on the left.
Now, in
line with the traditional contempt
shown by ruling-class theorists toward the vernacular and the thoughts of ordinary workers, Marcuse argues
as follows:
"Throughout the
work of the linguistic analysts, there is this familiarity with the chap on the
street whose talk plays such a leading role in linguistic philosophy. The
chumminess of speech is essential inasmuch as it excludes from the beginning the
high-brow vocabulary of 'metaphysics;' it militates against intelligent
non-conformity; it ridicules the egghead. The language of John Doe and Richard
Roe is the language which the man on the street actually speaks; it is the
language which expresses his behaviour; it is therefore the token of
concreteness. However, it is also the token of a false concreteness. The
language which provides most of the material for the analysis is a purged
language, purged not only of its 'unorthodox' vocabulary, but also of the means
for expressing any other contents than those furnished to the individuals by
their society. The linguistic analyst finds this purged language an accomplished
fact, and he takes the impoverished language as he finds it, insulating it from
that which is not expressed in it although it enters the established universe of
discourse as element and factor of meaning.
"Paying respect
to the prevailing variety of meanings and usages, to the power and common sense
of ordinary speech, while blocking (as extraneous material) analysis of what
this speech says about the society that speaks it, linguistic philosophy
suppresses once more what is continually suppressed in this universe of
discourse and behaviour. The authority of philosophy gives its blessing to the
forces which make this universe. Linguistic analysis abstracts from what
ordinary language reveals in speaking as it does -- the mutilation of man and
nature." [Marcuse (1968),
pp.142-43.]
From this it is quite plain that Marcuse
prefers the obscure and impenetrable jargon that ruling-class hacks regularly inflict on their
readers to the language of ordinary workers, and it isn't hard to see why.
Indeed, as was alleged above, Marcuse all but concedes that it is impossible to derive the empty theses of Traditional Philosophy
("metaphysics") if theorists confine themselves to the vernacular. [On this, see
Essay Twelve Part One.] And that is why he complains that
the language used by Wittgenstein, and others, has been "purged" of the
very jargon
upon which traditionalists like Marcuse dote -- which "purge" is in fact a move in
the right direction since it would prevent them from even attempting to
perform their
verbal tricks.
Arguing in this way, Marcuse plainly disagrees with Marx himself
(quoted earlier):
"The philosophers have only
to dissolve their language into the ordinary language, from which it is
abstracted, in order to recognise it, as the distorted language of the actual
world, and to realise that neither thoughts nor language in themselves form a
realm of their own, that they are only manifestations of actual life."
[Marx
and Engels (1970), p.118. Bold emphases added.]
It is also worth pointing out that,
again like many
others, Marcuse
has confused ordinary language with "common sense". As we
have seen, these two aren't at all the same. [On that, cf., Hallett (2008),
pp.91-99.] Moreover, Marcuse is wrong in what he says about "eggheads" -- in
fact, in all my years of studying Wittgenstein and OLP (to date, at least 40 years), I have yet to encounter anything that
remotely suggests this reading. It isn't surprising, therefore, to find that
Marcuse fails to quote, or even cite, a single passage in support of his wild allegations.
Furthermore, neither the OLP-ers nor
Wittgenstein raised objections against other uses of language, they
simply point out that it is a serious error to suppose one can answer
questions about knowledge, perception, time, space, thought, action, etc., by
using words in technical, or in other odd ways (a point Marx also made).
As Hanjo Glock notes:
"Wittgenstein's ambitious claim is that it is
constitutive of metaphysical theories and questions that their employment of
terms is at odds with their explanations and that they use deviant rules along
with the ordinary ones. As a result, traditional philosophers cannot coherently
explain the meaning of their questions and theories. They are confronted with a
trilemma: either their novel uses of terms remain unexplained
(unintelligibility), or...[they use] incompatible rules (inconsistency), or
their consistent employment of new concepts simply passes by the ordinary use --
including the standard use of technical terms -- and hence the concepts in terms
of which the philosophical problems were phrased." [Glock (1996), pp.261-62.
See also
this
quotation,
and
my comments in Essay Thirteen
Part One, as well as those I have posted at Wikipedia (here
and
here)
concerning the use of technical
terms in science.]
"For two and a half millennia some of the best minds in
European culture have wrestled with the problems of philosophy. If one were to
ask what knowledge has been achieved throughout these twenty-five centuries,
what theories have been established (on the model of well-confirmed theories in
the natural sciences), what laws have been discovered (on the model of the laws
of physics and chemistry), or where one can find the corpus of philosophical
propositions known to be true, silence must surely ensue. For there is no body of
philosophical knowledge. There are no well-established philosophical theories or
laws. And there are no philosophical handbooks on the model of handbooks of
dynamics or of biochemistry. To be sure, it is tempting for contemporary
philosophers, convinced they are hot on the trail of the truths and theories
which so long evaded the grasp of their forefathers, to claim that philosophy
has only just struggled out of its early stage into maturity.... We can at long
last expect a flood of new, startling and satisfying results -- tomorrow.
"One can blow the Last Trumpet once, not once a
century. In the seventeenth century Descartes thought he had discovered the
definitive method for attaining philosophical truths; in the eighteenth century
Kant believed that he had set metaphysics upon the true path of a science; in
the nineteenth century Hegel convinced himself that he had brought the history
of thought to its culmination; and Russell, early in the twentieth century,
claimed that he had at last found the correct scientific method in philosophy,
which would assure the subject the kind of steady progress that is attained by
the natural sciences. One may well harbour doubts about further millenarian
promises." [Hacker (2001c), pp.322-23.]
Comrades like Marcuse are welcome to this
monumental waste of ink and paper (to which Hacker alludes) -- and that
comment applies even more so to 'dialectical philosophy', which is definitely the poor relation
of this long slow detour
to
nowhere.
What of this, though?
"Moreover, all too often it
is not even the ordinary language which guides the analysis, but rather blown-up
atoms of language, silly scraps of speech that sound like baby talk such as
'This looks to me now like a man eating poppies,' 'He saw a robin', 'I had a
hat.' Wittgenstein devotes much acumen and spare to the analysis of 'My broom is
in the corner.'" [Marcuse (1968),
p.143.]
But, does Marcuse take Hegel or Engels to
task for their use of "The rose is red" (on that, see here
and here), or Lenin for his employment of "John is
a man"? Not a bit of it! In fact, Marcuse misses the point of using such simple
language: If we can't
get the latter right, we stand no chance with more complex propositions or
bodies of text. Indeed,
as we have seen (for example,
here, here
and here),
dialecticians can't even get "John is a man" right! [Which rather makes
my
point for me, one feels.]
However, Marcuse has an answer to this:
"To take another
illustration: sentences such as 'my broom is in the corner' might also occur in
Hegel's Logic, but there they would be revealed as inappropriate or even false
examples. They would only be rejects, to be surpassed by a discourse which, in
its concepts, style, and syntax, is of a different order -- a discourse for
which it is by no means 'clear that every sentence in our language is in order
as it is."' Rather the exact opposite is the case -- namely, that every sentence is
as little in order as the world is which this language communicates." [Ibid.,
p.144. Bold emphasis added.
Quotation marks altered to conform with the conventions adopted at this site.]
But, if the above were indeed so -- if "every sentence is as
little in order as the world is which this language communicates" then the
ordinary words and sentences Marcuse himself usescan't be "in order",
either, which means we can't take what they say at face value. [But, is
there another, deeper significance to his words?] We have
already seen that
attempts to argue that ordinary language is in some way (or in any way)
defective back-fire on anyone foolish enough to try. But, here we encounter the same
reckless bravado, for if Marcuse's words aren't "in order", what can they possibly
mean?
As Marcuse notes on the same page:
"Thus the analysis does not
terminate in the universe of ordinary discourse, it goes beyond it and opens a
qualitatively different universe, the terms of which may even contradict the
ordinary one." [Ibid.,
p.144.]
Except that here the tables are turned on
Marcuse, for if we analyse his words, and are able to follow his argument, we see that
(if he were correct) his
words would imply the opposite of what he intended -- that is, our ability to
comprehend what he says shows that his words
are in the "right order" and hence we can understand him after all! And yet, as soon as we
succeed in understanding
what he is telling us, we immediately see that
his words aren't in fact in the "right order", for he tells us that
none are!
--
"every sentence is as little in order as the world is which this language
communicates"
--,
and that they make no sense, therefore. [Yet another ironic 'dialectical inversion', one
feels.]
Next
we encounter this hackneyed criticism -- Marcuse (quoting Wittgenstein):
"The almost
masochistic reduction of speech to the humble and common is made into a program:
'if the words "language", "experience", "world", have a use, it must be as
humble a one as that of the words "table", "lamp", door."' We must 'stick to the
subjects of our every-day thinking, and not go astray and imagine that we have
to describe extreme subtleties...' -- as if this were the only alternative, and
as if the 'extreme subtleties' were not the suitable term for Wittgenstein's
language games rather than for Kant's Critique of Pure Reason. Thinking
(or at least its expression) is not only pressed into the straitjacket of common
usage, but also enjoined not to ask and seek solutions beyond those that are
already there. 'The problems are solved, not by giving new information,
but by arranging what we have always known.'
"The
self-styled poverty of philosophy, committed with all its concepts to the given
state of affairs, distrusts the possibility of a new experience. Subjection to
the rule of the established fact is total -- only linguistic facts, to be sure,
but the society speaks in its language, and we are told to obey. The
prohibitions are severe and authoritarian: 'Philosophy may in no way interfere
with the actual use of language.' 'And we may not advance any kind of theory.
There must not be anything hypothetical in our considerations. We must do
away with all explanation, and description alone must take its place.'
"One might ask
what remains of philosophy? What remains of thinking, intelligence, without
anything hypothetical, without any explanation? However, what is at stake is not
the definition or the dignity of philosophy. It is rather the chance of
preserving and protecting the fight, the need to think and speak in terms
other than those of common usage -- terms which are meaningful, rational, and
valid precisely because they are other terms. What is involved is the spread of
a new ideology which undertakes to describe what is happening (and meant) by
eliminating the concepts capable of understanding what is happening (and
meant)." [Ibid.,
pp.144-45.]
Marcuse has worked himself up into a right
old lather here, all the while missing the point. Once more, Wittgenstein was speaking here of his
new approach to philosophy, which, if correct, would mean that
traditional forms-of-thought, beloved of characters like Marcuse, are nothing
more than elaborate, insubstantial
"houses of cards". Wittgenstein is certainly not arguing against
"anything
hypothetical", or against "explanation" in other areas of theory (for example,
in science -- indeed, he developed a novel account of what it is
to reason hypothetically). Once more, in his haste to malign Wittgenstein,
Marcuse only succeeded in aiming a few blows at thin air.
And, far from the following being true,
the opposite is in fact the case:
"It is rather the chance of
preserving and protecting the fight, the need to think and speak in terms
other than those of common usage -- terms which are meaningful, rational, and
valid precisely because they are other terms. What is involved is the spread of
a new ideology which undertakes to describe what is happening (and meant) by
eliminating the concepts capable of understanding what is happening (and
meant)." [Ibid.]
The
obscure terminology that litters the pages of Traditional Thought, and particularly the
impenetrable jargon Hegel inflicted on his readers, actually prevents us
from understanding the world. As pointed out in Essay Twelve
Part One, the influence of Traditional Philosophy must be terminated in
order to facilitate the advance of scientific knowledge in general, and Marxism
in particular. [Here, of course, I am very loosely paraphrasing Kant.]
Marcuse's failure to get the point is further
underlined by this blindingly irrelevant comment:
"To begin with, an
irreducible difference exists between the universe of everyday thinking and
language on the one side, and that of philosophic thinking and language on the
other. In normal circumstances, ordinary language is indeed behavioural -- a
practical instrument. When somebody actually says 'My broom is in the corner,'
he probably intends that somebody else who had actually asked about the broom is
going to take it or leave it there, is going to be satisfied, or angry. In any
case, the sentence has fulfilled its function by causing a behavioural reaction:
'the effect devours the cause; the end absorbs the means.'" [Ibid.,
pp.145-46.]
Marcuse plainly didn't know -- perhaps because of his characteristically sloppy
research --, that when Wittgenstein used the sentence "My broom is in the
corner" [Wittgenstein
(2009), §60, p.33e (this links to a PDF)] he
was in fact criticising a view he himself had adopted in the Tractatus -- about
(i) The nature of logically simple names, (ii) The idea that a fact is a complex,
and (iii) The thesis that analysis is capable of revealing logical form, etc. [Wittgenstein (1972),
2-3.263, pp.7-25, and 5.5423, p.111 (this links to a PDF). On the background to this, see White (1974,
2006). On Investigations §§37-61
(the relevant sections), see Baker and Hacker (2005b), pp.112-42, Hallett
(1977), pp.112-39, and Hallett (2008),
pp.33-41.]
So, Wittgenstein
was here advancing a profound criticism of his earlier
way of seeing things. Now, whether or not one agrees with Wittgenstein (before or
after he changed mind -- or even at all!), the issues he raises aren't of
the everyday "behavioural" sort that Marcuse seems to think; they
concern the
logical nature of propositions and how they can be used to represent the world (that is,
if they can). These are hardly trivial.
But, there is more:
"In contrast, if, in a
philosophic text or discourse, the word 'substance,' 'idea,' 'man,' 'alienation'
becomes the subject of a proposition, no such transformation of meaning into a
behavioural reaction takes place or is intended to take place. The word remains,
as it were, unfulfilled -- except in thought, where it may give rise to other
thoughts. And through a long series of mediations within a historical continuum,
the proposition may help to form and guide a practice. But the proposition
remains unfulfilled even then -- only the hubris of absolute idealism asserts
the thesis of a final identity between thought and its object. The words with
which philosophy is concerned can therefore never have a use 'as humble ... as
that of the words "table", "lamp", "door"'.
"Thus, exactness and clarity
in philosophy cannot be attained within the universe of ordinary discourse. The
philosophic concepts aim at a dimension of fact and meaning which elucidates the
atomized phrases or words of ordinary discourse 'from without' by showing this
'without' as essential to the understanding of ordinary discourse. Or, if the
universe of ordinary discourse itself becomes the object of philosophic
analysis, the language of philosophy becomes a 'meta-language.' Even where it
moves in the humble terms of ordinary discourse, it remains antagonistic. It
dissolves the established experiential context of meaning into that of its
reality; it abstracts from the immediate concreteness in order to attain true
concreteness." [Ibid.,
p.146.]
Once again, as we have also seen, it is in
fact the obscure jargon, which
litters Traditional Philosophy, that
undermines clarity of
thought. In which case, it is no surprise to discover that, far from
constituting a "guide" to practice, dialectics has been roundly refuted by it.
[On that, see Essay Ten Part One.]
Moreover, as far as 'abstraction' is
concerned, Marcuse just helps himself to this word without any attempt to
explain the obscure 'process' that supposedly underlies it -- or, indeed, any attempt to show how it is even
possible to 'abstract' anything at all.
[On that, see Essay Three Parts One
and Two.]
"Ordinary language in its
'humble use' may indeed be of vital concern to critical philosophic thought, but
in the medium of this thought words lose their plain humility and reveal that
'hidden' something which is of no interest to Wittgenstein. Consider the
analysis of the 'here' and 'now' in Hegel's Phenomenology, or...Lenin's
suggestion on how to analyze adequately 'this glass of water' on the table. Such
an analysis uncovers the history in every-day speech as a hidden
dimension of meaning -- the rule of society over its language. And this
discovery shatters the natural and reified form in which the given universe of
discourse first appeals. The words reveal themselves as genuine terms not only
in a grammatical and formal-logical but also material sense; namely, as the
limits which define the meaning and its development -- the terms which society
imposes on discourse, and on behaviour. This historical dimension of meaning can
no longer be elucidated by examples such as 'my broom is in the corner' or
'there is cheese on the table.' To be sure, such statements can reveal many
ambiguities, puzzles, oddities, but they are all in the same realm of language games
and academic boredom." [Ibid.,
pp.147-48. I have corrected several serious typographical errors in the
on-line version of the last sentence.]
As we will see in Essay Twelve, Hegel's crass
analysis of spatial and temporal
indexicals (i.e., "here" and "now") is hardly a reassuring
advert for the
'superiority' of DL
over FL.
Moreover, we have already seen what
a mess Lenin got himself into with his 'analysis' of glass tumblers, using DL. In which
case, the alleged 'banalities of ordinary language' are much to be preferred over
the irredeemable confusion that has for two centuries oozed out of
Hegel's
Hermetic House of Horrors, clogging the minds of comrades like Marcuse.
Indeed, science has about as much to learn from this backwater of
Neoplatonic
Mysticism as it has from
dowsing or
crystal
gazing.
[FL = Formal Logic; DL =
Dialectical Logic.]
It is
also revealing to discover that Marcuse has an unhealthy interest in what is
"hidden" -- revealing because we have also seen that
it is a cornerstone of ruling-class ideology that there is a "hidden"
'reality' lying behind "appearances", which is accessible to thought alone and
which is more real than the world we see around us. Marcuse reveals yet again that even though he pretends to be a radical, he is nonetheless a philosophical conservative,
happy to ape the thought-forms of the last two-and-a-half millennia of boss-class
theory. [On that, see
these
comments from Essay Two.]
Marcuse continues:
"The
therapeutic character of the philosophic analysis is strongly emphasized -- to
cure from illusions, deceptions, obscurities, unsolvable riddles, unanswerable
questions, from ghosts and spectres. Who is the patient? Apparently a certain
sort of intellectual, whose mind and language do not conform to the terms of
ordinary discourse. There is indeed a goodly portion of psychoanalysis in this
philosophy -- analysis without Freud's fundamental insight that the patient's
trouble is rooted in a general sickness which cannot be cured by analytic
therapy. Or, in a sense, according to Freud, the patient's disease is a protest
reaction against the sick world in which he lives. But the physician must
disregard the 'moral' problem. He has to restore the patient's health, to make
him capable of functioning normally in his world.
"The
philosopher is not a physician; his job is not to cure individuals but to
comprehend the world in which they live -- to understand it in terms of what it
has done to man, and what it can do to man. For philosophy is (historically,
and its history is still valid) the contrary of what Wittgenstein made it out to
be when he proclaimed it as the renunciation of all theory, as the undertaking
that 'leaves everything as it is.' And philosophy knows of no more useless
'discovery' than that which 'gives philosophy peace, so that it is no longer
tormented by questions which bring itself in question.'" [Ibid.,
p.149.]
Here, Marcuse
unambiguously nails his colours to the mast,
for he is perfectly happy to assume the role Traditional Philosophers have
always arrogated to themselves (i.e., of possessing unique access to the
aforementioned, hidden
Super-Truths,
available to thought alone, but oddly enough not to scientists), even though
these Super-Theorists have yet to solve a single problem in over 2500 years (as
Peter Hacker
pointed out). In fact,
they have yet to agree what a 'solution' would even look like -- or even what the
right questions are! Marcuse makes no attempt to defend this age-old
view of Philosophy, except he appeals to the fact that it has always been
regarded
this way. So much for his radical credentials! But, how is it that pure
thought is able to gain such easy access to this 'hidden world'? No good looking
to Marcuse for an answer to that one; he is conspicuously silent about it.
Is
the philosopher a physician? Not if she follows in Marcuse's footsteps and
confuses Wittgenstein's new approach to the subject with Traditional Thought.
[On this aspect of
Wittgenstein's work, see Fischer (2011a, 2011b).]
Once again, Marcuse has a 'reply':
"This intellectual
dissolution and even subversion of the given facts is the historical task of
philosophy and the philosophic dimension. Scientific method, too, goes beyond
the facts and even against the facts of immediate experience. Scientific method
develops in the tension between appearance and reality. The mediation
between the subject and object of thought, however, is essentially different. In
science, the medium is the observing, measuring, calculating, experimenting
subject divested of all other qualities; the abstract subject projects and
defines the abstract object." [Ibid., p.150. Bold emphasis added.]
This appears to be an echo of Marx's claim:
"Vulgar economy actually does no more than
interpret, systematise and defend in doctrinaire fashion the conceptions of the
agents of bourgeois production who are entrapped in bourgeois production
relations. It should not astonish us, then, that vulgar economy feels
particularly at home in the estranged outward appearances of economic relations
in which these prima facie absurd and perfect contradictions appear and
that these relations seem the more self-evident the more their internal
relationships are concealed from it, although they are understandable to the
popular mind. But all science would be superfluous if the outward appearance
and the essence of things directly coincided." [Marx
(1981), p.956. Bold emphasis added.]
[I have dealt with this view of science in Essay
Twelve Part
One; the reader is directed there for more details. In addition, the distinction
Traditional Philosophers have drawn between "appearance" and "reality"
was
criticised in Essay Three
Part Two.]
Even so, as if to confirm an earlier
allegation (that the jargon and
thought-forms of Traditional Philosophy have thoroughly compromised the brains
of far too many Dialectical Marxists), Marcuse kindly provides us with yet more
evidence, presenting his readers with this prize example of
academic gobbledygook:
"In contrast, the objects of
philosophic thought are related to a consciousness for which the concrete
qualities enter into the concepts and into their interrelation. The philosophic
concepts retain and explicate the pre-scientific mediations (the work of
everyday practice, of economic organization, of political action) which have
made the object-world that which actually is -- a world in which all facts are
events, occurrences in a historical continuum." [Marcuse (1968),
pp.150-51.]
How
Marcuse knew all this a priori
psychology (about what is or what isn't related to "consciousness", or what does
or does not enter into "the concepts" and their "interrelation") he
unwisely kept to
himself. But, this is just par for the course; as
we have seen, every single dialectician does
likewise: they impose their dogmatic theories on reality and on human
'cognition' in a
thoroughly traditional manner, all the while claiming that this is precisely
what they aren't
doing!
"The separation of science
from philosophy is itself a historical event. Aristotelian physics was a part of
philosophy and, as such, preparatory to the 'first science' -- ontology. The
Aristotelian concept of matter is distinguished from the Galilean and
post-Galilean not only in terms of different stages in the development of
scientific method (and in the discovery of different 'layers' of reality), but
also, and perhaps primarily, in terms of different historical projects, of a
different historical enterprise which established a different nature as well as
society. Aristotelian physics becomes objectively wrong with the new
experience and apprehension of nature, with the historical establishment of a
new subject and object-world, and the falsification of Aristotelian physics then
extends backward into the past and surpassed experience and apprehension."
[Ibid.,
p.151. Emphasis in the original.]
Marcuse is right here; as we will see
in Essay Thirteen Part Two, the sciences gradually separated themselves from
Traditional Thought as scientists took the material world increasingly into
consideration in the formation of theory (through careful observation and
experiment), progressively (but not
completely) abandoning the Super-Scientific
approach championed by this Ancient Tradition. [Of course, the situation is far
more complex than these brief comments might suggest. A useful summary of the
approach I intend to take about the relation between theoretical and
experimental science can be found in Lerner (1991) --
although the reader mustn't assume I agree with everything Lerner has to say.] But, the
separation Marcuse mentions is in fact to the detriment of Traditional Thought,
which now reflects in a more pure form ruling-class ideology, an approach to a
priori theory Marcuse
unwisely sought to emulate.
"...However it is fair to say
that the most abstruse metaphysics has not exhibited such artificial and
jargonic worries as those which have arisen in connection with the problems of
reduction, translation, description, denotation, proper names, etc. Examples are
skilfully held in balance between seriousness and the joke: the differences
between Scott and the author of Waverly; the baldness of the present king
of France; Joe Doe meeting or not meeting the 'average taxpayer' Richard Roe on
the street; my seeing here and now a patch of red and saying 'this is red;' or
the revelation of the fact that people often describe feelings as thrills,
twinges, pangs, throbs, wrenches, itches, prickings, chills, glows, loads,
qualms, hankerings, curdlings, sinkings, tensions, gnawings and shocks." [Ibid.,
pp.151-52.]
However, the point of all this
is that unless we are capable of understanding the logic of simple language like this we stand
no chance with the impenetrable jargon Hegel and other philosophers
inflict on their readers.
Now, while Marcuse makes a more substantive point
(that the sort of analyses one finds in early
Analytic Philosophy
-- concerning, say,
Russell's Theory of Descriptions, which was the point of him mentioning
Waverly
and the King of France, by the way -- are even more jargonised than Traditional
Metaphysics), he failed to cite any examples. Be this as it may, the technical
language employed by Analytic Philosophers is easy to paraphrase in more
ordinary terms. The same can't be said for the gobbledygook that holds
Marcuse in its thrall. [Concerning the latter,
these comment of Chomsky's are entirely apposite.]
"In cleaning up this mess,
analytic philosophy conceptualizes the behaviour in the present technological
organization of reality, but it also accepts the verdicts of this organization;
the debunking of an old ideology becomes part of a new ideology. Not only the
illusions are debunked but also the truth in those illusions. The new ideology
finds its expression in such statements as 'philosophy only states what everyone
admits,' or that our common stock of words embodies 'the distinctions men have
found worth drawing.'" [Ibid.,
p.152.]
The two quotations at the end of this passage
are taken from Wittgenstein and John Austin respectively. Wittgenstein's point is
that the theses philosophers concoct are very often misconstrued rules of language.
As such, language users readily acknowledge their triviality when they have been stated
clearly (since they use such rules every day) -- as I have shown in Essay Twelve
Part One.
Moreover, Austin's comment is in
the past tense, and doesn't refer to the future. New distinctions are always
possible.
But, what about this?
"What is this 'common
stock'? Does it include Plato's 'idea,' Aristotle's 'essence,' Hegel's Geist,
Marx's Verdinglichung in whatever adequate translation? Does it
include the key words of poetic language? Of surrealist prose? And if so, does
it contain them in their negative connotation; that is, as invalidating the
universe of common usage? If not, then a whole body of distinctions which men
have found worth drawing is rejected, removed into the realm of fiction or
mythology; a mutilated, false consciousness is set up as the true consciousness
that decides on the meaning and expression of that which is. The rest is
denounced -- and endorsed -- as fiction or mythology."
[Ibid.,
p.152.]
The point is that unless the expressions
Marcuse lists (Plato's "idea", Hegel's "Geist", etc.) can be paraphrased
in ordinary language, they can't be used to help draw a single distinction (in that word's
ordinary sense), to begin with. As we
have seen, metaphysical jargon like this is devoid of meaning (or, rather,
it can only be 'explained' in terms of yet more empty jargon) and serves to
confuse only those naive enough to take it seriously. However, Marcuse's point about poetry is misplaced,
too, since neither Austin nor Wittgenstein would have wished to deny its
literary merit. I suspect this point has been lumped in here because of Marcuse's
penchant for making unfounded and sweeping allegations about a tradition in modern
philosophy he plainly struggled to comprehend.
Once more, Marcuse has an 'answer':
"Analytic philosophy often
spreads the atmosphere of denunciation and investigation by committee. The
intellectual is called on the carpet. What do you mean when you say....? Don't
you conceal something? You talk a language which is suspect. You don't talk like
the rest of us, like the man in the street, but rather like a foreigner who does
not belong here. We have to cut you down to size, expose your tricks, purge you.
We shall teach you to say what you have in mind, to 'come clear,' to 'put your
cards on the table.' Of course, we do not impose on you and your freedom of
thought and speech; you may think as you like. But once you speak, you have to
communicate your thoughts to us -- in our language or in yours. Certainly, you
may speak your own language, but it must be translatable, and it will be
translated. You may speak poetry -- that is all fight. We love poetry. But we
want to understand your poetry, and we can do so only if we can interpret your
symbols, metaphors, and images in terms of ordinary language.
"The poet might answer that
indeed he wants his poetry to be understandable and understood (that is why he
writes it), but if what he says could be said in terms of ordinary language he
would probably have done so in the first place. He might say: Understanding of
my poetry presupposes the collapse and invalidation of precisely that universe
of discourse and behaviour into which you want to translate it. My language can
be learned like any other language (in point of fact, it is also your own
language), then it will appear that my symbols, metaphors, etc. are not
symbols, metaphors, etc. but mean exactly what they say. Your tolerance is
deceptive. In reserving for me a special niche of meaning and significance, you
grant me exemption from sanity and reason, but in my view, the madhouse is
somewhere else." [Ibid.,
pp.155-56. Typo corrected.]
Readers will look long and hard, and to no
avail, in the writings of Austin or Wittgenstein for a single passage that even remotely resembles
this mendacious caricature of their view of poetry. Nowhere do they insist (or
even imply) that
what a poet has to say should be translated into ordinary language, which is,
of course,
why Marcuse failed to quote either of them to that effect.
However, unless Marcuse and other
metaphysicians are capable of making themselves clear (and actually manage to do
just that), they haven't in fact succeeded in
saying anything determinate. The onus is therefore on them to make their ideas
and theses
comprehensible. Marcuse blames Analytic Philosophers for pointing this out, as
if it is their
fault that he and others like him speak and write in riddles, or insist
on producing gobbledygook
by the cartload. That makes about as much sense as blaming the boy -- in the
famous story by
Hans Christian Andersen -- for pointing out that the
Emperor was naked!
At this point, Marcuse compounds his errors by
turning his fire on ordinary language (again, in a thoroughly traditional manner):
"But critical
analysis must dissociate itself from that which it strives to comprehend; the
philosophic terms must be other than the ordinary ones in order to elucidate the
full meaning of the latter. For the established universe of discourse bears
throughout the marks of the specific modes of domination, organization, and
manipulation to which the members of a society are subjected. People depend for
their living on bosses and politicians and jobs and neighbours who make them
speak and mean as they do; they are compelled, by societal necessity, to
identify the 'thing' (including their own person, mind, feeling) with its
functions. How do we know? Because we watch television, listen to the radio,
read the newspapers and magazines, talk to people.
"Under these
circumstances, the spoken phrase is an expression of the individual who speaks
it, and of those who make him speak as he does, and of whatever
tension or contradiction may interrelate them. In speaking their own language,
people also speak the language of their masters, benefactors, advertisers. Thus
they do not only express themselves, their own knowledge, feelings, and
aspirations, but also something other than themselves. Describing 'by
themselves' the political situation, either in their home town or in the
international scene, they (and 'they' includes us, the intellectuals who
know it and criticize it) describe what 'their' media of mass communication tell
them -- and this merges with what they really think and see and feel.
"Describing to
each other our loves and hatreds, sentiments and resentments, we must use the
terms of our advertisements, movies, politicians and best sellers. We must use
the same terms for describing our automobiles, foods and furniture, colleagues
and competitors -- and we understand each other perfectly. This must necessarily
be so, for language is nothing private and personal, or rather the private and
personal is mediated by the available linguistic material, which is societal
material. But this situation disqualifies ordinary language from fulfilling the
validating function which it performs in analytic philosophy. 'What people mean
when they say...' is related to what they don't say. Or, what they mean
cannot be taken at face value -- not because they lie, but because the universe of
thought and practice in which they live is a universe of manipulated
contradictions.
"Circumstances
like these may be irrelevant for the analysis of such statements as 'I itch,' or
'he eats poppies,' or 'this now looks red to me,' but they may become vitally
relevant where people really say something ('she just loved him,' 'he has no
heart,' 'this is not fair,' 'what can I do about it?'), and they are vital for
the linguistic analysis of ethics, politics, etc. Short of it, linguistic
analysis can achieve no other empirical exactness than that exacted from the
people by the given state of affairs, and no other clarity than that which is
permitted them in this state of affairs -- that is, it remains within the limits
of mystified and deceptive discourse."
[Ibid.,
pp.156-57. Emphases in the original.]
This
once again reveals Marcuse's
contempt for ordinary workers and their forms of discourse. To be sure, their
use of language can be corrupted in the way he says, but for every class-compromised sentence that can be uttered in ordinary language, there exists
its negation. That is why socialists can say things like "Women aren't
inferior", "Jews aren't sub-human", "Capitalism isn't fair",
"Gays aren't perverts!", etc. [More on that,
here. This
topic will be taken up again in detail in Essay Twelve Part Seven.]
It is
also clear that. like so may others, Marcuse has run-together speaker's
meaning with word meaning (a
topic I take up again later in much more detail); that can be seen in the
following passages:
"People depend for their living on
bosses and politicians and jobs and neighbours who make
them speak and mean as they do; they are compelled, by societal
necessity, to identify the 'thing' (including their own person, mind, feeling)
with its functions....
"In speaking
their own language, people also speak the language of their masters,
benefactors, advertisers. Thus they do not only express themselves,
their own knowledge, feelings, and aspirations, but also something other than
themselves....
"We
must use the same terms for describing our automobiles,
foods and furniture, colleagues and competitors -- and we understand each other
perfectly. This must necessarily be so, for language is nothing
private and personal, or rather the private and personal is mediated by the
available linguistic material, which is societal material. But this situation
disqualifies ordinary language from fulfilling the validating function which it
performs in analytic philosophy. 'What people mean when
they say...' is related to what they don't say. Or,
what they mean
cannot be taken at face value -- not because they lie, but because the universe of
thought and practice in which they live is a universe of manipulated
contradictions." [Ibid.]
[Speaker's meaning has been highlighted in red; word meaning in purple.
What
a person intends to achieve with his/her words is surely distinct from what
those words mean. So, if NN says (sarcastically) "Well done!" and means
(speaker's meaning) "You screwed up!"
that in no way alters what either "well" or "done" mean in English.]
Despite this, it is a bit rich of Marcuse
pointing his class-compromised finger at ordinary language when the traditional approach to Philosophy
which he champions is itself a clear expression of
ruling-class forms-of-thought. Hence, concerning what Marcuse has to say:
"[H]is (i.e., Marcuse's) words are an expression
of the individual who speaks it, and of those who make him speak as he does, and
of whatever tension or contradiction may interrelate them. In speaking his own
language, Marcuse also speaks the language of his masters, benefactors,
advertisers..." [Edited misquotation of Marcuse (1968).]
it is worth asking: What makes Marcuse think he can rise above
such social forces?6a01
At this point, it is important to
recall once again Marx's
comment:
"The philosophers have only
to dissolve their language into the ordinary language, from which it is
abstracted, in order to recognise it, as the distorted language of the actual
world, and to realise that neither thoughts nor language in themselves form a
realm of their own, that they are only manifestations of actual life." [Marx
and Engels (1970), p.118. Bold emphases added.]
It seems, then, that Marx didn't share
Marcuse's contempt for ordinary language. And no wonder; by the time he wrote
The German Ideology, he knew philosophy couldn't change the world. The
working class is essential to that end. Holding the vernacular in contempt is not, therefore, a good
place to start.
"Where it seems to go beyond
this discourse, as in its logical purifications, only the skeleton remains of
the same universe -- a ghost much more ghostly than those which the analysis
combats. If philosophy is more than an occupation, it shows the grounds which
made discourse a mutilated and deceptive universe. To leave this task to a
colleague in the Sociology or Psychology Department is to make the established
division of academic labour into a methodological principle. Nor can the task be
brushed aside with the modest insistence that linguistic analysis has only the
humble purpose of clarifying 'muddled' thinking and speaking. If such
clarification goes beyond a mere enumeration and classification of possible
meanings in possible contexts, leaving the choice wide open to anyone according
to circumstances, then it is anything but a humble task. Such clarification
would involve analyzing ordinary language in really controversial areas,
recognizing muddled thinking where it seems to be the least muddled,
uncovering the falsehood in so much normal and clear usage. Then linguistic
analysis would attain the level on which the specific societal processes which
shape and limit the universe of discourse become visible and understandable."
[Marcuse (1968),
p.157.]
Marcuse again mistakes the point of
Wittgenstein's method; it isn't just aimed at clearing up "muddles", but
at showing
that the distorted language of Traditional Philosophy can't deliver any
results at all -- that is, other than profound confusion. The last 2500 years of
wasted effort is testimony enough. In which case, we have no choice but to
turn things over to scientists and sociologists -- or, better still, to historical
materialists who haven't sold their radical souls for a
mess of
boss-class pottage.
"The range and extent of the
social system of meaning varies considerably in different historical periods and
in accordance with the attained level of culture, but its boundaries are clearly
enough defined if the communication refers to more than the non-controversial
implements and relations of daily life. Today, the social systems of meaning
unite different nation states and linguistic areas, and these large systems of
meaning tend to coincide with the orbit of the more or less advanced capitalist
societies on the one hand, and that of the advancing communist societies on the
other. While the determining function of the social system of meaning asserts
itself most rigidly in the controversial, political universe of discourse, it
also operates, in a much more covert, unconscious, emotional manner, in the
ordinary universe of discourse. A genuinely philosophic analysis of meaning has
to take all these dimensions of meaning into account because the linguistic
expressions partake of all of them. Consequently, linguistic analysis in
philosophy has an extra-linguistic commitment. If it decides on a distinction
between legitimate and non-legitimate usage, between authentic and illusory
meaning, sense and non-sense, it invokes a political, aesthetic, or moral
judgment." [Ibid.,
p.159.]
Indeed, and the political viewpoint (I
will say nothing of its aesthetic or moral aspects) of such an analysis is the
same as Marx's when he (and not just the OLP-ers) alleged that Philosophy is
based on linguistic distortion and confusion. Although, of course, this wasn't Wittgenstein's aim; it most
certainly is the present author's. But, this just shows that linguistic analysis
doesn't have to be as Marcuse depicted it. And sure, language changes,
but logical grammar does not. [I have given four examples of the latter,
here,
here,
here, and
here,
as well a many others right throughout this site. Readers who follow those link
will also see why logical grammar doesn't change (in the sense that it is ever
undone), it is simply augmented in line with social development.]
"It may be objected that such
an 'external' analysis (in quotation marks because it is actually not
external but rather the internal development of meaning) is particularly out of
place where the intent is to capture the meaning of terms by analyzing their
function and usage in ordinary discourse. But my contention is that this is
precisely what linguistic analysis in contemporary philosophy does not
do. And it does not do so inasmuch as it transfers ordinary discourse into a
special academic universe which is purified and synthetic even where (and just
where) it is filled with ordinary language. In this analytic treatment of
ordinary language, the latter is really sterilized and anesthetized.
Multi-dimensional language is made into one-dimensional language, in which
different and conflicting meanings no longer interpenetrate but are kept apart;
the explosive historical dimension of meaning is silenced." [Ibid.,
pp.159-60.]
Marcuse, of course, says all this from the
standpoint of his own "academic universe", which was long ago
compromised by
thought-forms dominated by ruling-class ideology. But, he is wrong about
linguistic analysis, and particularly in relation to how analysis is carried out in the
Wittgensteinian tradition. Not only has the latter motivated a wider application
of his method to the Arts (Poetry, Drama and Literature -- on that, see Perloff
(1996)), but also to Sociology, Politics and History (particularly the History
and Sociology of Science). Admittedly, much of this has unfolded since Marcuse
rushed into print, but that just shows how peremptory and parochial his accusations were.
[More on this in Essays Twelve Part Seven and Thirteen Part Two, when they are
published.]
"Wittgenstein's endless
language game with building stones, or the conversing Joe Doe and Dick Roe may
again serve as examples. In spite of the simple clarity of the example, the
speakers and their situation remain unidentified. They are x and y, no matter
how chummily they talk. But in the real universe of discourse, x and y are
'ghosts.' They don't exist; they are the product of the analytic philosopher. To
be sure, the talk of x and y is perfectly understandable, and the linguistic
analyst appeals righteously to the normal understanding of ordinary people. But
in reality, we understand each other only through whole areas of
misunderstanding and contradiction. The real universe of ordinary language is
that of the struggle for existence. It is indeed an ambiguous, vague, obscure
universe, and is certainly in need of clarification. Moreover, such
clarification may well fulfil a therapeutic function, and if philosophy would
become therapeutic, it would really come into its own." [Ibid.,
p.160.]
Marcuse has in fact made a valid a point
here, as I have argued
elsewhere. Having said that, this defect is easily rectified. As
noted earlier, unless we understand simple talk, we stand no chance with more
complex exchanges. What is more, we would be fools to look to those whose ideas don't work anyway, or which
imply that change is impossible.
And we would be even more foolish to look to the obscure language Hegel uses to
tell us anything about language. It is certainly true that, for example, the
parable of the parable of the builders in the Philosophical Investigations
(on this, see Rhees (1970b)) looks like it pictures cardboard cut-outs of
human beings, and much of Wittgenstein's work is a-historical. But
that is because he was concerned to investigate logical grammar, which, as I
noted above, is a permanent feature of language. Having said that, a
return to Hegel in order to provide any insight into actual human beings,
how they talk and think, would be like appealing to astrology to inform
astronomy.
"Philosophy
approaches this goal to the degree to which it frees thought from its
enslavement by the established universe of discourse and behaviour, elucidates
the negativity of the Establishment (its positive aspects are abundantly
publicized anyway) and projects its alternatives. To be sure, philosophy
contradicts and projects in thought only. It is ideology, and this ideological
character is the very fate of philosophy which no scientism and positivism can
overcome. Still, its ideological effort may be truly therapeutic -- to show
reality as that which it really is, and to show that which this reality prevents
from being.
"In the
totalitarian era, the therapeutic task of philosophy would be a political task,
since the established universe of ordinary language tends to coagulate into a
totally manipulated and indoctrinated universe. Then politics would appear in
philosophy, not as a special discipline or object of analysis, nor as a special
political philosophy, but as the intent of its concepts to comprehend the
unmutilated reality. If linguistic analysis does not contribute to such
understanding; if, instead, it contributes to enclosing thought in the circle of
the mutilated universe of ordinary discourse, it is at best entirely
inconsequential. And, at worst, it is an escape into the non-controversial, the
unreal, into that which is only academically controversial."
[Ibid.,
p.160.]
This is a
rather glib sales pitch, but like
most advertising it should be taken with a pinch of salt, for, as we have seen,
the only thing that Philosophy succeeds in generating is distorted language,
compounded by confusion, resulting in
non-sensical
and incoherent thought. But, yes, we should be wary of "indoctrination" and
guard against being "manipulated", but, if so, the accusatory finger should
rather be rotated through a full half circle, and pointed at those who look to
boss-class hacks like Hegel for guidance.
Marcuse continues in the same vein in Chapter
Eight of his book, except his target is Analytic Philosophy in general, and not so much
Wittgenstein. In that chapter, he focuses on how Analytic Philosophers have
handled "Universals". I do not propose to deal with this chapter in any great
detail since I have said more-or-less all I want to say on this in Essay Three Parts
One and
Two. However, a few paragraphs
require comment:
"The commitment of analytic
philosophy to the mutilated reality of thought and speech shows forth strikingly
in its treatment of universals. The problem was mentioned before, as part
of the inherent historical and at the same time transcendent, general character
of philosophic concepts. It now requires a more detailed discussion. Far from
being only an abstract question of epistemology, or a pseudo-concrete question
of language and its use, the question of the status of universals is at the very
centre of philosophic thought. For the treatment of universals reveals the
position of a philosophy in the intellectual culture -- its historical
function." [Ibid.,
p.161.]
We saw in the aforementioned Essays, that far
from dealing with genuine universals (i.e., general terms/concepts),
Traditional Philosophers and DM-theorists in fact turn them into the
Proper Names of
Abstract
Particulars, destroying their capacity to express generality. How does
Marcuse manage to side-step this bear trap? The answer is, he
doesn't. Because of his adoption of traditional forms-of-thought,
he blunders right into it:
"Contemporary analytic
philosophy is out to exorcize such 'myths' or metaphysical 'ghosts' as Mind,
Consciousness, Will, Soul, Self, by dissolving the intent of these concepts into
statements on particular identifiable operations, performances, powers,
dispositions, propensities, skills, etc. The result shows, in a strange way, the
impotence of the destruction -- the ghost continues to haunt. While every
interpretation or translation may describe adequately a particular mental
process, an act of imagining what I mean when I say 'I,' or what the priest
means when he says that Mary is a 'good girl,' not a single one of these
reformulations, nor their sum-total, seems to capture or even circumscribe the
full meaning of such terms as Mind, Will, Self, Good. These universals continue
to persist in common as well as 'poetic' usage, and either usage distinguishes
them from the various modes of behaviour or disposition that, according to the
analytic philosopher, fulfil their meaning." [Ibid.
p.161. Bold emphasis added.]
For Marcuse, these universals aren't general
terms, they are Proper Names of the aforementioned Abstract Particulars,
which vacates them of whatever generality they might once have had. [The
disastrous consequences that has on our ability
to say anything at all were discussed at length in Essay Three
Part One.]
We will leave to one side whether or not
these 'terms' persist in common usage since Marcuse offers no evidence that they
do. To be sure, ordinary speakers use words like "mind", "self" and "will" all
the time, but it is open to considerable doubt that when they do so they are
referring, or even alluding, to the artificial 'Universals' of
philosophical lore. Certainly, Marcuse
offers no evidence (or argument) that they do. And it is far from clear that
anyone else has managed to do so, either -- least of all those who lionise Marcuse's work.
In fact, we can draw the opposite
conclusion: since ordinary speakers say general things about
whatever it is that they wish to say such things about (for example, "The boss
is a crook", "Tony Blair is a war criminal", "The Nile is
longer than any other river on earth", "Fighting austerity is
socialist priority", or even "Anyone with half a mind should enter politics, since that's all
you need"), they plainly do not refer, or even allude, to the Abstract
Particulars Marcuse focussed upon. If they were to do that, they wouldn't be able
to make such
general points.
It could be objected that Marcuse does offer his
readers an argument in support of the claims he makes about ordinary speech,
namely this:
"However, this dissolution itself must be questioned -- not only on behalf of
the philosopher, but on behalf of the ordinary people in whose life and
discourse such dissolution takes place. It is not their own doing and their own
saying: it happens to them and it violates them as they are compelled, by the
'circumstances,' to identify their mind with the mental processes, their self
with the roles and functions which they have to perform in their society. If
philosophy does not comprehend these processes of translation and identification
as societal processes -- i.e., as a mutilation of the mind (and the body)
inflicted upon the individuals by their society -- philosophy struggles only
with the ghost of the substance which it wishes to de-mystify. The mystifying
character adheres, not to the concepts of 'mind,' 'self,' 'consciousness,' etc.
but rather to their behavioural translation. The translation is deceptive
precisely because it translates the concept faithfully into modes of actual
behaviour, propensities, and dispositions and, in so doing, it takes the
mutilated and organized appearances (themselves real enough!) for the reality."
[Ibid.,
p.162. Several typos corrected.]
Here, Marcuse undoubtedly advances a substantive point about
what he thinks ordinary folk mean by their use of "mind", etc. However, instead
of looking at how we/they actually employ this word, he imposed an a priori
interpretation and structure on it. [For an illuminating lecture on how such
words are used, readers are encouraged to watch
this video of Peter Hacker speaking on the topic.]
This is doubly unfortunate since one of
the books that Marcuse lists in his 'rogues gallery' is Gilbert Ryle's The Concept of
Mind [i.e.,
Ryle (1949a)
-- this links to a PDF], which shows a far greater sensitivity to our
complex use of psychological vocabulary than anything Marcuse and other
dialecticians have yet managed to cobble together -- indeed, as we will see as this Essay unfolds.
I won't bang on, but I will end with just two more examples (one taken from
Chapter Eight, and one from earlier in the book):
"The 'whole'
that here comes to view must be cleared from all misunderstanding in terms of an
independent entity, of a 'Gestalt,' and the like. The concept somehow expresses
the difference and tension between potentiality and actuality -- identity in
this difference. It appears in the relation between the qualities (white, hard;
but also beautiful, free, just) and the corresponding concepts (whiteness,
hardness, beauty, freedom, justice). The abstract character of the latter seems
to designate the more concrete qualities as part-realizations, aspects,
manifestations of a more universal and more 'excellent' quality, which is
experienced in the concrete.
"And by virtue
of this relation, the concrete quality seems to represent a negation as well as
realization of the universal. Snow is white but not 'whiteness'; a girl may be
beautiful, even a beauty, but not 'beauty'; a country may be free (in
comparison with others) because its people have certain liberties, but it is not
the very embodiment of freedom. Moreover, the concepts are meaningful only in
experienced contrast with their opposites: white with not white, beautiful with
not beautiful. Negative statements can sometimes be translated into positive
ones: 'black' or 'grey' for 'not white,' 'ugly' for not 'beautiful.'
"These
formulations do not alter the relation between the abstract concept and its
concrete realizations: the universal concept denotes that which the particular
entity is, and is not. The translation can eliminate the hidden negation
by reformulating the meaning in a non-contradictory proposition, but the
untranslated statement suggests a real want. There is more in the abstract noun
(beauty, freedom) than in the qualities ('beautiful,' 'free') attributed to the
particular person, thing or condition. The substantive universal intends
qualities which surpass all particular experience, but persist in the mind, not
as a figment of imagination nor as more logical possibilities but as the 'stuff'
of which our world consists. No snow is pure white, nor is any cruel beast or
man all the cruelty man knows -- knows as an almost inexhaustible force in
history and imagination." [Ibid.,
pp.168-69.]
This is in fact a faint echo of Hegel's reference to
what I have called "Spinoza's Greedy Principle" [SGP] (in Essay Eleven
Part Two) --
i.e., "Every
determination is also a negation". However, this is an unreliable principle
(even where any sense can be made of it!), not least because it confuses what we
do with words with the means by which we do it. Of course, that move is about as
brainless as confusing, say, a holiday with the aeroplane you might board in
order travel there, or even a map with a trek in the hills! [The serious
weaknesses of the SGP have been exposed in Essay Three
Part One. They will be
more completely revealed in Essay Twelve Part Five, but they are
connected with the points I have made
here.]
Ignoring for the moment the fact that Marcuse confuses
concepts with words, it isn't even true that:
"the concepts are meaningful
only in experienced contrast with their opposites: white with not white,
beautiful with not beautiful. Negative statements can sometimes be translated
into positive ones: 'black' or 'grey' for 'not white,' 'ugly' for 'not
beautiful.'" [Ibid.]
Colour concepts are meaningful, among other
things, because of the
colour octahedron [this links to a PDF] not because we have met in
experience "not-white" (or whatever). If someone has no understanding of colour
words, they can swim in "not-white" all day long for all the good it will do
them. Much the same can be said about the mastery of other words.6a
But, the above errors are connected with
a much
deeper logical issue. This brings us to the final passage from One Dimensional
Man that I propose to examine:
"In the
classical logic, the judgment which constituted the original core of dialectical
thought was formalised in the propositional form, 'S is p.' But this form
conceals rather than reveals the basic dialectical proposition, which states the
negative character of the empirical reality. Judged in the light of their
essence and idea, men and things exist as other than they are; consequently
thought contradicts that which is (given), opposes its truth to that of the
given reality. The truth envisaged by thought is the Idea. As such it is, in
terms of the given reality, 'mere' Idea, 'mere' essence -- potentiality.
"But the
essential potentiality is not like the many possibilities which are contained in
the given universe of discourse and action; the essential potentiality is of a
very different order. Its realisation involves subversion of the established
order, for thinking in accordance with truth is the commitment to exist in
accordance with truth. (In Plato, the extreme concepts which illustrate this
subversion are: death as the beginning of the philosopher's life, and the
violent liberation from the Cave.) Thus, the subversive character of truth
inflicts upon thought an imperative quality. Logic centres on judgments which
are, as demonstrative propositions, imperatives, -- the predicative 'is' implies
an 'ought'.
"This
contradictory, two-dimensional style of thought is the inner form not only of
dialectical logic but of all philosophy which comes to grips with reality. The
propositions which define reality affirm as true something that is not
(immediately) the case; thus they contradict that which is the case, and they
deny its truth. The affirmative judgment contains a negation which disappears in
the propositional form (S is p). For example, 'virtue is knowledge'; 'justice is
that state in which everyone performs the function for which his nature is best
suited'; 'the perfectly real is the perfectly knowable'...; 'man is free'; 'the
State is the reality of Reason.'
"If these
propositions are to be true, then the copula 'is' states an 'ought,' a
desideratum. It judges conditions in which virtue is not knowledge, in which men
do not perform the function for which their nature best suits them, in which
they are not free, etc. Or, the categorical S-p form states that (S) is not (S);
(S) is defined as other-than-itself. Verification of the proposition involves a
process in fact as well as in thought: (S) must become that which it is.
The categorical statement thus turns into a categorical imperative; it
does not state a fact but the necessity to bring about a fact. For
example, it could be read as follows: man is not (in fact) free, endowed
with inalienable rights, etc., but he ought to be, because he is free in
the eyes of God, by nature, etc." [Ibid.,
pp.110-11. Link added. I have used the on-line text here, and
have corrected any typographical errors I managed to spot.]
We have
already seen that
dialecticians en masse have bought into a
defective theory of
predication,
so it is no surprise to see Marcuse following suite. His claim that the traditional
logic of subject (S) and predicate (p) "conceals
rather than reveals the basic dialectical proposition, which states the negative
character of the empirical reality" may or may not be true --, but if it isn't,
then that is all to the good since "reality" has neither a "negative" nor a
positive "character". In fact, it is only because Marcuse has limited
himself to a very
narrow range of examples that his assertions might seem (to some) to be
plausible. As was noted in Essay Three
Part One:
For example, how would the following be classified?
H1: Every sailor loves a girl who reminds him of
anyone other than his mother.
H2: Anyone who knows Marx's work will also know that
he is second to none in his analysis of all the economic forces operating in
Capitalism, and most of those constitutive of other Modes of Production.
H3: Any prime factor of an even number between
two and one hundred is less than a composite number not equal to but greater
than fifty.
H4: Some who admire most of those who do not
despise themselves often avoid sitting opposite any who criticise those who
claim membership of the minority break-away faction of the Socrates
Appreciation Society.
H5: Today, Blair met some of those who think his
policy in Iraq is a betrayal of his few remaining socialist principles.
Are these universal,
particular, negative, or positive? Are they judgements or propositions? But
these are the sort of propositions that feature in mathematics and the sciences
all the time (to say nothing of everyday speech -- excepting perhaps H4). Indeed,
the serious limitations of the restrictive old logic, with its incapacity to
handle complex sentences in mathematics, inspired Frege to recast the entire
discipline in its modern form over a hundred and twenty years ago. [On this, see
Essay Four.]
Some might object that these aren't the sort
of "judgements" with which 'traditional logic', or even Hegel, concerned itself/himself, but that is
precisely the
point. It is only because Marcuse, along with other dialecticians, has relied on a
bowdlerised form of the (already antiquated) Aristotelian Logic extant in
Hegel's day that his argument even seems to gain a
slender toe-hold. [On this, see Kenny (2006b), pp11-13.]
However, let us assume
for the moment that Marcuse's
analysis is impeccable. Even then, what he alleges would still
be incorrect:
"In the classical logic, the
judgment which constituted the original core of dialectical thought was
formalised in the propositional form, 'S is p.' But this form conceals rather
than reveals the basic dialectical proposition, which states the negative
character of the empirical reality. Judged in the light of their essence and
idea, men and things exist as other than they are; consequently thought
contradicts that which is (given), opposes its truth to that of the given
reality. The truth envisaged by thought is the Idea. As such it is, in terms of
the given reality, 'mere' Idea, 'mere' essence -- potentiality.
"...Or, the categorical S-p
form states that (S) is not (S);
(S) is defined as other-than-itself. Verification of the proposition involves a
process in fact as well as in thought: (S) must become that which it is.
The categorical statement thus turns into a categorical imperative; it
does not state a fact but the necessity to bring about a fact. For
example, it could be read as follows: man is not (in fact) free, endowed
with inalienable rights, etc., but he ought to be, because be is
free in the eyes of God, by nature, etc." [Ibid.]
Clearly, this argument depends on "men and things" each having an essence, which
Marcuse simply takes for granted. Of course, to mystics like Hegel and
Aristotle, it seemed clear that "men and things" did indeed have an
"essence" (which was for them something that had been decided upon by 'god'), but this is just another example of
ruling-class ideology
dominating their thought.
But, even if
that allegation were itself incorrect, what
is Marcuse going to say about propositions like the following?
M1: Human beings are mortal.
M2: Tables and chairs are
often made of wood.
Do
these "oppose" the "truth of reality"? Are we to assume that humans are 'really'
immortal, and that they oughtn't be like this -- i.e., mortal? Or,
that ordinary objects are in 'reality' non-material, and that there is
an 'imperative' here which means that we should all struggle to make them material?
If not, Marcuse's analysis can't be relied upon to reveal 'the truth' even about
mundane matters of fact --
which shouldn't surprise us in view of the preceding paragraphs -- that
is, considering the
defective logic Marcuse
appropriated and then put to misuse in order to arrive at most of his
rather odd conclusions.
Finally, if what Marcuse asserts were correct:
"consequently thought
contradicts that which is (given), opposes its truth to that of the given
reality",
then
this 'thought' would itself 'contradict' reality.
And, if that were so, it would mean that in reality there are no such 'contradictions'
-- otherwise Marcuse's own comments wouldn't represent things as they 'really
are'. As we have seen
many times already,
Diabolical Logic like this soon self-destructs.
It is time we left this prime example of
boss-class confusion and turned instead to consider several others who have
drifted off into
deep waters.
A more recent swipe at Wittgenstein comes
from my old friend Ben Watson (in a book that is openly contemptuous of academic
standards -- a dismissive approach that Marx himself would have criticised, to
say nothing of Lenin and Trotsky):
"Take Ludwig Wittgenstein.
Deprived of the benefit of Trotsky's optical materialism, his commitment to
Aristotelian formal logic drives him into madness…. Wittgenstein's 'play of the
imagination' is incipient schizophrenia, the confusion of reality with symbolic
systems used to represent it…. The 'logical' analytic philosophers, whose
attempt to live in the flatland of symbolic representation, drove themselves
crazy." [Watson (1998), p.121.]
To be fair, the first set of dots in
the above passage conceals the omission of a long quotation from Wittgenstein's
Philosophical Investigations -- which Watson then, alas predictably,
proceeds to misrepresent, as we will see.
[FL = Formal Logic; AFL =
Aristotelian Formal Logic; LOI = Law Of Identity; LOC = Law Of Contradiction;
LEM = Law Of Excluded Middle; HM= Historical Materialism.]
The reference to Trotsky's "optical
materialism" is no less unfortunate. As I demonstrated in
Essay Six: If we are more accurate
and honest,
Trotsky's "optical materialism" rather more closely resembles 'Dialectical
Myopia'.
Even so, the presence of these relatively
minor flaws shouldn't detract from the book's more egregious
errors.
First, as far as logic is concerned,
Wittgenstein was a Fregean (even if he adopted a critical but deferential
stance toward his work).7In fact, Watson is invited to try to find a single reference in
Wittgenstein's entire corpus (of over five million words) that commits him to
AFL, or any aspect of it.8
Indeed, Wittgenstein's first published work, as short review of Coffey's two
volume logic text, he had this to say:
"In no branch of learning can an author disregard the results of honest research
with so much impunity as he can in Philosophy and Logic. To this circumstance we
owe the publication of such a book as Mr Coffey's
Science of Logic: and only as a
typical example of the work of many logicians of to-day does this book deserve
consideration. The author's Logic is that of the scholastic philosophers, and he
makes all their mistakes -- of course with the usual references to Aristotle.
(Aristotle, whose name is taken so much in vain by our logicians, would turn in
his grave if he knew that so many Logicians know no more about Logic to-day than
he did 2,000 years ago). The author has not taken the slightest notice of the
great work of the modern mathematical logicians -- work which has brought about
an advance in Logic comparable only to that which made Astronomy out of
Astrology, and Chemistry out of Alchemy....
"[Summarising
Coffey's errors -- RL]:
"[1] The author
believes that all propositions are of the subject predicate form....
"[3] He confounds the
copula 'is' with the word 'is' expressing identity....
"The worst of such books is that they
prejudice sensible people against the study of Logic...."
[Wittgenstein (1913a),
pp.2-3.]
So,
while the above shows he had great respect for Aristotle, he had much greater
respect for Frege and Russell's mathematical logic, indeed, as he said in his
second published work:
"I do not wish to judge how
far my efforts coincide with those of other philosophers. Indeed, what I have
written here makes no claim to novelty in detail, and the reason why I give no
sources is that it is a matter of indifference to me whether the thoughts that I
have had have been anticipated by someone else.
"I will only mention that I am
indebted to Frege's great works and of the writings of my friend Mr Bertrand
Russell for much of the stimulation of my thoughts." [Wittgenstein
(1972), p3.]
Here, he described Frege's work as "great",
a term he applied to no
other Philosopher or Logician.
Second, far from confusing symbols with reality,
Wittgenstein was in fact one of the few leading Philosophers in the entire
history of the subject consistently to strive to do the opposite,
arguing that most of what passes for Traditional Philosophy was guilty of this very failing
(a point that has been reiterated throughout this site). Hence, it is a little rich of
Watson to
raise this particular point when he himself is an avid fan of dialectics, whose
theorists constantly do precisely what he accuses Wittgenstein of doing
(substantiated here and throughout this
site, but especially in Essays Three Parts One
and Two,
and Twelve
Part One).
Finally, and with respect to the passage
quoted above (from
the Philosophical Investigations, aimed at discussing the LOI), Watson has
clearly missed the point. Onthevery same page,
Wittgenstein himself admits the following about his earlier approach:
"My symbolic expression [in
the Tractatus] was really a mythological description of the use of a
rule." [Wittgenstein (1958), p.85e, §221.
(This links to a PDF.)]
Now, even the most superficial reading of the
Philosophical Investigations will reveal that Wittgenstein is arguing
against the mythology surrounding our use of symbols -- including his own
earlier misdemeanours in this respect --, that is, against what I have called
the
"fetishisation of language".9
This is one reason why
Wittgenstein himself took the LOI to task in both his earlier and later
periods. On this issue, he argued that those who regard this 'law' as a particularly deep sort of truth
misconstrue a rule for the use of certain symbols/words as if it were a scientific or
metaphysical truth about reality. Indeed, and as we have seen, this is precisely
how Trotsky, for example, misinterpreted the LOI, even if, following Hegel, he declared
this 'law'
always false -- or, perhaps, both false and true, or, at least, not
always unconditionally true. It is this
tradition that Wittgenstein sought to undermine: a pattern of thought that
Hegel, Trotsky and other DM-theorists share with card-carrying defenders of a
ruling-class view of the world.
One of the main aims of
Wittgenstein's method was to show that philosophical theses (like those that are based on
a traditional reading of the LOI, the LOC, and the LEM) were predicated on a
systematic misconstrual of rules as if they were substantive truths
about the world. To be sure, Wittgenstein might not have concurred with the following
observation, but it is worth making all the same: such rules become fetishised when alienated
forms-of-thought encourage theorists to mistake contingent features of
language for necessary relations, objects or processes in reality (i.e., with
all those underlying, mysterious "essences").
Hence, what had once been the product of
the social relations between human beings (i.e., language) became inverted and
then systematically misconstrued as the real relations between things -- and, in
the case of the LOI, the real relation between "a thing and itself" -- or,
between two or more events/states of affairs.
Of course, this
misidentification had always been taken seriously by Traditional Thinkers, and
not just in relation to the LOI.
Because of this it
was easy for them to project this error back onto nature to give spurious
'objectivity' to their theories about 'Ultimate Reality'. In the Ideal World they had conjured into existence, the
socially-sanctioned relationships between words were mistaken for the real
relationships between things, or even those things themselves. The material world
was now interpreted through
this distorted,
idealised
view of language, and in such a way that contingent features of discourse were
regarded as objective features of 'Reality'. In this way, these distorted linguistic
forms came to determine the fundamental nature of 'Reality', which was in fact
just a projection of fetishised discourse back onto it! [We saw this was also the case with Marcuse,
above,
and we will see it several more times throughout the rest of this Essay.]
Even though
dialecticians have tried to distance themselves from Idealist moves like these,
by means of their invention of scientific-sounding
'philosophical reasoning' -- through which they attempt to argue that the LOI, the
LEM and the LOC are
empirically false while being, in some sense, 'ideally'/'abstractly' true
-- these
moves only succeed in reduplicating the 'problems' with which they began, as we saw in Essays
Four,
Five,
Six
and Eight Part
Three.
If, for example, the usual
interpretation of the LOI (as a 'necessary'/metaphysical truth) is in fact the
result of a confusion over the use of certain symbols, then the standard
DM-criticism of that 'Law' will only ever be self-defeating. That is because
that critique is directed against a mythological representation of a rule of language,
not against an empirical falsehood (or even an 'ideal'/'abstract' truth). Indeed, as
we saw in Essay Six, such a ham-fisted 'attack' on the LOI cannot succeed because it is
aimed at a mirage; hence an 'attack' like this will always backfire on those
held in
thrall to this mythological picture. That is because this wrong-headed approach undermines the meaning
of the words used to that end -- e.g., "same", "equal",
"exact", "identical", and "different". The result is that anyone foolish enough
to stray down that path will only ever
succeed in vitiating their own use of these very words.
And this is precisely what we witnessed earlier in
connection with Trotsky's (and derivatively, Hegel's)
'analysis' of the LOI;
any attempt to undermine that 'law' cannot fail to undermine the application of
the terms used in that very 'attack'.
Hence, a misplaced assault on the LOI is
forced to employ symbols whose own identities (and hence denotations) are
simultaneously called into question and not called into
question. That is, this approach aims to show how limited this 'law' is, but its
execution requires this 'law' to be valid so that the symbols it uses don't
change while the argument is being prosecuted. In which case, the argument is entirely misconceived, since, if, per impossible,
it were valid all the words an erstwhile critic must use would (as phenomenal
objects) cease to be identical from moment to moment, which would mean that the following
hackneyed 'dialectical' criticisms of the 'equality' of two letter "A"s would fall flat:
"[T]he first of [the
universal Laws of Thought], the maxim of Identity, reads: Everything is
identical with itself, A = A…." [Hegel (1975), p.167.]
"In this remark, I will
consider in more detail identity as the law of identity which is usually adduced
as the first law of thought.
"This proposition in its positive expression
A = A is, in the first instance, nothing more than the expression of an empty
tautology." [Hegel (1999), p.413.]
"Abstract Identity (a = a…) is
likewise inapplicable in organic nature. The plant, the animal, every cell is at
every moment of its life identical with itself and yet becoming distinct from
itself….The law of identity in the old metaphysical sense is the
fundamental law of the old outlook: a = a." [Engels (1954), pp.214-15.]
"The 'fundamental laws of
thinking' are considered to be three in number: 1) The Law of Identity… [which]
states that 'A is A' or A = A…." [Plekhanov (1908), p.89.]
"…Hegel elucidates the
one-sidedness, the incorrectness of the 'law of identity' (A = A)…." [Lenin
(1961), p.134.]
"The Aristotelian logic of
the simple syllogism starts from the proposition that 'A' is equal to 'A'. This
postulate is accepted as an axiom for a multitude of practical human actions and
elementary generalisations. But in reality 'A' is not equal to 'A'." [Trotsky
(1971, p.63.]
"Formal Logic starts from the
proposition that A is always equal to A. We know that this law of identity
contains some measure of truth…. Now…when we go to reality and look for evidence
of the truth of the proposition: A equals A…we find that the opposite of this
axiom is far closer to the truth." [Novack (1971), pp.32-33.]
"Formal Logic asserts:
'A is A'.
Dialectical Logic is not saying 'A is not-A'…. It says: A is indeed A, but A is
also not-A precisely so far as the proposition 'A is A' is not a tautology but
has real content." [Lefebvre (1968), p.41.]
"The Law of identity is
usually expressed in the form, A is A. That is, each thing is identical with
itself." [Somerville (1946), p.183.]
"The Aristotelian conception
of the laws basic to correct thinking may be stated as follows: 1. Law of
Identity: Each existence is identical with itself. A is A…." [Somerville (1967),
pp.44-45.]
"Classical, Aristotelian
logic takes as its fundamental premise the Law of Identity, the statement that a
thing is identical with itself. Expressed in a formula: A is A…. In Aristotle's
formal logic A is A, and never non-A. In Hegel's dialectics A is A as well as
non-A." [Baghavan (1987), pp.75-76.]
"The biggest contradiction of
all lies in the fundamental premises of formal logic itself…. The basic
laws…are:
1) The law of Identity ('A' =
'A')…." [Woods
and Grant (1995), pp.90-91.]
"Dialectics, or the logic of motion, is distinct from formal or static logic.
Formal logic is based on three fundamental laws:
"(a) The
law of identity: A is equal to A; a thing is always equal to itself." [Mandel
(1979), p.160.]
"The laws of logic are based
on two main propositions. The first is that of identity or of self-conformity.
The proposition very simply states: 'A is A,' that is every concept is equal to
itself. A man is a man, a hen is a hen, a potato is a potato. This proposition
forms one basis of logic." [Thalheimer (1936), pp.88.]
"[In
FL] things are defined statically, according to certain fixed properties -–
colour, weight, size, and so on. This is denoted by the expression 'A is equal
to A'." [Rees (1998a), p.272.]
If two such letters are "never equal" to one another, then sentences
stand no chance.
In practice
this means that no one -- not Hegel, not Engels, not Plekhanov, not Lenin, not Trotsky,
not Mao... -- would have access to
identically the same message that they had committed to paper
the previous day, let alone those written by others seventy, eighty,
or one hundred and eighty
years earlier, since, on this account, there would be no such thing.
But, critics who have
arrived at the same conclusion as Trotsky (or Hegel) about the LOI must clearly have
done just something their theory says cannot be done. They must have access to the exact message
they had committed to paper -- which message now tells
them there can be no such thing!
Clearly, this undermines any
conclusions such critics might draw -- but not the LOI. Indeed, that 'law'
(or rather this rule of language) will
have just been used, and must always be used, in this charade aimed at to deriving
this self-defeating result. Hence, their own implicit (or explicit) use of
identity -- in this instance, involving the identity of symbols, meanings and
the use of language over many generations -- to criticise the 'law' under scrutiny counts as
a practical refutation of that
very criticism! With that, their attack on the LOI self-destructs, which is,
of course, also part of the reason why so many 'dialectical theses' so readily collapse into incoherence
-- as, indeed, we saw
in Essay Six.
[The "relative stability" defence has
been defused here.]
As Wittgenstein noted, we can't get outside
language (and we can't even try) in order to state 'philosophical truths',
or those which masquerade as particularly deep
'philosophical truths' about discourse, let alone about
'Reality'.
By implication, this can't be done
either with the more radical aim
of undermining the application of fundamental rules of language (such
as those expressed by the LOI and the LOC). Anyone attempting to do this will find that they
first have to employ these self same rules in order to undermine them, which
will, naturally, fatally damage that attack. And, that is why theorists can't even try to challenge logical features/rules
of language like these.
[This is a summary of three much longer arguments found
here,
here and
here
(where I also respond to one or two obvious objections).]
Nevertheless, in a later part of the same
book, Watson offered his readers the following thoughts:
"The radical democracy of
Voloshinov's linguistics is a model for any theory of the superstructure. It
stems from the fact that he does not abstract speech from its actual use in
society. This is the very opposite of philosophers who build a system by
wondering what it means to stare at their desk. It is a slap in the face for
cretins who think it is clever to read Wittgenstein." [Watson (1998), p.334. Bold
emphasis added.]
However, if these "cretins" have
read their Wittgenstein with the same 'careful' attention to detail that certain
comrades have devoted to the same task (no irony intended) then this epithet is well-deserved.
To hammer the point home, Watson very helpfully provided his readers with an
example of Voloshinov's careful use ofordinary speech in this
further quotation from the latter's book:
"The separation of word
meaning from evaluation inevitably deprives meaning of its place in the living
social process (where meaning is always permeated with value judgement), to its
being ontologized and transformed into ideal Being divorced from the historical
process of Becoming….
"Meaning -- an abstract
self-identical element -- is subsumed under theme and torn apart by theme's
living contradictions so as to return in the shape of a new meaning with a
fixity and self-identity only for the while, just as it had before." [Voloshinov
(1973), pp.105-06, quoted in ibid., pp.334-35.]
Now, I'm sure Watson can clearly recall the last
time he heard ordinary folk talking like this at work, down the pub or even on
a picket line, discussing how the bosses are always "ontologizing" their jobs,
or downsizing them so that the number of operatives is no longer
"self-identical" with whatever it had been a month earlier. In fact, observers
of everyday conversations regularly note how it is nigh on impossible to stop
working people constantly
talking about "ideal Being", "theme" and "Becoming".
Indeed, and on a personal note, I can vividly
recall selling revolutionary papers alongside Ben in XXXX in the late 1980s -- how we
happily shouted catchy slogans about "Being", "Becoming" and "theme".
We definitely sold a record number of papers as a result.
Cheap debating points? Perhaps so. But,
Watson will
need to research his work a little more carefully if he hopes to substantiate
the allegations he levelled against Wittgenstein -- or, indeed, if he wants to
establish his claim that Wittgenstein is at all representative of
twentieth century
Analytic Philosophy.
In fact, Wittgenstein's method was, and still
is, ignored by the vast majority of Analytic Philosophers
(and by practically all professional Philosophers).10 Even when his
approach was slightly more 'in vogue', as it were, only a tiny minority of Analytic
Philosophers fully embraced it. One reason for this is that
in his later work Wittgenstein insisted on using the vernacularwherever possible --
unlike, one might add, Voloshinov and Hegel -- and, dare I say it, Watson. Another
reason is that his method reveals how
non-sensical, confused and
useless Traditional Philosophy
is -- which approach would bring the entire subject to its long overdue end.
In that case,
naturally, the fact that professional Philosophers almost en masse ignore
Wittgenstein's method is no more surprising than the fact that members of the UK
Royal Family aren't prominent Republicans.
An apposite quotation from Larry Laudan (although aimed at French Philosophers) springs to mind, here:
"Foucault
has benefited from that curious Anglo-American view that if a Frenchman talks
nonsense it must rest on a profundity which is too deep for a speaker of English
to comprehend." [Laudan (1977), p.241. I owe this reference to Kitcher (1998),
p.55. Link added.]
If Foucault's name and the phrase "a
Frenchman" are replaced by "Voloshinov" and "a sort-of-Bolshevik", respectively,
then this might help explain what prompted Watson to write a 400-page book
eulogising similar "profundities",
and worse.
Finally, this almost unseemly dismissal of a
fellow comrade's work in fact finds ample justification in the subtitle Watson
gave his book (Art,
Class and Cleavage): viz.: Quantulumcunque Concerning Materialist
Esthetix. A
dog's dinner of a title, for sure -- but a genuine slap in the
face for those who think it clever, or helpful, to confuse revolutionary socialism with the
intellectual equivalent of rabies.
This brings us to Voloshinov. Recently, his work has been
reviewed by a number of
comrades: John Parrington, Marnie Holborow, Sean Doherty, Dave McNally, and Chik
Collins.11
Because what Voloshinov 'appears' to have said about language flatly contradicts
much that is contained in the Essays published at this site -- and in view of several of
the unfavourable things said about his work above -- detailed comments about his
work are clearly
in order.12
Marnie Holborow summarised one of Voloshinov's main
insights in the following way:
"A fundamental element of
Volosinov's critique of abstract objectivism is his view of language being able
to generate new meanings…. This generative quality arises from the fact that
language is inseparable from its context and its users…. The meanings and
different connotations for a word or a piece of language are constructed by the
speakers, who give each utterance their particular evaluative accent.
"Let us take an example…. [:]
I'm hungry conjures up a general concept. When, however, we look at
different contexts in which the phrase might be used, we see how the evaluative
accent changes everything. A child saying this to her mother might be
indirectly a request for the mother to get her something, an enquiry about what
there is to eat, or a statement that she just feels like something to eat.
One adult saying it to another might mean that it's time for lunch and be a
suggestion that they go somewhere to eat…. In each case the context is not
merely the gloss on the meaning but constitutes different meanings --
different inevery aspect…." [Holborow (1999), p.28. Bold emphasis
added. Italic emphasis in the original On Holborow's spelling of Voloshinov's name, see
here.]
Unfortunately, Holborow has chosen what Philosophers of language call an
avowal, which possess rather unique features. So, it is far from clear
that any useful, or reliable, generalisations across every use of
language can be inferred from them. What, we might ask, is the 'evaluative
content' of any of the following, if someone utters them:
U1: New York is bigger than
Athens.
U2:
Voloshinov wrote Marxism and the Philosophy of Language.
U3:
A fundamental element of Volosinov's
critique of abstract objectivism is his view of language being able to generate
new meanings.
Furthermore, when Holborow says that the sentence
"I'm hungry conjures up a general concept", it isn't at all clear
what she means. What precisely is "general" about it? Unless we suppose, perhaps, that several people utter this sentence all at
once as part of a synchronised plea for food, say -- or one person suffering
from a multiple personality disorder comes out with it -- no
generality seems to
be implied here at all. I think she might mean that this sentence can mean many
things. We will return to consider this possibility later.
Perhaps Holborow also meant that this sentence when considered in
isolation from an occasion of its use possesses certain non-specific general features -- or,
that maybe the sentence itself might
suggest them to us. What these are Holborow unfortunately failed to say; nor does
she indicate why they are general, or, indeed, why they are concepts,
as opposed to propositions, requests or orders, for example.13
Moreover, Holborow's claim that utterances
have an "evaluative accent" is puzzling, too, since it is unclear what evaluation
has to do with the type of request she herself considered. Had she
interpreted such a plea (i.e., "I'm hungry") as the equivalent of something
like: "I like food", or "Food is sacred" -- or even "Food is theft"
-- her point
might have been a little clearer. But, what sort of "evaluative accent" does a
plain and simple request for food possess or suggest? Again, Holborow failed to
tell us. [To be sure, Voloshinov does make some attempt to say; his comments will be examined
presently.] Even less obvious is how an "evaluative accent" could affect the meaning
of any of the words used -- as opposed to altering what a speaker might
consequentially, or incidentally, intend to convey by means of them. As
we shall see, these two aren't at all the same.
Of course, Holborow is simply summarising
Voloshinov's view:
"Any word used in actual
speech possesses not only theme and meaning in the referential, or content,
sense of these words, but also value judgement: i.e., all referential contents
produced in living speech are said or written in conjunction with a specific evaluative accent. There is no such thing as a word without evaluative
accent." [Voloshinov (1973), p.103. Italic emphasis in the original.]13a
Voloshinov offered no proof of these rather sweeping statements, and none of the
comrades listed above have even thought to fill in the evidential gaps, in the
meantime. Nevertheless, he then proceeded to connect "evaluative accent" with
"expressive intonation", but he failed to say why the latter are in any way
"evaluative". The same can be said about Holborow's commitment to this idea. To be sure,
Voloshinov quoted a long passage from
Dostoyevsky to clarify his point, adding:
"All six 'speech
performances' by the artisans [in the quoted passage from Dostoyevsky -- RL] are
different, despite the fact that they all consisted of one and the same word.
The conversation was conducted in intonations expressing the value judgements of
the speakers. These value judgements and their corresponding intonations were
wholly determined by the immediate social situation of the talk...." [Ibid.,
p.104.]
But, is this true of every utterance? If it
is, then, as noted above, Voloshinov neglected to include the data establishing
this for a fact.
Despite this, what Voloshinov says is highly
implausible in itself. What, for instance, is the "evaluative accent" of this response
(and many more like it):
Questioner: "What can you tell me about the
River Nile?"
Maybe these do contain,
or imply, an "evaluative accent";
who can say? But, until we are told what an "evaluative accent" is, little more
can be done with such vague claims.
Nevertheless, Holborow also argued that
different contexts of utterance (or, is it the different "evaluative accents" of
each utterance?) constitute entirely new meanings each time, which are "different in every aspect".
Once more, Holborow failed to explain how the
same words could take on these new meanings in this way. To be sure, different
connotations can be promoted by
prosody (i.e., intonation, rhythm, or stress, etc., as Voloshinov
himself noted). For example: "I'm hungry" suggests something different from
"I'm hungry". But, even then, these words still don't change their meaning
-- why would anyone use them if they meant something different? What changes here is what a speaker might hope to convey by the use
of familiar words accentuated differently. Otherwise, speakers might just as
well say "I'm cold" and 'mean' the same as someone else might 'mean' by
their use of
"I'm hungry".
Again, it could be that an injudicious choice of
examples has distorted Holborow's conclusions, since the sentence "I'm hungry" uses an
indexical
expression (viz., "I'm") -- that is, this particular sentence depends for
its incidental import on one of its words being relativised to a time, a
speaker, a place and possibly also an occasion.
On the other hand, from what little else the
above passage says, it is
reasonably clear that Holborow probably doesn't have this aspect of
sentence/word use in mind -- i.e., pointing out the obvious fact that as each
individual utters the words "I'm hungry" the "I" could relate to someone new,
changing the import of what was said by adverting to a different speaker. That is because the mere fact that one person might utter it one minute,
and another the next, doesn't warrant the conclusion Holborow draws that these
words convey
a different meaning "in every aspect", each time.
Perhaps this is
being too hasty? If so, it is
worth considering Holborow's claims more closely. Just like each 20 cent coin,
or each 50 pence coin is a
tokens of the same type
(they are all examples of the same coin -- assuming they aren't counterfeit),
so, each inscription, or each utterance of "I'm hungry" is a token of the same
sentence type. Hence, Holborow might be claiming that different
tokens of the same type utterance,
"I'm hungry", could be used to say different things, and that the
meaning of each of these speech acts is entirely dissimilar, since the
occasions of utterance can't fail to vary, accordingly:
"A
child saying this to her mother might be indirectly a
request for the mother to get her something, an enquiry about what there is to
eat, or a statement that she just feels like something to eat. One adult
saying it to another might mean that it’s time for lunch and be a suggestion
that they go somewhere to eat…." [Ibid., p.28.]
Anyway, what Holborow then goes on to say
doesn't seem at all correct:
"In each case the context is
not merely the gloss on the meaning but constitutes different meanings --
different in every aspect…." [Ibid., p.28. Bold emphasis added.]
Holborow can't seriously be suggesting that
words have new meanings ("different in every aspect") each time they are
uttered. If she were, then these particular words (i.e., "I'm hungry") would be of
no use to anyone, since no one would be able understand what they (the words) meant from
occasion to occasion.
Well, perhaps these 'new meanings' could be inferred from the
intentions of each speaker, or from the context of each utterance? But, in that
case, Holborow's own suggested translations (i.e., that "I'm hungry" means "It's time
for lunch", etc) would surely be subject to the very same
equivocation, in that it, too, would be occasion-sensitive and in need of its
own translation -- just as each of these new translations would, as well, and
so on. Naturally, this would imply that the supposed translation (i.e., "It's time for
lunch") could itself mean "I'm bored with this conversation", or
"I can't see the point of this", or..., which in turn
could mean, "I wonder what's for tea", or "I can't make out what she
is saying", which
themselves could mean…, and so on.
Indeed, when Holborow wrote:
"A
child saying this to her mother might be indirectly a
request for the mother to get her something, an enquiry about what there is to
eat, or a statement that she just feels like something to eat. One adult
saying it to another might mean that it's time for lunch and be a suggestion
that they go somewhere to eat…" [Ibid., p.28.],
maybe she really meant something
different, too? Given her own intentions, she could have meant: "I think Voloshinov is correct
and I want you to agree with me". In that case, "I think Voloshinov is correct
and I want you to agree with me" and:
"A
child saying this to her mother might be indirectly a
request for the mother to get her something, an enquiry about what there is to
eat, or a statement that she just feels like something to eat. One adult
saying it to another might mean that it's time for lunch and be a suggestion
that they go somewhere to eat…" [Ibid., p.28.]
would be synonymous!
If intending something can
change the usual meaning of a word to the extent that totally different passages
and sentences become synonymous, then each and every one of our words/sentences could mean
anything whatsoever. Hence, in this case, not only would "I'm hungry" mean the
same as "It's time for lunch", it could mean the same as "I think
Voloshinov is correct and I want you to agree with me" -- as well as "I'm not the
least bit hungry". Who can say?
Furthermore, the (above) employment here of
these very same words (by me, RL) implies that all three passages must now mean:
"I (RL) disagree, and think Voloshinov is seriously confused", since that is
what I intended to convey by this passage of text. This implies that all three
sentences must mean at least two or more totally different things!
I have now used four sentences to mean "I (RL) am right and Holborow is wrong".
Hence, if Holborow were right,
these must all mean
the same thing (viz.: "I (RL) am right and Holborow is wrong"), even while
all these (now) five sentences could mean something entirely different if
someone else used them to advertise, say, the sale of a garden gnome on
E-Bay!
In that case, by
extrapolation, these words could now be led by the nose to mean anything
whatsoever -- and hence nothing at all. [That is, these words would
have no intrinsic meaning.13b] Of course, since I (RL) intended that all
these passages should end up this way (i.e., that they should mean nothing), that
must mean my use of these sentences indicates that they do indeed mean nothing. So,
when I use Holborow's words, they mean what I want them to mean, not what she
intended -- i.e., nothing at all!
Is anyone convinced by any of this? They should be if they
agree with Holborow -- or, perhaps even with Humpty Dumpty from
Alice Through the Looking Glass.14
It could be argued that the
context of
utterance will succeed in eliminating many of these fanciful 'translations';
unfortunately, as we will see, that isn't even remotely correct.
Once more, therefore, if the words "I'm hungry" do
in fact
mean the same as "It's time for lunch", why do we need the translation?
As competent speakers of English (or whatever language in which this sentence
had been expressed), we would all know what this sentence says. For example, who
(saving small children, those suffering from some sort of brain disease, and
those ignorant of English/some-other-language) needs
to have the word "puppy" translated as "infant dog" each
time it is used? Indeed, only someone with a
rather poor grasp of their native tongue would need to have a perfectly ordinary
sentence translated into another perfectly ordinary sentence (unless, of
course, the former contained a coded message of some sort), for them to be able
to understand it.
Again, if all such sentences required
translation, then why not also those that are offered as their 'real' meaning? If
we need to be told what "I'm hungry" really means, how can we be sure we
understand, say, "It's time for lunch"? Perhaps, as already noted, the replacement/translated
sentence means something else, too? On the other hand, if the replacement
sentence "It's time for lunch" is already understood, and needs
no further sentence to make it clear, why isn't that the case with "I'm
hungry?" Why is the first in need of translation and not the second?
It is worth stressing here that I am not
denying that speakers can often intend to convey a message that it is time for
lunch by saying "I'm hungry"; what is being questioned is whether such an
intention can change what the words in "I'm hungry" actually mean.
Of course, Holborow isn't trying to
translate one sentence into another; she is offering an analysis of the various
uses toward which we put language -- amplified by the observation that discourse is context-dependent. This topic
will be examined in more detail
below.
The fact that several speakers can intend to
produce different effects by the use of typographically identical words/sentences
depends on the words used
having relatively fixed meanings already. If that weren't so, then, as
noted above, with
respect to a particular utterance (i.e., each physical speech act), any
words would do. The child in Holborow's example could say: "My socks are wet" and
'mean' that she wants her mother to get her some food -- or that her
mother should join the Foreign Legion, or that this week's TV Guide
had been delivered, or that her goldfish has died, or, indeed,
anything whatsoever. If context determined the (public) meaning of our words, all of these would be
possible. Why choose these words if any will do? Could it be
because of what these words already mean, and all competent English
speakers know what they mean, and they/we all agree over what they mean?
Plainly, Holborow failed to consider the
most obvious and plausible option here: What if "I'm hungry" is in fact being used to communicate
the same thing each time, or most of the time? What if it is being used
to indicate that the speaker is indeed hungry (period)? In that case, wouldn't
we be tempted to say that a perfectly ordinary act of communication had
occurred? Isn't this how we ordinarily address and comprehend each other?
Why is this so mysterious? Why do we need to be bamboozled into thinking
otherwise? Why do we need to eulogise the work of someone (i.e, Voloshinov) who has made something
that is
patently obvious seem hopelessly obscure?
Even on Holborow's account, we certainly
can't rule out the
possibility that "I'm hungry" might be being used to say the same thing many
times over. On the other hand, if this possibility could be ruled out, and anything could
mean anything (and was dependent on context), no one else would be able to
indicate, for instance, that it was time for lunch by the use of the words "I'm
hungry" -- including Holborow and her own translated suggestion to that effect!
This is because, if
meaning were that sensitive to each occasion of use -- to the extent that the meaning
of what had been said changed in "every aspect", every time --, then no one else could ever utter
"I'm hungry" and mean "It's time for lunch". In fact, no one else could ever
make the point that has just been made by me in the last sentence, using the same words in just
that way in any other sentence! Once used, the meaning of any given set of
words, or sentence, would
have been allused up, so to speak; and any words uttered thereby would have to be
forever sealed away in the
archives never to see the light of day again.
Presumably, therefore, we aren't
being asked to suppose that once these words have been used in this
uncontroversial manner to indicate that the speaker was indeed hungry, no other
speaker would ever be able to use them this way again. Holborow can't possibly mean that. But, if not,
what is the force of her claim that each utterance changes meaning in "every
aspect"? If in this clear, everyday example this doesn't happen, and at
least two utterances of the same token words can (and do) have the same
meaning (i.e., the same as the 'unspoken utterance', "It's time for lunch"), what can Holborow possibly have meant by what she herself
said?
Moreover, the meaning of any translation sentences/words ("It's time for lunch",
etc.) must be fixed forever in Platonic heaven, otherwise there would be nothing
for the original utterance to be translated into, that didn't also need
translating. And yet, if that were so, why can't it be the case with the words
in the original utterance?
However, let us suppose for a moment that Holborow is right,
and each utterance of "I'm hungry" does mean something
(completely) different each time, and that the
context,
aims and intentions of speakers can actually change the meaning of any of the words used.
In such circumstances, as was argued earlier, who would then be able to say what
such words actually meant? Certainly not the person uttering them; any attempt
to explain his/her own meaning (even if this were 'internally voiced')
would
surely be subject to the very same equivocation/translation. The words used to do just
that would also have to change
in meaning upon being uttered -- or, rather, any of the explanatory words
employed to that end would
themselves be sensitive to such changes on each new occasion of use/translation. Still less would hearers of these
words be able to say what they meant; they could now only guess what these elusive meanings might be, or
might have been -- and, incidentally, whose ownguesses would in
turn be subject to the same sort of equivocation/re-interpretation/translation, too.
In that case, if the meaning of every word is
occasion-sensitive, then so is that of any word that appears in a putative translation
or explanation of it
-- including Holborow's.
To be sure, what someone intends to achieve
by what they say does affect how we interpret the aims and intentions underlying
what they have just said, but this can't affect what the words they use mean.
Why
this is so will now be examined in more detail in the next sub-section.
As seems clear, Holborow failed to
distinguish
speaker's meaning from word or sentence
meaning. What a person intends to achieve with or by his/her words is surely
distinct from what those words mean. If that weren't so then we would have to admit that
the sentences listed below, for example, all meant the same if they were aimed
at making the same point.14a So, if someone uttered
each of these sentences with, say,
the intention of alarming their listeners, then that would
imply that they all meant the same -- i.e., they would all be synonymous!
V3: "Those pickets will stop
you strike-breaking!"
V4: "The Nazis know where you
live!"
V5: "Margaret Thatcher is
your biological mother!"
V6: "Tony Blair really
admires you!"
T1: "I want to alarm you!"
T2: "Holborow is wrong about
meaning being sensitive to intended effect and/or occasion of use."
Consider, for example, V1 and V2: if these
were uttered with the aim of alarming whoever they were directed at, then they
would both have to mean "I want to alarm you!", if Occasionalism were
true. In that case, presumably, "move"
and "your" from V1 and V2 must now mean "I" from T1; "and" and "house has" from
V1 and V2 must mean "want to" from T1, and so on. If not, then what
precisely is implied
by this view of meaning?
In fact, all the above were written
with the intention of showing that Holborow's ideas on this issue are misguided.
So, does this then imply that "Those pickets will stop you
strike-breaking!", for example, means "Holborow and Voloshinov
are wrong about word meaning being sensitive to intended effect and/or occasion
of use", if the use of both of these sentences had the same aim? Is this true of all the rest? Does
the sentence "Those pickets will stop you strike-breaking!" mean the same as
"The Nazis know where you live!" --, whose meaning actually is:
"Holborow and Voloshinov are wrong about meaning being sensitive to intended effect and/or
occasion of use"? On the account under review here, V3 must mean the same as T2!
This alone shows that
context can't narrow
down the options here, ruling certain 'translations' out as fanciful, since all of the
above (and countless more) could be used with the same intention
(expressed in T2, for instance), and in the same surroundings (of this Essay), even if all
of them are totally dissimilar and seemingly unrelated.
Worse
still: if these sentences
are synonymous, it must be possible to use them all
interchangeably. So, the next time someone wants to tell you that Tony Blair
admires you, all he/she has to do is say "Your house has just burnt down!", and
if they want to inform you that the Nazis know where you live, all they need do
is utter "Margaret Thatcher is your biological mother!" -- or, indeed,
"Holborow
and Voloshinov are wrong about meaning being
sensitive to intended effect and/or occasion of use".
If the occasion of
use means that these are synonymous when used with the same intention, then all this, and more, must surely
be the case. And if that is so, it isn't easy to see how any conceivable
context of utterance (short of the highly fanciful) would be able to tell you
that when someone says "Those pickets will stop you strike-breaking!" they
really mean "Your house has just burnt down!".
Of course, all of the above are
written, not spoken, examples of word use, but that can't form the basis of a successful
counter-response to the above objections. [Anyway, Holborow's reported utterance (i.e., "I'm hungry")
was written, too, as were Voloshinov's.] But, the same points could have
been made verbally, so they don't depend
specifically on
the written word. There is surely no significant philosophical or linguistic
difference (at least with respect to the meaning of words/sentences) if, say, V1-V6 were
to be printed
in a hard copy version of this Essay, or if they
were recorded and then played back as part of a speaking book, or even yelled in public
in the main square by the
Town Crier
at noon.
It could be maintained that there is a difference. Voloshinov was at pains to distinguish the
living, interactive use of
language between speakers, and the written word.
Or so it
might be argued...
That particular objection
has been dealt with
here.
Exception might be taken to the
above in view of the fact that Voloshinov and Holborow (along with several other
comrades) draw a clear distinction between the usual meaning of words
(their linguistic meaning) and their occasional meaning (which, from time to
time, Voloshinov seems to call, or associate with, "theme"). However, based on what the
aforementioned comrades actually
say, this distinction can't be sustained. Given their account, only un-used
words would actually have a linguistic meaning, while used words would
have an occasional meaning. Their theory seems to hold that words acquire new
meanings every time they are used. In that case, words would surely have a
linguistic meaningonly if they were never used! [On this, see
Note 16and
Note 17, below.]
That
neutralises the above
objection (i.e., that Voloshinov acknowledges that words have standard
meanings), since if anything can mean anything, then the phrase "standard
meaning" can, too, and we are back once more in that dialectical hole DM-fans seem to want to occupy.
An appeal to the dictionary meaning of words
would be to no avail, either, for those meanings are all culled from past
usage, which, given Voloshinov's theory, implies that only if a word has never
been used in the past (and hence fails to appear in a dictionary) would it have a
linguistic meaning!
Once more,
if this theory were correct, it seems that if a speaker (such as myself)
actually used the word "word" to mean "This expression means itinerant noises
like coughs", then its dictionary entry would have to be revised accordingly.
Either that, or, the linguistic/dictionary
meaning of a word like "cough" would actually be irrelevant, or
even unrelated, to its occasional meaning. But, if that were so, why would
anyone use that particular word? Any word, it seems will do, given the right
intention. And, if that were so, no one would bother compiling
dictionaries, since they would be full of useless definitions that no one observed,
and words that no one ever used.15
This is quite apart from the obvious fact
that if speaker's meaning determined the meaning of words, "dictionary" and
"entry" could mean anything, too!
[It is important to note once more that, in
the above comments, the distinction between the meaning of a word and the sense
of a sentence has been deliberately blurred. That is because the writers whose
views are being criticised here invariably fail to distinguish between them.
So, in order to expose the ridiculous nature of their theories, I am force to
employ words as they use them! It is
also arguable that part of the reason their ideas lapse into confusion so readily is that
they fail to notice this (rather obvious!) fact about the use of language. The distinction itself will be explored in more
detail below.]
Someone might object that the above examples (i.e.,
V1-V6) are highly fanciful and contrived; in
which case they can't be used against Voloshinov or Holborow. But, that isn't so. Sentences like these are uttered every day.
Anyway, the real point is that according to Voloshinov and Holborow any sentences
uttered with the aim of alarming hearers (and who can deny that this
happens many times a day around the world?) will all have the same meaning. The
actual choice of examples is therefore irrelevant.
With respect
to the use of a non-English sentence -- such as "La
plume de ma tante est sur le bureau de mon oncle"
("My aunt's pen is on my uncle's bureau") -- if it were uttered with the
intention of alarming someone, it needn't actually be translated since all
would know it meant, "I want to alarm you". Who then would need translators? Who
would need to learn a foreign language? Who would need a foreign language
dictionary of phrase book? So, when in Spain, for example, if a tourist wanted
to give her details to a Customs official, all she need say is "La
plume de ma tante est sur le bureau de mon oncle" if what she intended to
say was this "Hola,
yo me llamo Frieda
Farfenickle"
("Hi, my name is Frieda Farfenickle").
[Or are
Proper Names exempt from such occasionalist rules? If so, on what basis do they
get a pass?]
Despite this, it could be argued that the
circumstances surrounding the utterance of each of the above sentences (i.e.,
V1-V6) would all be different, and although they would all mean the same in the
abstract, their "theme" would be different, and thus their concrete meaning
would be different, too.
However, even if this were the case, any theory that had the consequence that in
the abstract, V1-V6 all meant the same (even if we knew what an 'abstract
meaning' was!) would still be subject to the objections advanced above.
Moreover, as we will soon see, Voloshinov is hopelessly unclear what he meant by
"theme", and his commentators are no less unclear, too. In that case, an appeal
to "theme" to rescue this theory would be about as helpful as a 'solution' to a
conundrum that had been written in the language of The Voynich Manuscript.
Having said
that, I do not wish to suggest that word meaning (which in turn involves the
content of an utterance) and speaker's meaning are hermetically-sealed aspects
of our use of language. I think Michael Dummett made the point rather well in an
interview he gave back in 1987, reprinted in one of his books:
"There
may be many cases in which the application is difficult, but the principle of
the distinction [i.e., between the content of an utterance and the point of
saying it -- RL] is quite clear. Even to understand one another we have
constantly to assess the motives or the intentions lying behind what is said;
sometimes this is perfectly obvious and sometimes it's difficult. 'Is he saying
that as a joke or is he serious?'. 'Did he think his remark relevant to the
previous conversation, or was he changing the subject?'.
"We need
constantly, in the course of conversation, to ask ourselves questions of this
kind. It is essential that linguistic utterances are in general voluntary,
rational actions and that we have to assess them as such. But the salient point
is that such assessments are like those of any non-linguistic action, as when we
ask ourselves, 'Why did he do that?', or 'What was he aiming at in doing that?'
One does not have, as it were, to learn that as part of acquiring the
language. One simply picks it up in the course of learning how to respond to
other people and interact with other people." [Dummett (1993b), p.182. Quotation
marks altered to conform with the conventions adopted at this site. Italic
emphasis in the original.]
When we try to
understand one another we often have to take into account both the meaning of
the words an interlocutor uses and why they are being used.
In
fact, this theory implies that a cough, for example, would actually mean the
same as a sneeze if it were intended to make someone jump --
and that a child's cry was synonymous with an alarm bell if both were aimed at
waking up the child-minder.
It
could be objected that a cough
or a child's cry aren't
linguistic expressions, hence they are inapt counter-examples.
However, if meaning were indeed
occasion-sensitive (as opposed to it being a feature of the public use of words
drawn from a finite vocabulary, etc.), then any sound or sign could count as a
linguistic move. If, say, someone coughed and they meant (speaker's meaning)
on that occasion: "Look out, the boss is coming!", then it seems, according
to this theory that that noise would mean (linguistically) the same
as: "Look out, the boss is coming!". In which case, for Holborow and Voloshinov,
it looks as if a cough would be just as much a linguistic act as uttering the
words: "Look out, the boss is coming!". Indeed, if that were so, the sentence "Look out, the boss is coming!"
would be
dispensable, and we could all use a cough from now on whenever we wanted to warn of
the boss's approach --, or, indeed, to report on that possibility in this Essay.
So, when I wrote:
If, say, someone coughed and
they meant (speaker's meaning) on that occasion: "Look out, the boss is
coming!", then it seems that (according to Holborow) this noise would mean
(linguistically) the same as the words: "Look out, the boss is coming!".
I could just as well from now on write:
If, say, someone coughed and
they meant (speaker's meaning) on that occasion: "COUGH!", then it seems that
(according to Holborow) this noise would mean (linguistically) the same as the
word: "COUGH!"
Which everyone committed to this theory would
understand, since the sentence "Look out, the boss is coming!" would, for
them, mean the same as a cough, or even "cough".
In that case, the ridiculous nature of the above should
now speak, or cough, for itself.
Again, it could be objected that this response only succeeds in undermining the
argument advanced in
this Essay (which was that the meaning of words and the sense of sentences
aren't in general dependent on contexts of utterance), for if the meaning of, say,
a cough is now admitted to be occasion-sensitive, then meaning in general must be
occasion-sensitive, contrary to what had been claimed.
That
objection is misguided. Given the theory under
consideration, and the example used above,
we would now have nothing into which we could 'translate' the said cough, since the original sentence "Look out, the boss is coming!" is
dispensable (it having been replaced, along with its meaning, by a
cough). If so, either (i) coughs would become meaningless by default -- they would not now be
translatable because the sentence they replaced, and which could be used to
translate them, has dropped from the language --, or (ii) if
coughs retained some sort of a meaning, it would then be equivalent to the now
unusable (or, from-now-on-and-forever-to-be-unused) sentence "Look out, the boss is
coming!" -- once again, it having passed from the language. Either way, coughs would thus have taken on
the role of the now defunct type sentence "Look out, the boss is coming!".
As should seem clear, coughs would thus
become occasion-insensitive, since they would now have this
meaning: "Look out, the boss is coming!", and no other. The whole point of the exercise
would be lost and occasion-sensitivity will have been transformed into its alter-ego: occasion-insensitivity!
[Of course, this would create problems for
those who cough because they have a tickle in the throat, or are suffering from
a chest complaint. Might they come to be described as serial boss-approach-warners? And
what are we to say of the patients in tuberculosis wards? Are they all warning
one another of the same or different bosses?]
On the other hand, (iii) even assuming that the
sentence "Look out, the boss is coming!" doesn't slip from the language in the
manner suggested above (or in some other way), the point at issue here would
still be that whatever handle we have on occasion-sensitive acts of communication,
it must rely on linguistic expressions that aren't themselves constrained by
occasion-sensitivity.
So, the point made in the main body of this
Essay wasn't that
nothing is occasion-sensitive, but that not everything could possibly
be occasion-sensitive.
If the translation into language of coughs and
other assorted random noises -- so that they could be taken to mean things like "Look out,
the boss is coming!" -- were itself dependent on nothing but occasion-sensitive
materials (including the sentence "Look out, the boss is coming!"), we
would be involved in trying to comprehend something (the cough) in terms of
something else (its supposed sentential equivalent) that would itself be in need
of an unravelling process all of its own. Down such a road, I fear, lies another
infinite regress, in which impenetrable thicket all meaning would soon become lost.
Again, it could be objected that this still
fails to address the main issue: coughs (etc.) are non-linguistic acts;
hence, they aren't at all what Holborow was adverting to.
The
point about using examples such as coughs and cries (etc.) is that Holborow's
view can't in the end distinguish between the
occasion-sensitivity of such sounds (etc.) and genuine linguistic acts, which she
says are also constrained in this way (i.e., in that they, too, are subject to
the constraints of occasion-sensitivity). If the meaning of both is occasion-sensitive
(whereas the view advanced in this Essay is that only the meaning of the former
is so constrained, when used in the manner suggested), then
Holborow still needs other criteria to tell them apart. If a cough could mean
(speaker's meaning) the same as "Look out, the boss is coming!" (and who can doubt that?),
and any other randomly chosen sentence (such as "My gerbil is dead!") could
also mean "Look out, the boss is coming!" (as it seems it could, given
Holborow's view; that is, if the person using "My gerbil is dead!" actually
meant it, or intended it, as the coded message/warning, "Look out, the boss is coming!"), then the
distinction between linguistic expressions and mere sounds would be
lost, and the points raised in the main body of this Essay would stand.
Despite this, it might be felt that since
coughs and itinerant noises aren't part of a standardised vocabulary, they can't
be interpreted along the lines outlined above. But, if a linguistic expression
can be used to mean anything whatsoever (even something wildly divergent
from the norm -- or to use Holborow's words: it could be "different in every
aspect") then standardised vocabularies must surely drop out as irrelevant. For
example, if the word "cough" (not the actual noise, or action, but the
word itself) could mean, say, "My armadillo is sick", then any connection it
once might have had with its own dictionary entry (or its established meaning) would be lost (as
would those of the other four words used: "my", "armadillo", "is" and "sick").
In that case, the links that the word "cough" had with its standard meaning
would be severed, too. And, if that is the case, an actual cough could
then mean the same as "My armadillo is sick", or any other word or set of words
in the dictionary or the language, which could in turn mean anything themselves, including
coughs.
It might now be objected that an actual
cough isn't a word, so it can't perform the roles assigned to it
in the above paragraphs.
But, if anything can mean anything,
we must
surely lose touch with the meaning of the word "word" itself. On this view, the
word "word" could in fact mean: "This expression actually means
itinerant noises like coughs"
if it were so 'intended' by deviant linguists (or if I so intend it
here). If Occasionalism were true, this possibility can't be ruled out.
Occasionalism permits any word to mean anything if it is so intended, or if the
circumstances suggest it. And that includes words and phrases
like "meaning", "sentence", "word", "cough", "and so on"..., and so on.
Perhaps more significantly, given this view,
it would be possible for plain gibberish to have the same effect on an
audience that a perfectly ordinary sentence had on those who heard both, and
hence for the two to
mean the same, if there were an intention to that effect. In which case, we should have to admit that a nonsensical string
of letters, such as:
[Or,
indeed, if both were intended to annoy or
perplex supporters of Voloshinov's 'theory' of meaning, and succeeded in doing
one or both.]
Moreover, if, as
Voloshinov argues, sentence and word meaning
(not speaker's meaning) were dependent on context and occasion of use, then words
divorced from every context would have no meaning at all.16
So, for instance, the sentence "Voloshinov is correct about meaning and theme"
would mean nothing until someone actually uttered it in a particular context
with a specific intention. But, if it
had no meaning, why would anyone choose to utter it? Why would anyone select
such a meaningless string of words? They might just as well say something
genuinely meaningless like: "BuBuBu" --, which, on this theory, should gain a sense
from being uttered with a special aim in mind. But, wouldn't they rather utter
"I'm hungry" in order to mean "Voloshinov is correct about meaning and theme"?
In fact, the situation is far worse than this; if a sentence such as "Voloshinov
is correct about meaning and theme" is meaningless -- that is, it is
meaningless until it is uttered, if Voloshinov is to be believed --,
then surely no one would use "I'm hungry" to mean something that is
meaningless, i.e., "Voloshinov is correct about
meaning and theme", since both of these sentences would fail to mean anything if
someone uttered "I'm
hungry"!
Plainly, that is because the sentence "Voloshinov is correct about meaning and
theme" wouldn't have been used --
merely
indirectly adverted to --, and so would stillbe meaningless.
Hence, at least
here, a sentence like "I'm hungry", used in a given context, couldn't have its
meaning determined by its occasional use, since it would now have no meaning at
all! Again, that is because it is supposed 'to mean' the same as the un-uttered
sentence, "Voloshinov is correct about meaning and theme", which, because it
hasn't been uttered, has no meaning -- thus implying that "I'm hungry" has no meaning,
either!
This must be the case with all sentences
before they are uttered. "I'm hungry" can't mean anything, no matter how we try
to translate or make sense of it since, if it is used in the way that Voloshinov
and Holborow imagine. It could only 'mean the same' as some other sentence that
wasn't uttered, merely alluded to --
which, because it hasn't been uttered, can have no meaning. Consequently, if true, Voloshinov's theory
concerning the occasion-sensitivity of words and/or sentences would, ipso facto,
become untrue
-- a result that represents yet another ironic 'dialectical inversion', one feels.17
Again, someone could object that this ignores
Voloshinov's distinction between "theme" and meaning. That response will be dealt with
presently.
Alternatively, Voloshinov's theory seems to
imply that
interlocutors must ascertain each other's aims and intentions
before they can be expected to grasp what was said. This would then involve the latter in
having to link aspects and surroundings of any utterance (which, we must recall, are as yet meaningless
to each hearer) -- that
are relevant
to that end in that context -- to an indeterminate number of possible meanings. But, if the
said utterance has no meaning until
it is interpreted (or even until it is uttered!) what is there that hearers could latch on to in the local
environment that might help them to that end?18
It is little use replying that speakers and hearers accomplish this every day, since, on
this view, it is hard to see how that is possible. To be sure, we often make an
educated guess when we encounter the odd things we sometimes hear, but this
typically takes place against a background that consists of an array of words we
already comprehend. But, if all the words we
hear or read were meaningless before we interpreted them, or even before
they were uttered, we would surely be like those trapped in a foreign land,
confronted by a language we had never encountered before.18a0
In fact, we would normally
say that interlocutors communicate because they possess a common language,
which benefits from a shared vocabulary with reasonably settled meanings, and
which both parties
already understand. What
they do not normally do is revise language during every
conversation. And yet, it seems that they must do this if Voloshinov were correct.18a
And, we certainly can't appeal to past
experience to help out here. That is,
interlocutors can't rely on
a previous
use of the same words in the same contexts to ascertain what is intended in or by their
current
employment, for Holborow and Voloshinov tell us that any and all words have
completely different meanings each time they see the light of day. Not only
that, but the circumstances surrounding the vast majority of (if not all) utterances are completely novel,
too.18b
Nevertheless, each new context brings with it
new meanings, according to Voloshinov and Holborow. These novel connotations would not only have
to be supplied by both parties to a conversation, they would have to coincide
for each of the parties to that conversation if communication is to succeed. But, how might
this be
achieved if
neither
interlocutor understands what is said in advance of it being said (since it
is meaningless until it said, given this view), or if
neither party has experienced exactly these circumstances before? Indeed,
given the fact that no two human beings are completely alike, nor have they even
remotely shared the same experiences, this theory implies that no act of communication
would ever succeed. [There is much more on this
below.]
In fact, anyone overhearing such a
conversation, and not knowing the aims or intentions of the interlocutors, wouldn't understand what they had overheard, either -- if Voloshinov were
correct. In general, this is patently
incorrect. We readily understand things not addressed to us. We might
sometimes miss the point of why it was said (just as we might not always
grasp every single detail (if, say, some of the individuals mentioned in such a
conversation were unknown to us)), but that is an entirely separate matter.
More difficult to explain, however, is the fact that
hearers would have to express to themselves in their own language (i.e.,
in their own
idiolect, or
their "inner speech") the aims and intentions of their interlocutors. This would
involve them in representing the latter in a language that wasn't itself subject to the same
constraints. Let us call such a language (i.e., one that
is comprehensible without recourse to any further occasion-sensitive
protocols)
an "immediate language".
Hence, a language understood
without the need for any
further processing would be an immediate language. But, if such an 'internal'
language is indeed immediate, then language itself at some point must be
occasion-independent -- namely, just here, internally. And yet, if
some language is internally immediate, why not the language we use 'externally'? The only possible reason
for denying this would seem to
be that the hypothesised language here is 'internal' to an individual. That option will be considered
presently, and neutralised.
Conversely, if a hearer's own 'internal language' is also occasion-sensitive -- that is, if
it isn't an
immediate language, after all --, then an infinite regress must ensue as interpretation upon
interpretation is layered on top of each incoming message, and each subsequent
translation (and translation of a translation of a translation…) is rendered into that
individual's
inner, inner, inner..., idiolect.
As already noted, this theory would mean that a word in fact possessed no meaning
at all (i.e., no intrinsic meaning) until someone deigned to give it one by using
it. But, if that were
the case, no one would be
able to ascertain whether or not they had settled on the same meaning as
that which had been latched on to by any of their interlocutors. Not only that, any attempt to resolve even
this
quandary would itself stall until a decision had been made (in no language at all,
presumably!) whether or not each party to a conversation meant the same by the
phrase "same meaning", let alone any of the other terms on offer.
And how might that minor miracle be
achieved, for goodness sake?19
On the contrary, if a hearer hasn't
already grasped what is said to him/her, the assumed (internal) process of
interpretation can't even begin. That is because hearers wouldn't be able to
distinguish what was meaningfully communicated to them from irrelevant or
pointless remarks -- or, for that matter, from gibberish and incidental noises
(as we saw earlier).
If they had to decipher words directed at them based only on contexts of
utterance and/or on the aims and intentions of speakers, then they would also have to be able to ascertain
which aspects of those contexts were relevant to that end (again, as noted earlier). [As
we have seen, "context of utterance is itself hopelessly vague!] But, that would
involve them in understanding the said utterance first. Otherwise,
anything could be counted as 'relevant'. If we are to interpret the aims and
intentions of fellow speakers successfully, they must address us in terms we
already understand so that we can layer on top of whatever they say any
additional gloss we deem appropriate -- as we try to discern their intentions,
and as each occasion demands, or otherwise. It can't work the other way round.
We don't divine what others intend by a sort of magical, languageless intuition,
which subsequently enables us to put meaning to their words. That is why we
don't have to wait to ascertain the point of someone uttering, say, "The
BNP is a Nazi Party" before we understand it. Once more, we must
first grasp what is said if we are to figure out the point of someone
saying whatever they say. Hence, the point behind the present author's inclusion of
this sentence about the BNP (which was in fact to argue that Voloshinov is wrong in what he
says) has nothing at all to do with what those words mean. Although, my
intentions certainly affected what I meant (speaker's meaning) to achieve by
using them.
In fact, it is quite easy to see that
Voloshinov's suppositions aren't viable since we already
understand the exemplary sentences from earlier (i.e., V1-V6) before we
know their context of utterance, or the point, or the purpose, anyone might intend by uttering
them:
V1: "Move, and you're dead!"
V2: "Your house has just
burnt down!"
V3: "Those pickets will stop
you strike-breaking!"
V4: "The Nazis know where you
live!"
V5: "Margaret Thatcher is
your biological mother!"
V6: "Tony Blair really
admires you!"
Moreover,
because of our facility with language, and as a result of our socialisation, we also know, or
can form, an educated guess concerning the sorts of contexts
in which such sentences could plausibly feature, or be uttered, and it is this that
helps us
interpret the aims and intentions of others when they arise.20
Of course, we do this with such ease
that we don't notice it, just as we can,
for example,
walk without noticing or knowing how we do it. And that is why we feel we can exclude (as highly
unlikely) most of the fanciful interpretations advanced above of what the
hypothetical child, for example, might have meant by "I'm hungry". That
is also why readers who have made it this far can easily comprehend sentences
like V1-V6 whether or not they are aware that these sentences have all been
fabricated to a specific end, and have no context other than the spurious ones provided here, or
mentioned earlier.
And it is a safe bet that that won't have affected the reader's understanding of
these perfectly ordinary sentences. That fact would be totally inexplicable if
linguistic meaning were context-dependent.
The seeming plausibility of Voloshinov and Holborow's examples (or,
indeed, any imaginative interpretation put upon them) trades on a facility
possessed by all competent language users: that is, of being able to understand sentences independently
of their context of utterance -- saving, of course, those that have
indexical features
(etc.). And this still remains the case even when a reference to the
context of utterance could help hearers ascertain the aims and intentions
of their interlocutors. That is why it isn't necessary for Voloshinov's readers
to know the contexts surrounding his particular use of language in order
to understand him, or his work; indeed, it is because they already grasp the words he chose to use
that they can recognise in general the types of contexts in which the examples
he cites might plausibly occur (should they want to do this), as well as the sorts of aims and intentions they
might reveal or express. That is also why the implication that sentences like V1-V6 have
the same meaning strikes us as completely bizarre, and why we can see
that, despite the fact that uttering any or all of them could have the same effect,
or arise from
the same intention, they don't
have the same meaning.20a
Finally, it is also why we can all see that inscriptions like those in V7 (i.e., "BBB XXX ZZZ QQQ TTT") are totally meaningless,
despite the fact that V7 could have had the same effect on someone as a
meaningful sentence, and be employed to the same ends. Even though V7 has a use -- for example, to
make the very point that it is meaningless21
-- it is, nevertheless, mere babble. Using it to make that very point doesn't show that its
meaning is that it is meaningless. Plainly not, otherwise it would have no
meaning by meaning that -- indeed, in that case, its meaning would be that it had no meaning!22
As we will see later in this Essay,
despite what they might appear to say, the reason why theorists like
Voloshinov (and, it seems, the other comrades mentioned above) find
Occasionalism so
attractive is that they have accepted the traditional view that 'acts of
meaning' are 'inner mental events', private to each individual. Quite apart from
the fact that this theory sits awkwardly with the belief that language is a social phenomenon, if it were true it would actually prevent -- not
facilitate -- communication. Indeed, since this view of meaning is plainly based
on the representational model, it is hardly surprising that it undermines
communication.
Now, there are places in Voloshinov's work where he sort of
half
recognises this, but his grasp of this idea isn't secure enough for him to
appreciate that he has only succeeded in undermining it because of the other things he says about meaning
-- or, indeed, vice versa.23
"Speech had first to come
into being and develop in the process of the social intercourse of organisms
so that afterward it could enter within the organism and become inner speech."
[Voloshinov (1973), p.39. Bold emphases added.]
Here, the order of events is plainly as
follows: speech first, "inner speech"
second. But, if to understand something is:
"...to refer a particular inner sign to a unity consisting of
other inner signs...." [Ibid., p.35.]
And:
"...understanding itself can come about only with in
some kind of semiotic material (e.g., inner speech), that sign bears upon sign,
that consciousness itself can arise and become a viable fact only in the
material embodiment of signs...understanding is a response to a
sign with
signs...." [Ibid., p.11. Italic emphasis in the original.]
Then this can only mean that
before "inner speech" developed, no one would be able to understand a single thing that was
said to them! Why then would anyone want to speak or engage in conversation? What use would language
be if no one could comprehend what anyone else was saying?
Furthermore, if understanding
is accomplished by "a response to a sign with signs", how would it be
possible for this
"response"
to have originally begun?
Indeed, how could outer signs ever have become 'inner signs' in the first place?
And yet, this idea is central to Voloshinov's theory. Unfortunately, however, he torpedoed
his own ideas when he said:
"There is no outer sign
without an inner sign." [Ibid., p.39.]
This can only mean that the development of
speech can't have originally used signs! Recall the order of events, according
to Voloshinov, is as follows:
(1) Outer speech first,
"inner speech" second.
(2) Understanding requires "inner speech" and
a "response" between signs.
(3) But, there can be no outer signs (speech) without
inner signs.
(4)
Therefore, (i) outer speech can't use any signs, and hence (ii) there can be
no outer speech!
That is because when language began there
were, as yet, no inner signs!
But, there could be no outer signs if there are no inner signs [Point (3)],
and there can be no inner signs if there are no outer signs [Point (1)]!
The
whole process can't begin!
If so, no sign could ever have become part of,
or incorporated into, inner speech,
either. In which case, if Voloshinov is to be believed, that must mean that no one in human
history will have understood anything that had ever been
said to them!
This predicament would extend to children
starting to learn to speak, too. Manifestly, they have no inner signs when they
are born. Hence, for them there can be no outer signs, either! [Point (3), once
more.] But, if there are no outer signs, then no child could
build up her own stock of inner signs. In that case, no
child could develop "inner speech". Thus, no child would ever
understand a single thing said to it!
The only avenue of escape for Voloshinov would
seem to be for
him to argue that "inner speech" developed first (thus rejecting
Point (1)). Given such a
scenario, the private use of
signs in "inner speech" would what allowed outer speech to develop. But, this is just the
bourgeois individualist theory (which is
an early modern version of the much older
Platonic-Christian Paradigm)
that has
dominated 'western' thought since Ancient
Greek times -- and yet this is the only way Voloshinov's theory can get off the
ground. His social theory thus inverts into an anti-social theory. Of
course, this
might help explain why Voloshinov equivocated between a social and an individualistic
account of meaning and speech.
Once again, we see the untoward consequences
of buying into Traditional Thought, and (here) the idea that understanding is an
'inner
process'.
It might be thought that Voloshinov's
introduction of "theme"
is capable of breathing life into these dead signs, and thus of neutralising the above
points. Unfortunately, as we will
soon discover, instead of
breathing life into signs, "theme" injects them full of
Formalin.
Anyway, exactly how "theme" can create outer
signs if there are as yet no inner signs is still unclear. Naturally, this means
that not even "theme" can rescue Voloshinov's theory from the nonsensical
implication that understanding would be impossible. Hence, the introduction of
"theme", here, would be about as much use as a margarine cement mixer.
Naturally, this isn't to deny that
languages change, nor is it to reject the idea that the spoken word is part of a living
system of inter-communication -- and neither is it to repudiate the view that context
(among other things) can affect speaker's
meaning --, nor even that social parameters, or forces, have a decisive effect on the
development of language. The above comments are merely aimed at reminding us
that whatever its is that lends to sentences the sense they have (and to words their
meaning), it can't be context of utterance, or the use of inner signs. Speaker's meaning is
parasitic on much more fundamental aspects of the social nature of language --
those that Voloshinov and the other comrades mentioned above appear not to have
noticed. What these features are have been hinted at throughout this site
(especially here).
Other important logical aspects of language will be rehearsed below.
As pointed out
here and in
Note 29andNote 86,
Voloshinov and the other comrades mentioned above seem to have ignored the
important distinction between
the meaning of words and the sense of
indicative sentences. This is a failing they
share with the majority of Philosophers who have written on this
subject --, that is, up until just over a hundred years ago. Beginning with ideas that
were first seriously mooted in
Frege's work, Philosophers working in the
Analytic tradition have generally (but not unanimously) maintained the
opinion
that an account of language that ignores the above distinction is radically flawed.24
This observation, of course, isn't based on
supposition, nor is it mere dogma. We all recognise it to be the case when
we are reminded of it. Even competent speakers of a language would fail to
comprehend what was said to them if it contained words they had never encountered before; and they would remain in that state until
the meaning of those words had been explained to them. In stark contrast, all
of us readily understand sentences wehave never heard before
(saving, of course, those that contain such novel words). This indicates that
word-meaning and sentence-sense can't be the same, otherwise this wouldn't happen.
For example, the words in the previous
paragraph could be reassembled into different combinations, and, providing each
new arrangement
satisfied certain syntactic and pragmatic constraints, they would be readily
understood by most competent speakers of English. However, if the names
"Tony Blair" and "Leon Trotsky" were mixed up to give "Leon
Blair" and "Tony Trotsky" no one would know who was being referred to -- even if they knew who the
original characters were.
Moreover, if the following sentence were
uttered:
V9: "Tony Blair's wrist watch
has been eaten by a Koala Bear",
the vast majority of English language
speakers would understand it even though that sentence (in all probability) has
never been written, spoken or heard before by anyone in human history; and they would comprehend it
without knowing whether it was true or whether it was false, since they would know under what
circumstances it would or could be either one of these. In short, they would understand its sense.
[This theme has been greatly expanded upon in Essay Twelve
Part One.]
Contrast V9 with the following:
V10: Bogomil.
Now, it is highly likely that most English
speakers have never encountered this word before. Even though it contains
familiar letters, no one would understand it until its meaning had been
explained to them, or they looked it up in a dictionary.25
If this word appeared in a sentence, that sentence would similarly remain
incomprehensible until its meaning had been clarified -- unless, of
course, on the rare occasion it had been
worked out by means of an educated guess, perhaps.
This difference between how we read, receive or understand words and sentences shows
that the sense of the latter and the meaning of the former are distinct logical
features of our use of language.25a
All traditional, ancient, and most modern 'theories of meaning' founder on this fact alone.26
Nevertheless, this isn't an obscure feature of language, something
that only
those who study Linguistics or Analytic Philosophy are aware of. All of us
appreciate its validity (and recognise its force) when it is pointed out to us
since we depend
on it to communicate everyday of our lives. We all understand sentences we
have never seen or heard before, and we all fail to comprehend words we have never
previously encountered. The fact that this distinction had been ignored
for thousands of years (and is still largely ignored today) by Traditional Theorists shows how
divorced from ordinary life -- and how obsessed with atomistic theories of
language -- such thinkers had (and have) become (and this is so for reasons
examined in Essay Twelve (summary
here)).26a
Naturally,
this means that serious errors were introduced into thought by
previous generations of Philosophers, who not only ignored, they disdained, the vernacular,
preferring instead a fetishised view of language.27
[Again, why this is so and why it is
significant are explained in more detail in Essay Twelve
Part One.]
Moreover, any
analysis of language that
tried to explain the meaning of words and the sense of sentences by an
indiscriminate appeal to speaker'susage (i.e., to what a speaker
idiosyncratically intends to convey or achieve by employing certain
sentences or words) would similarly fail to account for the phenomenon noted
above. If the sense of a sentence were (generally) based on the use to which a
speaker might (idiosyncratically) put it, then competent speakers of a language
wouldn't be able to understand sentences they had never heard before -- just as they now
fail to comprehend novel words they have never previously encountered. If the idiosyncratic use of words determined
meaning, and if intentions or contexts of utterance determined the sense of sentences, word
meaning and sentence sense would be all of a piece.28
In that case, language users wouldn't be able to understand both words and
sentences they had never met before, rather than just failing to comprehend
words they had never heard before. Since most of the sentences we encounter
are novel, while the words they contain are not, we would in such circumstances
fail to understand anything said to us until everything had been explained -- but which
explanation would also need explaining..., and so on --, and for the same reason.
This might be why Socialist Worker
had the following to say a few years ago (ironically, just before the
UK-SWP became engulfed in the scandal around
'Comrade Delta' and allegations of rape, which succeeded in
halving its membership
over the next year following a series of mass resignations):
"But language
can't be divorced from society and used however
we like. Meanings can't be dictated by
individual intentions.... [One] example is last
year's 'Slutwalk' protests against sexism. Women
organised these demonstrations after
a police
officer in Canada suggested that women should
avoid 'dressing like sluts' to protect
themselves from rape. 'Slut' is a derogatory
term used to insult women based on their
supposed sexual behaviour. It promotes the false
idea of women as naturally pure and chaste in
order to denounce those who don't conform to
this stereotype. The protests highlighted
women's oppression and attracted people who
wanted to fight it. But adopting the word 'slut'
didn't change its meaning in wider society. It
created divisions in the ranks of those who
wanted to fight sexism. The bigots can draw
confidence from this. They see women using the
word 'slut' as a licence for them to use it too.
So 'reclaiming' oppressive words doesn't
ultimately strengthen the fight against
oppression -- it makes it harder." [Socialist
Worker 2290, 18/02/2012, p.9. Quotation
marks altered to conform with the conventions
adopted at this site. Paragraphs merged; bold
emphasis and link added.]
In
addition to the above, a plausible account of language must be able to relate the clear
distinction we draw between the sense of sentences and the meaning of words to our capacity to form
and comprehend novel sentences -- the senses of which are related (sometimes
systematically, sometimes not) to the manner in which their constituent words
have been combined (etc.), coupled with the meaning of the words used. It must, therefore, connect the sense of
a sentence to
rules of syntax, which in turn must be related to something other
than idiosyncratic use.29
It is here, perhaps, where the weakness
of Voloshinov's 'theory' is most obvious: the confusion of the meaning of words
with the sense of sentences.30
Again, as already noted, he isn't alone in taking this wrong turn; it is a
major failing of all Traditional Theories of language (and many modern ones,
too).
Oddly enough, this atomistic approach to discourse also plagues accounts of language
written by several prominent Wittgensteinians -- including, it seems, practically every
'Wittgensteinian' who is also a social scientist.31
At this point, it could be objected that
Voloshinov's theory of language isn't susceptible to the above criticisms. That is because of (a) The distinction he drew between meaning and "theme",
(b) His insistence that written words
and spoken words are subject to different criteria, and because of (c)
His opposition to what he
calls "abstract objectivism".
Item (b) has already been discussed
(here), so I will consider (a)
first.
Unfortunately, Voloshinov's comments on "theme" are far too
sketchy and
confused for anyone to be able to say what he actually meant by this word!
To compound matters, Holborow, Parrington and Doherty provide us with little
help in this regard, either. Indeed, Parrington doesn't even use the term!
Nevertheless, Voloshinov had this to say about
"theme":
"Let us agree to call the
entity which becomes the object of a sign the theme of the sign. Each
fully fledged sign has its theme. And so every verbal performance has its theme.
"An ideological theme is
always socially accentuated. Of course, all the social accents of ideological
themes make their way into the individual consciousness (which, as we know, is
ideological through and through) and there take on the semblance of individual
accents, since the individual consciousness assimilated them as its own. However,
the source of these accents is not the individual consciousness. Accent, as
such, is interindividual....
"The theme of an ideological
sign and the form of an ideological sign are inextricably bound together and are
separable only in the abstract…." [Voloshinov (1973), p.22.]
"A definite and unitary
meaning, a unitary significance, is a property belonging to any utterance as
a whole. Let us call the significance of a whole utterance its theme…."
[Added in a footnote: "The
term is, of course, a provisional one. Theme in our sense embraces its
implementation as well; therefore our concept must not be confused with that of
a theme in a literary work. The concept of 'thematic unity' would be closer to
what we mean."]
"…The theme must be unitary,
otherwise we would have no basis for talking about any one utterance. The theme
of an utterance is individual and unreproducible, just as the utterance itself
is individual and unreproducible. The theme is the expression of the concrete,
historical situation that engendered the utterance. The utterance 'What time is
it?' has a different meaning each time it is used, and hence, in accordance with
our terminology, has a different theme, depending on the concrete historical
situation ('historical' here in microscopic dimensions) during which it is
enunciated and of which, in essence, it is a part.
"It follows, then, that the
theme of an utterance is determined not only by the linguistic forms that
comprise it -- words, morphological and syntactic structures, sounds, and
intonation -- but also by extraverbal factors of the situation. Should we miss
these situational factors, we would be as little able to understand an utterance
as if we were to miss its most important words. The theme of an utterance is
concrete -- as concrete as the historical instant to which the utterance belongs.
Only an utterance taken in its full, concrete scope as an historical
phenomenon possesses a theme. That is what is meant by the theme of an
utterance.
"...Together with theme or,
rather, within the theme, there is also the meaning that belongs to an
utterance. By meaning, as distinguished from theme, we understand all those
aspects of the utterance that are reproducible and self-identical
in all instances of repetition. Of course, these aspects are abstract: they have
no concrete, autonomous existence in an artificially isolated form, but, at the
same time, they do constitute an essential and inseparable part of the
utterance. The theme of an utterance is, in essence, indivisible. The meaning of
an utterance, on the contrary, does break down into a set of meanings belonging
to each of the various linguistic elements of which the utterance consists. The
unreproducible theme of the utterance 'What time is it?' taken in its
indissoluble connection with the concrete historical situation, cannot be
divided into elements. The meaning of the utterance 'What time is it?' -- a
meaning that, of course, remains the same in all historical instances of its
enunciation -- is made up of the meanings of the words…that form the
construction of the utterance.
"Theme is a complex,
dynamic system of signs that attempts to be adequate to a given instant of
generative process. Theme is reaction by the consciousness in its generative
process to the generative process of existence. Meaning is the technical
apparatus for the implementation of theme. Of course, no absolute,
mechanistic boundary can be drawn between theme and meaning. There is no theme
without meaning and no meaning without theme. Moreover, it is even impossible to
convey the meaning of a particular word…without having made it an element of
theme, i.e., without having constructed an 'example' utterance. On the other
hand, a theme must base itself on some kind of fixity of meaning; otherwise it
loses its connection with what came before and what comes after -- i.e., it
altogether loses its significance….
[Quoting
Marr]
"'But was such an all-meaning word in fact a word?' we might be asked. Yes,
precisely a word. If, on the contrary, a certain sound complex had only one
single, inert, and invariable meaning, then such a complex would not be a word,
not a sign, but only a signal. Multiplicity of meanings is the constitutive
feature of a word. As regard the all-meaning word of which Marr speaks, we
can say the following: such a word in essence has virtually no meaning; it is
all theme. Its meaning is inseparable from the concrete situation of its
implementation. This meaning is different each time, just as the situation
is different each time. Thus the theme, in this case, subsumed meaning under
itself and dissolved it before meaning had any chance to consolidate and
congeal. But as language developed further, as its stock of sound complexes
expanded, meaning began to congeal along lines that were basic and most frequent
in the life of the community for the thematic application of this or that word.
"Theme, as we have said, is
an attribute of a whole utterance only; it can belong to a separate word only
inasmuch as that word operates in the capacity of a whole utterance…. Meaning,
on the other hand, belongs to an element or aggregate of elements in their
relation to the whole. Of course, if we entirely disregard this relation to the
whole (i.e., to the utterance), we shall entirely forfeit meaning. That is the
reason why a sharp boundary between theme and meaning cannot be drawn.
"The most accurate way of
formulating the interrelationship between theme and meaning is in the following
terms. Theme is the upper, actual limit of linguistic significance; in
essence, only theme means something definite. Meaning is the lower limit
of linguistic significance. Meaning, in essence, means nothing; it only
possesses potentiality -- the possibility of having a meaning within a concrete
theme. Investigation of the meaning of one or another linguistic element can
proceed, in terms of our definition, in one of two directions: either in the
direction of the upper limit, toward theme, in which case it would be
investigation of the contextual meaning of a given word within the conditions of
a concrete utterance; or investigation can aim toward the lower limit, the limit
of meaning, in which case it would be investigation of the meaning of a word in
the system of language or, in other words, investigation of a dictionary word.
"A distinction between theme
and meaning and a proper understanding of their interrelationship are vital
steps in constructing a genuine science of meanings. Total failure to comprehend
their importance has persisted to the present day. Such discriminations as those
between a word's usual and occasional meanings, between its
central and lateral meanings, between its denotation and connotation, etc., are
fundamentally unsatisfactory. The basic tendency underlying all such
discriminations -- the tendency to ascribe greater value to the central, usual
aspect of meaning, presupposing that that aspect really does exist and is stable
-- is completely fallacious. Moreover, it would leave theme unaccounted for,
since, theme, of course, can by no means be reduced to the status of the
occasional or lateral meaning of words." [Ibid., pp.99-102. Italic
emphases in the original.]31a
It would, of course, be unfair to criticise
Voloshinov too much for the sketchy nature of these comments since he admitted his ideas were
provisional. He was prevented from developing them into full, or even fuller,
coherence by the fact that he disappeared during the Stalinist purges, which
began soon after he wrote his book. Those who have followed in his footsteps
and who have simply regurgitated his ideas aren't so easily excused.
Even so, it is worth pointing out that
Voloshinov supplied his readers with little or no evidence to
substantiate this distinction between "theme" and meaning -- or, indeed, much
else that he dogmatically asserts in his
book. In fact, readers will find no experimental results, observations,
surveys, tables, graphs or
figures (or even so much as a single reference to other studies which record
or report such data/information!) and no statistical analyses and
mathematics in support of a single substantive conclusion he draws in the entire work. That
fact alone ought to worry comrades who regard Voloshinov's book as a major
contribution to the Science of Linguistics.
However, the fact that it
doesn't, should surprise no one who is familiar with the
a priori and dogmatic nature
of 'dialectical philosophy'.
In like manner, Parrington, Holborow
and Doherty offer little (or any) empirical evidence to back-up their claims that Voloshinov's ideas are of
any scientific merit -- or,
indeed, for allaying the concerns of those who might be tempted to conclude that his
ideas have been
imposed on reality, contrary to what
we are told dialecticians never do.32
Anyway, given the nature of what
Voloshinov actually said, no evidence could ever have been, or could ever be found
to support his claims. That is because, by definition, "theme" is
totally inaccessible
since
it is essentially occasion-sensitive. This means that not only is "theme"
inaccessible to scientific enquiry, it is inaccessible to each and every
participant in a conversation --
since, as Voloshinov himself admits, "theme" is affected by the microscopic
differences between cases:
"[T]heme must be unitary,
otherwise we would have no basis for talking about any one utterance. The theme
of an utterance is individual andunreproducible, just as the
utterance itself is individual and unreproducible. The theme is the
expression of the concrete, historical situation that engendered the utterance.
The utterance 'What time is it?' has a different meaning each time it is
used, and hence, in accordance with our terminology, has a different theme,
depending on the concrete historical situation ('historical' here in
microscopic dimensions) during which it is enunciated and of which, in
essence, it is a part." [Ibid., p.99. Bold emphases added.]
In that case, "theme" can serve no part in
effecting communication, even if we knew what "theme" was. [Those who think we
do know what "theme" is are encouraged to continue reading, after which their premature feelings
of confidence should emerge somewhat..., shall we say..., shaken.]
If the above is indeed the case, it might well be
wondered how anyone could possibly tell whether an utterance does, or does not, have a "theme".
If something is intrinsically unique, has a transient nature and is
ephemeral
in the extreme, how
might its existence even be detected, let alone confirmed?32a
In fact, in the place of supporting evidence
Voloshinov presented his readers with what looks suspiciously like a
Transcendental Argument to demonstrate the existence of "theme".33
Ex hypothesi, that is all he could have offered anyway, since whatever evidence there might have been for
the existence of a particular "theme" must have (of necessity) arrived far too
late on the scene for it to be of much use. A split-second delay would be
far too
long to wait, if, as Voloshinov says, even microscopic changes alter "theme".
Hence, by the time any of this elusive 'evidence' became apparent, the alleged
"theme" would have changed, or would have disappeared. Naturally, this means that it would
be impossible for anyone to confirm this aspect of Voloshinov's theory.
Even film, video or recorded evidence would be of no use; these couldn't possibly preserve the
microscopic details surrounding the original utterance.34
Indeed, it is unclear whether it is possible for anyone to begin to form the
faintest idea of
what such confirmation might even look like.
Recall what we were told about "theme":
"Let us agree to call the
entity which becomes the object of a sign the theme of the sign. Each
fully fledged sign has its theme. And so every verbal performance has its theme."
[Ibid.]
Hence, the "object" of a
sign is intimately connected with the unique, occasional use of certain signs.
In that case, such an "object" plainly can't be identified, let alone
studied independently of singular events like these. Since these are in
principle unrepeatable they are
uncheckable, and if that is so, no scientific investigation would be
able to confirm this aspect of Voloshinov's theory. What could be measured, observed, or tested in such circumstances,
anyway? Even if there were anything to observe, how might test
results be confirmed if the "object" studied is irreduciblyuniqueandephemeral in the extreme?
This situation isn't at all like the experiments carried out in
High
Energy Physics, for
instance, where things happen extremely quickly, too. There such events are
reproducible since they aren't unique, and they aren't
occasion-sensitive. With "theme", this isn't the case. Hence, not only did
Voloshinov fail to provide any evidence to support his claims, none could have been
offered by him, or by anyone on his behalf, for that matter --, now or ever.
What then are we to make of claims like
these?
"Finally, for me there could be no question of
superimposing the laws of dialectics on nature but of discovering them in it and
developing them from it." [Engels (1976),
p.13. Bold emphasis
added.]
"The dialectic does not liberate the investigator from
painstaking study of the facts, quite the contrary: it requires it."
[Trotsky (1986), p.92. Bold emphasis added]
"Dialectics and materialism are the basic elements in the
Marxist cognition of the world. But this does not mean at all that they can be
applied to any sphere of knowledge, like an ever ready master key. Dialectics
cannot be imposed on facts; it has to be deduced from facts, from their
nature and development…." [Trotsky (1973), p.233.
Bold emphasis added.]
"'[The dialectic is not a] magic master key for
all questions.' The dialectic is not a calculator into which it is possible to
punch the problem and allow it to compute the solution. This would be an
idealist method. A materialist dialectic must grow from a patient,
empirical examination of the facts and not be imposed on them…."
[Rees (1998a), p.271. Bold emphases added.]
"A consistent materialism cannot proceed from
principles which are validated by appeal to abstract reason, intuition,
self-evidence or some other subjective or purely theoretical source. Idealisms
may do this. But the materialist philosophy has to be based upon evidence taken
from objective material sources and verified by demonstration in practice...."
[Novack (1965), p.17. Bold emphasis added.]
"Marxism, therefore, seeks to base our ideas
of things on nothing but the actual investigation of them, arising from and
tested by experience and practice. It does not invent a 'system' as previous
philosophers have done, and then try to make everything fit into it…."
[Cornforth (1976), pp.14-15. Bold emphases added.]
As we have seen (in Essay
Two), comrades who say such things, or who assent to them, quite happily do
the exact opposite in the very next breath, and readily impose their ideas on nature
and society, just
like Voloshinov.
Furthermore, even if there were some corroborating
evidence, it would surely have to be expressed in linguistic form, at some point. In that case,
it would itself be subject to the very same strictures applied to its own
"theme" and meaning, and so on ad infinitem. How would it be
possible to identify the "theme" of any sentence expressing or reporting this evidence, or
confirm whether or not it even had a "theme" if its own "theme" is
equally ephemeral and elusive?35
In addition, if the "microscopic" details
surrounding an "utterance" are essentially unique then within
nanoseconds of an "utterance" ending those involved in the conversation
would be at a loss themselves as to what its "theme" had been --, that
is, if it indeed
had one. [Or, even if they knew what they were looking for!] In fact, even as the sound waves carrying each utterance were
travelling between speaker and hearer the "microscopic" details surrounding the
original speech act would be altering, changing the "theme", or
perhaps even losing
it, forever. Worse still, during
vocalisation the "microscopic" details
proximate to each and every
nascent
speech act would be changing
diachronically. This means that,
while a speaker
was
speaking, the theme of what he or she was in the act of saying would be
altering -- unless, of course, we are to suppose that each "theme" is timed to
coincide with the beginning and end of each speech act.36
Indeed, unless "themes" were timed to begin or end miraculously like this, it would mean that each
utterance must possess an indefinite number of "unitary themes", depending on how
fast its originator spoke, how many micro-phonetic parts it contained, or
how often a speaker coughed, sneezed, or was interrupted during in a
conversation. Of course, anyone with a stammer would be doubly handicapped.36a
It could be objected that this is all
rather unfair since Voloshinov speaks of a "unitary theme" that
belongs to each
utterance, and he tells us that the "theme" of an utterance is "indivisible".
However, what these phrases actually mean will, of course, depend on how we
count utterances. Voloshinov appears to believe that "themes" and utterances
can be paired-off, one-one. In fact, something like this would have to
be the case, otherwise the ascription of a "unitary" and "indivisible" "theme" to an utterance would
be entirely empty. If so, it might look as if "themes" can be individuated
by the utterances they 'accompany'. But, that can't be right since "themes"
are circumstance-sensitive, which implies that any particular
type
utterance could in fact be the expression of countless different themes at
different times for each of its
tokens, as the "microscopic" details (etc.)
surrounding them varied. We saw as much above with respect to the sentences Holborow
herself discussed:
"Let us take an example…. [:]
I'm hungry conjures up a general concept. When, however, we look at
different contexts in which the phrase might be used, we see how the evaluative
accent changes everything. A child saying this to her mother might be
indirectly a request for the mother to get her something, an enquiry about what
there is to eat, or a statement that she just feels like something to eat.
One adult saying it to another might mean that it's time for lunch and be a
suggestion that they go somewhere to eat…. In each case the context is not
merely the gloss on the meaning but constitutes different meanings -- different
in every aspect…." [Holborow (1999), p.28.]
Here, each token utterance of the type "I'm
hungry" means something entirely new. If so, they must presumably have different
"themes". Indeed, as Voloshinov himself says:
"The theme of an utterance is
individual and unreproducible, just as the utterance itself is individual and
unreproducible. The theme is the expression of the concrete, historical
situation that engendered the utterance. The utterance 'What time is it?' has
a different meaning each time it is used, and hence, in accordance with our
terminology, has a different theme, depending on the concrete historical
situation ('historical' here in microscopic dimensions) during which it is
enunciated and of which, in essence, it is a part." [Voloshinov (1973), p.99.
Bold emphasis added.]
Since
we aren't allowed to consider the
meaning of type utterances (the idea is foreign to Voloshinov, it seems
-- but, meanings might be part of what he refers to as the "self-identical" aspects of an
utterance; p.100), Voloshinov's theory appears to indicate that token utterances of a certain
type are in fact pairable with
an indefinite number of "themes" according to circumstances.36a1
Conversely, the 'same' "theme" could be expressed by different type utterances.
Voloshinov doesn't explicitly rule this alternative out; even though it seems to be inconsistent with
some of the things he said, it is implied by other things he wrote, too. [On this, see Note 36a1, and
below.] So, the "theme" above (if it is one!), expressed by
the child's desire for her mother to get her some food, could in fact be
expressed in a number of different ways. The child could say any of the
following -- each expressing or instantiating the
same "theme":
P1: "Please get me some
food."
P2: "I'm
starving/famished/ravenous."
P3: "Is it nearly time for
dinner?"
P4: "I want a
biscuit/apple/burger/banana/pizza…"
P5: "I want to eat
something."
P6: "My tummy's rumbling."
P7: "My stomach thinks my
throat is cut."
And so on. In fact there are countless ways
this hypothetical child might express the 'very same' "theme".
These possibilities now raise serious
questions about how the pairing of "themes" and utterances is supposed to work.
More pointedly: Which "theme(s)" is/are to be paired with which utterance(s) if,
in theory, an utterance might represent a 'different' theme at different times,
and the 'same' "theme(s)" might be expressed by different utterances at same or
different times? Indeed, how are we to rule out the possibility that one
utterance could in fact express two or more "themes" at once (which seems
to be a viable option since two or more "themes" -- as allowed for by the theory
-- could be expressed by one and the same utterance on different occasions of
use)?36b
For example:
P8: "I'm hungry,"
could indicate that the one saying it wanted
feeding as well as expressing a veiled criticism of the one not doing the
feeding. In that case, it would have two "themes" --
if, of course,
this is what a "theme" is! Who can say?
Someone uttering P8
could thus mean (i.e., speaker's meaning), "Get me some food" as
well as
"I think you are a rather poor carer." Indeed, there might be other "themes"
'themed' by this one sentence on the same occasion (such as "I'm more important
than him/her, so feed me first", "You always treat me worse than
him/her", or "You are my employee, do as I say!", and so on).
To be sure, one or more of the above could in
principle be ruled out by a suitable definition or
stipulation.
It could then
simply be baldly asserted that each token utterance was paired one-one with
exactly one "theme". [How that could be checked is even
less clear!] Unfortunately, that would mean that whenever the 'same
utterance' was produced, the 'same' "theme" would have to be present, as a matter of definition.
That would, of course, make a mockery of the occasion-sensitivity of "theme"!
Either that, or it would mean that (despite appearances to the contrary) no
utterance was actually repeatable since each would be identified and
individuated by its own unique and unrepeatable "theme" -- which would
superglue each utterance to a unique set of circumstances. Indeed, this latest observation
seems to be consistent with some of the other things Voloshinov says:
"[T]heme must be unitary,
otherwise we would have no basis for talking about any one utterance. The theme
of an utterance is individual andunreproducible, just as the
utterance itself is individual and unreproducible. The theme is the
expression of the concrete, historical situation that engendered the utterance."
[Voloshinov (1973), p.99. Bold emphasis added.]
Unfortunately, he then went on to
say:
"The utterance 'What time is
it?' has a different meaning each time it is used, and hence, in accordance
with our terminology, has a different theme, depending on the concrete
historical situation ('historical' here in microscopic dimensions)
during which it is enunciated and of which, in essence, it is a part." [Ibid.,
p.99. Bold emphasis added.]
In that case, it seems reasonably clear that
Voloshinov wouldn't have pointed out that "The utterance 'What time is it?'
has a different meaning each time it is used…" (emphasis added), if the same utterance
hadn't in fact been used -- otherwise, the second "it" would dangle with no referent.
This indicates that, confused as he was, Voloshinov wanted to appeal to the
possible use of the same type utterance, tokened in new circumstances,
all the while clinging on to the idea that each utterance is totally unique!
With the best will in the world, it isn't easy to see how any of this is feasible --, nor is it
easy to figure a simple way out of this dialectical thicket. It is even less easy to see why
anyone (least of all the comrades mentioned above) would voluntarily propel
themselves right into centre of this impenetrable briar patch.
It could be argued that even though "What time
is it?" might be uttered countless times in many different circumstances, each
one is individuated by the occasion of its use, and hence by its "theme" (a
point in fact made a few paragraphs back!). Maybe so, but this simply commits
Voloshinov to a belief in type utterances, something he would have rejected (as
abstract objectivism). Anyway, we have already seen that each utterance could
have many different "themes", so "theme" itself seems incapable of individuating
anything, least of all an utterance. [We will return to this
below.]
On the other hand, it could be argued that
"theme" might be identified by the 'thought' conveyed by each utterance.
However, that option would itself risk becoming bogged down in a metaphysical
dispute over the precise nature of 'thoughts', and how they too might be
individuated! Anyway, we have already seen (in
Note 23) that Voloshinov had
himself blocked that escape route:
"In point of fact, the speech
act, or more accurately, its product -- the utterance, cannot under any
circumstances be considered an individual phenomenon in the precise meaning of
the word and cannot be explained in terms of the individual psychological or
psychophysiological conditions of the speaker. The utterance is a social
phenomenon." [Ibid., p.82. Italic emphasis in the original.]
Nevertheless, if we ignore this intractable problem for the moment, the
question would still remain: How do we individuate 'thoughts' except by
reference to the utterances they supposedly accompany, or instantiate?37
But, that just loops the discussion back in on itself. The whole point of the volunteered
response outlined in the previous paragraph seemed
to be aimed at trying to identify, or individuate, a "theme" by means of an
accompanying
'thought'. It now looks as if this can't be done without defining a 'thought' in
terms of utterances that are themselves supposedly identified by a "unitary theme".
But, that
in turn seems to mean that "themes" may be individuated only if they have
already been individuated!
Maybe we could pair-off 'thoughts' with type
utterances? Unfortunately, this would only serve to undermine "theme's" context-dependency, since the same utterance would implicate the same
'thought', and hence
the same "theme", and that would just loop the discussion back to where it was at
the end of the last paragraph.
Perhaps an appeal to meaning might help? But, again, if meaning
is parasitic on "theme", we are no further forward. Maybe the physical
properties of an utterance -- that is, the sound patterns associated with
specific sets of vibrating atoms or molecules -- could supply the principle of
individuation for "unitary themes"? Unfortunately, criteria of
individuation for sets of already identical atoms and molecules (distinguishable
only by an appeal to even more problematic spatial and temporal coordinates,
scalar energy and vector fields) aren't all that easy to construct (as
we have already seen). But, even if they were, this would still be of
little help. That is because those criteria would have to be
expressed in linguistic form, too, which would in turn attract the very same
difficulties that bedevilled the alleged "theme" they supposedly accompanied!
This doesn't look like a
very promising way out of this dialectical
dungeon.38
Furthermore -- and returning to an earlier theme
(no pun intended) --, if
utterances are to be individuated by means of circumstances, and the
latter are still microstate-sensitive, a finite set of words could conceivably
represent a potentially infinite (or indefinitely large finite) set of such
token utterances (since there seems to be no upper limit on the different
circumstances surrounding each utterance if any one of the latter is paired-off
with one of the former), all with their own "unitary themes". Hence,
and once more, the question, "What time is it?" could in fact mean
countless different things because of the indefinite number of surrounding circumstances that might accompany,
or occasion, each
of its exemplary utterances, all of which would presumably instantiate their own "themes". Naturally, this
would seem to imply that since the meaning of "What time is it?" isn't fixed by context-independent considerations
(according to Voloshinov), it could exemplify any number of such "unitary
themes", as the micro-details of each nascent utterance required -- including
those indicated or suggested during vocalisation, or those attendant upon that utterance while it was in
the process of being
registered in a hearer's 'consciousness', and so on.
Consequently, unless far more clearly defined
criteria are provided (by those sympathetic to Voloshinov's ideas) for counting,
distinguishing, or identifying utterances and "themes" (etc.), it seems
impossible to decide whether there are in fact countless "unitary themes"
pairable, one-one, one-many, many-one, or many-many, with utterances
(interpreted as identifiable spoken tokens, etc.) --, or whether there exist
more complex sets of functional relations between utterance tokens and "theme"
tokens, or between utterance types and "theme" tokens, and so on ad nauseam.
Having said that, it is worth pointing out
that the difficulties we face trying to comprehend what Voloshinov could possibly
have meant by what he wrote are largely
the result of the confused way in which he expressed himself. For example, on
the topic in hand (i.e., the individuation of "theme"), he had this to say:
"The theme must be unitary,
otherwise we would have no basis for talking about any one utterance. The theme
of an utterance is individual and unreproducible, just as the utterance itself
is individual and unreproducible. The theme is the expression of the concrete,
historical situation that engendered the utterance. The utterance 'What time is
it?' has a different meaning each time it is used, and hence, in accordance with
our terminology, has a different theme, depending on the concrete historical
situation…." [Ibid., p.99.]
From this,
as we saw earlier, it looks like Voloshinov thought
that an utterance could be individuated by its "theme":
"The theme must be unitary,
otherwise we would have no basis for talking about any one utterance…." [Ibid.]
On the other hand, he clearly thought that
"theme" was dependent on the concrete circumstances surrounding it:
"The utterance 'What time is
it?' has a different meaning each time it is used, and hence, in accordance with
our terminology, has a different theme, depending on the concrete historical
situation…." [Ibid.]
But, he also appears to have believed
that concrete circumstances were expressed by "theme":
"The theme is the expression
of the concrete, historical situation that engendered the utterance." [Ibid.]
In addition, it looks like Voloshinov thought
that not only was "theme" unreproducible, so were utterances:
"The theme of an utterance is
individual and unreproducible, just as the utterance itself is individual and
unreproducible." [Ibid.]
And yet, as noted above, he then spoke about
utterances being repeated:
"The utterance 'What time is
it?' has a different meaning each time it is used…." [Ibid. Emphasis
added.]
So, an utterance is and isn't
repeatable, hence its "theme" is and isn't unreproducible!
"The meaning of the utterance
'What time is it?' -- a meaning that, of course, remains the same in all
historical instances of its enunciation", [Ibid., p.100.]
it also looks like he believed that meaning
is fixed, after all, but only when the same utterance is produced,
something he had just said couldn't happen!
Unfortunately, the bemused reader will search in vain in the articles written by
the aforementioned comrades for any help in comprehendingwhat on earth
Voloshinov was banging on about!39
Again, it could be objected that the above
seriously misrepresents Voloshinov in that it ignores the clear distinction
he drew between meaning and "theme":
"A definite and unitary
meaning, a unitary significance, is a property belonging to any utterance as
a whole. Let us call the significance of a whole utterance its theme. The theme must be unitary,
otherwise we would have no basis for talking about any one utterance. The theme
of an utterance is individual and unreproducible, just as the utterance itself
is individual and unreproducible. The theme is the expression of the concrete,
historical situation that engendered the utterance. The utterance 'What time is
it?' has a different meaning each time it is used, and hence, in accordance with
our terminology, has a different theme, depending on the concrete historical
situation….
"It follows, then, that the
theme of an utterance is determined not only by the linguistic forms that
comprise it -- words, morphological and syntactic structures, sounds, and
intonation -- but also by extraverbal factors of the situation. Should we miss
these situational factors, we would be as little able to understand an utterance
as if we were to miss its most important words. The theme of an utterance is
concrete -- as concrete as the historical instant to which the utterance
belongs….
"Together with theme or,
rather, within the theme, there is also the meaning that belongs to an
utterance. By meaning, as distinguished from theme, we understand all those
aspects of the utterance that are reproducible and self-identical
in all instances of repetition. Of course, these aspects are abstract: they have
no concrete, autonomous existence in an artificially isolated form, but, at the
same time, they do constitute an essential and inseparable part of the
utterance. The theme of an utterance is, in essence, indivisible. The meaning of
an utterance, on the contrary, does break down into a set of meanings belonging
to each of the various linguistic elements of which the utterance consists. The
unreproducible theme of the utterance 'What time is it?' taken in its
indissoluble connection with the concrete historical situation, cannot be
divided into elements. The meaning of the utterance 'What time is it?' -- a
meaning that, of course, remains the same in all historical instances of its
enunciation -- is made up of the meanings of the words…that form the construction
of the utterance.
"…On the other hand, a theme
must base itself on some kind of fixity of meaning; otherwise it loses
its connection with what came before and what comes after -- i.e., it altogether
loses its significance…." [Ibid., pp.99-100. Bold emphasis alone added.]
From
this it could be argued that Voloshinov actually acknowledged many of the points
made above, and consequently they can't be used against him. Unfortunately, however, there are other things he said that undermine this 'sympathetic' interpretation of his intentions:
"A definite and unitary
meaning, a unitary significance, is a property belonging to any utterance as
a whole. Let us call the significance of a whole utterance its
theme….
"…The theme must be unitary,
otherwise we would have no basis for talking about any one utterance. The theme
of an utterance is individual and unreproducible, just as the utterance itself
is individual and unreproducible. The theme is the expression of the concrete,
historical situation that engendered the utterance. The utterance 'What time
is it?' has a different meaning each time it isused, and hence, in
accordance with our terminology, has a different theme, depending on the
concrete historical situation ('historical' here in microscopic dimensions)
during which it is enunciated and of which, in essence, it is a part….
"Together with theme or,
rather, within the theme, there is also the meaning that belongs to an
utterance. By meaning, as distinguished from theme, we understand all those
aspects of the utterance that are reproducible and self-identical
in all instances of repetition. Of course, these aspects are abstract: they
have no concrete, autonomous existence in an artificially isolated form,
but, at the same time, they do constitute an essential and inseparable part of
the utterance. The theme of an utterance is, in essence, indivisible. The
meaning of an utterance, on the contrary, does break down into a set of meanings
belonging to each of the various linguistic elements of which the utterance
consists….
"Theme is a complex,
dynamic system of signs that attempts to be adequate to a given instant of
generative process. Theme is reaction by the consciousness in its generative
process to the generative process of existence. Meaning is the technical
apparatus for the implementation of theme. Of course, no absolute,
mechanistic boundary can be drawn between theme and meaning. There is no theme
without meaning and no meaning without theme. Moreover, it is even impossible
to convey the meaning of a particular word…without having made it an element of
theme, i.e., without having constructed an 'example' utterance….
"Theme, as we have said, is
an attribute of a whole utterance only; it can belong to a separate word only
inasmuch as that word operates in the capacity of a whole utterance….
Meaning, on the other hand, belongs to an element or aggregate of elements in
their relation to the whole. Of course, if we entirely disregard this
relation to the whole (i.e., to the utterance), we shall entirely forfeit
meaning. That is the reason why a sharp boundary between theme and meaning
cannot be drawn.
"The most accurate way of
formulating the interrelationship between theme and meaning is in the following
terms. Theme is the upper, actual limit of linguistic significance; in
essence, only theme means something definite. Meaning is the lower limit
of linguistic significance. Meaning, in essence, means nothing; it only
possesses potentiality -- the possibility of having a meaning within a concretetheme…." [Ibid., pp.99-101. Bold emphases added; italic
emphases in the original.]
In this extract, while Voloshinov
distinguished "theme" from meaning, he also identified the two, saying that:
"A
definite and unitary meaning, a unitary significance,
is a property belonging to any utterance as a whole. Let us call the
significance of a whole utterance its theme…." [Ibid., p.99.
Bold emphases added.]
"Unitary meaning", "unitary significance" and
"theme" are one and the same here.40
To be sure, Voloshinov later acknowledged that words (etc.) possess fixed
meanings, but he had already torpedoed that idea by his prior
equation of meaning with "theme" -- since the latter isn't fixed. Moreover, he added
the following thoughts:
"…The utterance 'What time is
it?' has a different meaning each time it is used, and hence, in
accordance with our terminology, has a different theme….
"There
is no theme without meaning and no meaning without theme. Moreover, it is even
impossible to convey the meaning of a particular word…without having made it
an element of theme, i.e., without having constructed an 'example'
utterance….
"…Meaning, on the
other hand, belongs to an element or aggregate of elements intheir
relation to the whole….
“…Meaning, in essence,
means nothing; it only possesses potentiality -- the possibility of having a
meaning within a concretetheme….
"...Therefore, there is no
reason for saying that meaning belongs to a word assuch. In essence,
meaning belongs to a word in its position between speakers; that is,
meaning is realized only in the process of active, responsive understanding…."
[Ibid., pp.99-102. Bold emphases added.]
All
of these appear to make fixity of meaning a rather empty notion for Voloshinov
-- that is, if meaning is indeed occasion-specific and context-dependent, or if
it can change with each utterance (and can't be
ascertained apart from them), or, indeed, if it is speaker-relative, too.
It could be objected that this still
misrepresents Voloshinov in that he is quite clear that while there is no
clear boundary separating these two notions, at the extreme end of this continuum they are entirely different:
"The most accurate way of
formulating the interrelationship between theme and meaning is in the following
terms. Theme is the upper, actual limit of linguistic significance; in
essence, only theme means something definite. Meaning is the lower limit
of linguistic significance. Meaning, in essence, means nothing; it only
possesses potentiality -- the possibility of having a meaning within a concrete
theme…." [Ibid., p.101; italic
emphases in the original.]
But, according to this, without an
association with "theme", meaning "means nothing". Moreover, we are given no
clues as to how meaning can slowly appear along this alleged continuum. Is,
therefore, meaning like, say, the temperature of a metal bar as it is being heated from cold
to warm, and then to hot? But, what would be an example of a 'tepid' sort of meaning? An
utterance that was mumbled? Or, one that was cut-off in mid-stream? Or, one that
was uttered between two distinct surrounding circumstances or locations, on the run, as it
were? Indeed, what sense can be made of half a meaning, or 25% of one?
Hence, although Voloshinov does try to distinguish
"theme" and meaning, the other things he says identifies them, as pointed out
above. For example:
"A definite and unitary
meaning, a unitary significance, is a property belonging to any utterance as
a whole. Let us call the significance of a whole utterance its theme. The theme must be unitary,
otherwise we would have no basis for talking about any one utterance. The theme
of an utterance is individual and unreproducible, just as the utterance itself
is individual and unreproducible. The theme is the expression of the concrete,
historical situation that engendered the utterance. The utterance 'What time is
it?' has a different meaning each time it is used, and hence, in accordance with
our terminology, has a different theme, depending on the concrete historical
situation…." [Ibid, p.99. Bold emphases added; italic emphases in the
original.]
Of course, this quandary isn't helped by
the fact that we still haven't got the faintest idea what "theme" is!
The above might still be regarded as a little
unfair to Voloshinov, for he went on to connect "theme" with "understanding":
"The distinction between
theme and meaning acquires particular clarity in connection with the problem
of understanding….
"Any genuine kind of
understanding will be active and will constitute the germ of a response. Only
active understanding can grasp theme -- a generative process can be grasped only
with the aid of another generative process.
"To understand another
person's utterance means to orient oneself with respect to it, to find the
proper place for it in the corresponding context. For each word of the utterance
that we are in process of understanding, we, as it were, lay down a set of our
own answering words. The greater their number and weight, the deeper and more
substantial our understanding will be.
"Thus each of the
distinguishable significative elements of an utterance and the entire utterance
as a whole entity are translated in our minds into another, active and
responsive, context. Any true understanding is dialogic in nature….
Understanding strives to match the speaker's word with a counter word….
"Therefore,
there is no reason for saying that meaning belongs to a word as such. In
essence, meaning belongs to a word in its position between speakers; that is,
meaning is realized only in the process of active, responsive understanding.
Meaning does not reside in the word or in the soul of the speaker or in the soul
of the listener. Meaning is the effect of interaction between speaker and
listener produced via the material of a particular sound complex. It is like
an electric spark that occurs only when two different terminals are hooked
together. Those who ignore theme (which is accessible only to active, responsive
understanding) and who, in attempting to define the meaning of a word, approach
its lower, stable, self-identical limit, want, in effect, to turn on a light
bulb after having switched off the current…." [Ibid., pp.102-03. Bold
emphasis added; italic emphasis in the original.]41
Here, meaning is no longer linguistic (i.e.,
it no longer belongs to the use of a word), it is essentially psychological,
and is now a
feature of the interaction between at least two 'minds'. In that case, meaning
isn't:
"the lower limit
of linguistic significance. Meaning, in essence, means nothing; it only
possesses potentiality -- the possibility of having a meaning within a concrete theme…." [Ibid., p.101.]
Nor is it what we were earlier led to believe:
"The meaning of an
utterance, on the contrary, does break down into a set of meanings belonging to
each of the various linguistic elements of which the utterance consists. The
unreproducible theme of the utterance 'What time is it?' taken in its
indissoluble connection with the concrete historical situation, cannot be
divided into elements. The meaning of the utterance 'What time is it?' -- a
meaning that, of course, remains the same in all historical instances of its
enunciation -- is made up of the meanings of the words…that form the
construction of the utterance." [Ibid., p.100. Bold emphases added.]
Meaning has now become:
"the effect of interaction between speaker and
listener produced via the material of a particular sound complex. It is like
an electric spark that occurs only when two different terminals are hooked
together. Those who ignore theme (which is accessible only to active, responsive
understanding) and who, in attempting to define the meaning of a word, approach
its lower, stable, self-identical limit, want, in effect, to turn on a light
bulb after having switched off the current…." [Ibid., p.102.]
[As we will
see later, Voloshinov has
clearly run-together several different meanings of "meaning"! In fact, as Note 23shows, if Voloshinov were
correct, inter-subjective understanding would be impossible.]
Nevertheless, even if we ignore these serious
difficulties for now, the above passages still can't help
us in our understanding of Voloshinov's theory, since we are now entirely
unclear about both "theme" and meaning!
Moreover, if
understanding
were in fact
dependent on translation, that would compound the problems facing
Voloshinov's theory even further. That is because a listener would have no way of knowing whether his/her
translated words accurately represented the "theme" that his/her
interlocutor had (in fact?) intended, or had associated with their own words
when they were uttered, or, indeed, were associated with what they had said by
the 'concrete circumstances' of their utterance.
Instead of having merely to understand a speaker, a hearer would now have
to unravel an intrinsically inaccessible and un-reproducible "theme" before
understanding could even begin!
Worse still, the "theme" associated
with an utterance (according to the
'definition' we were given) is totally unique; it can't have been
experienced by that individual, or by anyone else, for that matter, in all of
human history -- ever. How then could anyone use this totally unique
"theme" to assist in the understanding of someone else's words? Naturally, this
means that far from assisting linguists or psychologists find a solution to
the 'problem' of understanding, the introduction of this radically obscure
notion ("theme") isn't just a
hindrance, it presents them with an completely insurmountable obstacle,
the equivalent of throwing an anvil at a drowning man.
Furthermore, if all understanding involves
translation, then speakers themselves would fail to understand even their
own words. As seems reasonably obvious, if translation is to be successful, it must represent
that which is
to be translated in a medium that is already understood. But, if this
prior understanding itself requires still further translation (which it must do if,
given
this theory, all understanding requires translation), then that just
introduces yet another infinite regress, with translation upon translation
stacking-up in order to facilitate each episode of "understanding".
Translation has to take
place in some language or other, which according to Voloshinov must itself be
"theme"-dominated, and hence occasion-sensitive. That being
the case, these translations would
now depend on resources which are themselves intrinsically inaccessible, and
would relate to speech-acts and surrounding circumstances that are themselves sensitive to
changes that take place every microsecond (according
to Voloshinov). Moreover, since no speaker has
access to any of their own past "themes" (or any at all!) to assist him or her
in this endless "theme"-hunt, they would be searching for the terminally ephemeral by means
of the psychologically unattainable. This is the equivalent of looking for a
needle in a haystack when (i) you have never seen a needle before, (ii) no one
else has either, and (iii) no one knows what a haystack is!
Consequently, on this account, speakers would
fail
even to understand themselves!
In that case, the following would be
impossible:
"Thus each of the
distinguishable significative elements of an utterance and the entire utterance
as a whole entity are translated in our minds into another, active and
responsive, context. Any true understanding is dialogic in nature….
Understanding strives to match the speaker's word with a counter word…."
[Ibid., p.102.]
If such "counter words" have any meaning, they
must also have a "theme", and if that is so, hearers will fail to
understand these words, and they will do so for the above reasons. Furthermore,
even if the "theme" of an utterance were
in fact accessible to speaker and hearer alike, only a minor miracle would make the
"theme" of a listener's "counter word" coincide with the "theme" of the original
utterance. Either way, and once again, understanding and communication would fail.
The usual response to this line of argument runs something like this:
E1: Well, we never really
understand one another, do we?
But, if that were the case, we must fail to
understand E1, too. And, if that is so, any response that depends on E1 must
also fail, for we wouldn't know what we were being informed of by means of it!
This is quite apart
from the incongruity of being presented with a theory of understanding that ends up
denying there isn't any of it to be had!
[That is, if anyone actually understands
this theory to begin with -- which, if E1 were true, they wouldn't, would they?]41a0
Conversely, if understanding a particular
translation required no further acts of understanding, or of translation
(that is, if the regress outlined above were short-circuited somehow), it would
then be pertinent to ask why translation was introduced to account for
understanding in the first place. If at some unspecified point we succeed in comprehending our own
'translated' words un-mediated, as it were, by any further
acts of translation (as this short-circuited variant would have it), why can't
we do this directly with the words others send our way? Why can't we just
understand them? Not only would this prevent the above regress (and the
subsequent attempt to block it on an ad
hoc basis), it would conform with what we already mean by
"understanding" (as will be demonstrated below).
The only conceivable reason for accepting the
sort of psychologistic detour we met above (which appeals to what appear to be
occult acts of 'inner translation' to account for understanding and the use of
language) would seem to
be that comprehension is a private mental process that we accomplish directly
by means of "inner speech", or some such.
Now, we don't have to appeal to the
definitive case mounted by Wittgenstein against
the possibility of there being a "private language"
to oppose this approach to 'cognition'; Voloshinov's own precepts rule it out,
for he declares that:
"Meaning does not reside
in the word or in the soul of the speaker or in the soul of the listener.
Meaning is the effect of interaction between speaker and listener produced
via the material of a particular sound complex." [Ibid., pp.102-03. Bold
emphasis alone added.]
In that case, it is difficult to see why Voloshinov
(or, indeed, any of his epigones) found he had to appeal to translation
in order to account for
our ability to understand one another, when, given his own theory, it drops out of
the picture. If anything, our understanding accounts for
translation, not the other way round -- otherwise, as noted above, the
individual concerned wouldn't know whether or not she/he had translated a given "sound
complex" correctly.
Of
course, if could be argued that "correct" has no place here, since meaning is a
negotiation between speakers. But, if that were so, there would be no
distinction between translation and mistranslation, which would, naturally
undermine the possibility of translation itself, just as calculation would be
undermined if there were no distinction between calculation and miscalculation.
Be
this as it may, an earlier allegation that there is a tension in Voloshinov's
work -- whereby, on the one hand, he wants to see language as a social product
or phenomenon, while, on the other, his ideas about "understanding" suggest that
he has fallen prey to the traditional theory that language and our capacity to
understand what is said depends on 'inner acts of intellection' -- now seems
correct.41a
Here is what Voloshinov says, for example, about
"understanding":
"Idealism and
psychologism alike overlook the fact that
understanding itself can come about only within some kind of semiotic material
(e.g., inner speech), that sign bears upon sign,
that consciousness itself can arise and become a viable fact only in the
material embodiment of signs...understanding is a response to a sign with
signs." [Ibid., p.11. Italic emphasis in the original.]
"In the first instance, to
understand means to refer a particular inner sign to a unity consisting of
other inner signs, to perceive it in the context of a particular psyche....
"Self-observation
(introspection) is the understanding of one's own inner sign.... We do not see or
feel an experience we understand it. This means that in the process of
introspection we engage our experience into a context made up of other signs we
understand. A sign can only be illuminated with the help of another sign."
[Ibid., pp.35-36. Italic emphases in the original.]
This is unfortunate since, if the above were the case, human beings
couldn't evenbegin to
"understand" anything. That is because we aren't born with 'signs' in our heads (or in our
'consciousness') -- unless we assume that a baby has a set of 'innate' signs in her/his
'psyche'. Hence, if acts of understanding were indeed a function of the relation between
signs, as Voloshinov says, they couldn't take place. After all, "a sign can
only be illuminated with the help of another sign", so, if we have none to begin
with, the process of "illumination" can't even begin. [On this, see
Note 23.]
Despite this, it is rather
odd to say that our heads are full of "signs" --, or, perhaps that
"consciousness itself can arise and become a viable fact only in the material
embodiment of signs" -- since Voloshinov isn't too clear what he means by "sign". Hence, not much can be
done with this peculiar
idea of his. [However, I will return to this topic later.]
Anyway, the above comments at least
pin Voloshinov's flag to the traditionalist mast: understanding for him is (in
the "first instance") an inner, private affair. Despite his other gestures to the
contrary, he has clearly failed to break decisively with Platonic, Christian, and
Cartesian Mythology. Plainly,
this is
one "ruling idea" that has landed, and set up home in, yet another
radical "psyche".
[Alas, other comrades seem to have caught the same bug.]
One response to the above might run along the
lines that
hearers have to (in Voloshinov's words) "orient" themselves toward a
speaker's utterance:
"To understand another
person's utterance means to orient oneself with respect to it, to find the
proper place for it in the corresponding context. For each word of the utterance
that we are in process of understanding, we, as it were, lay down a set of our
own answering words. The greater their number and weight, the deeper and more
substantial our understanding will be." [Ibid., p.102. Bold emphasis added.]
Clearly, this must involve the translation of the latter's words
into the listener's own
idiolect -- or perhaps into their own "inner speech":
"Idealism and
psychologism alike overlook the fact that
understanding itself can come about only within some kind of semiotic material
(e.g., inner speech), that sign bears upon sign,
that consciousness itself can arise and become a viable fact only in the
material embodiment of signs...understanding is a response to a sign with
signs." [Ibid., p.11. Italic emphasis in the original.]
Quite apart from the fact that Voloshinov offered no empirical evidence to
substantiate this bizarre idea (that we accompany the speech of others with
strings of our own words -- i.e., "inner speech", or "counter words" -- in order to comprehend
our interlocutors), any parallel dialogue like this would actually get in the way
of our attending to what was being said. It would
be rather like having to put up with an irritating 'inner
i-Pod' -- which
couldn't be ignored, turned down or switched off while we struggled to listen to what
was being said to us.41b
Even if such an 'inner running-commentary' actually
took place, it still wouldn't explain how we succeed in understanding anything
said to us, for it would clearly fail to account for our immediate
comprehension of the words (i.e., the "signs" that appear in "inner speech")
that these 'inner i-Pods' themselves constantly pump into
our 'inner ears'. If all understanding requires
such "inner speech", the constant din of this inner nuisance would surely
have to be accompanied by an even 'inner inner i-Pod', ('inner, inner
speech') if it, too, is to be
comprehended, and so on, ad infinitem.
On the other hand, if we
directly understand our own individual 'inner i-Pods' (this "inner speech") without recourse to any
further such devices (that is, if this infinite regress is terminated at the
first stage), then what reason could there be for not stopping
it one stage earlier still? Why may we not understand each other's words directly and
dispense with these spooky 'inner voices'/'inner signs'? If we understand "inner speech"
directly, then why not 'outer' speech?
[As we saw above, the only reason for
supposing that we can't understand 'outer speech' directly without the
intercession of these 'inner voices' -- the latter of which we seem to be able
to
comprehend directly without any further intercession by an 'inner, inner voice'
-- is that, for Voloshinov, understanding is an individualised, secret,
'inner process'. On this, see the
next section.]
Furthermore, the mere correlation
of two parallel streams of language (wherein an 'inner' dialogue supposedly
accompanies its 'outer' correlate, as one or both are processed in the Central
Nervous System [CNS], one presumes) doesn't establish that one of them is the
translation of the other, any more than talking aloud in English while a Russian
film is on TV counts as translating it. And this remains the case even if the one doing the
talking actually understands Russian. Hence, even if we could comprehend the
nature of "inner
speech", it wouldn't establish
that a successful translation had been accomplished by means of it.
It
would count as a translation of the former only if the words they contained had the same meaning (and
presumably the same "theme") as those which they sought to translate, but according to this theory,
no two utterances can have the same "theme"
(and thus not even the same meaning), so these annoying 'inner voices' would
be no use at all (even if they existed!). Just like the antics of an incompetent
translator, this ghostly charade would get no translation right, since there is,
on this view, no such thing to be had!
The
sensible theorist, therefore, will switch this annoying device off -- or,
perhaps better still, question its existence to begin with.42
As noted earlier, the only apparent reason for rejecting
the above objections would have to be based on the belief
that "inner speech" is immediate to 'consciousness', and is therefore instantly
comprehensible -- simply because it is "inner". This view in turn
trades on the further idea that when something is inside our heads (or is part of the CNS, perhaps?), a sort of
internal, ghostly viewer or listener takes over and does the
translating and the understanding (directly and without further
translation), for us. There is no other way to make sense of the metaphors
Voloshinov and others use here.
If so, intimate,
internal proximity seems to be the
factor that renders such speech automatically comprehensible. In contrast, speech that is 'outer'
somehow prevents, or, at least, fails to facilitate, understanding (in the "first
instance -- p.35.).43
In that case, it looks like the mere fact that such speech is inner means
that it is capable of being grasped directly without the need for another
even more inner, inner 'meta-translating' device to act as the next intercessor
in the chain. But, as noted above, if "inner speech" is indeed speech, presumably it
too must be occasion-sensitive. And yet, if that is so, the elusive "theme"
associated with each inner representation of the utterances encountered in 'outer speech'
will be
even more inaccessible than the "theme" allegedly associated with its
intended outer correlate. And, as noted above, short of a minor miracle, there is no way
these two speech episodes could have identical "themes", given the strictures
Voloshinov placed on "theme".
If, on the other hand, "inner speech"
isn't occasion-sensitive (and thus has no "theme") then we are owed an
explanation as to why it should be called "speech" in the first place, and why
(in the second) this use of "signs" is exempt from -- while their outer
correlates are still subject to -- occasion-sensitivity. Indeed, if "inner
speech" isn't
itself
occasion-sensitive, then how could it help translate "theme" accurately if the
"theme" of 'outer speech' is occasion-sensitive? And, if
"inner speech"
has no "theme" then how might it be understood? Hw might it have any
meaning?43a
"There is no theme without meaning and no meaning without theme. Moreover, it is even
impossible to convey the meaning of a particular word…without having made it
an element of theme, i.e., without having constructed an 'example'
utterance…." [Ibid., p.100. Bold emphasis added.]
Indeed, "inner speech" can't be understood if it has
no "theme". This isn't just my view, it is Voloshinov's:
"…The theme must be unitary,
otherwise we would have no basis for talking about any one utterance. The theme
of an utterance is individual and unreproducible, just as the utterance itself
is individual and unreproducible. The theme is the expression of the concrete,
historical situation that engendered the utterance. The utterance 'What time is
it?' has a different meaning each time it is used, and hence, in accordance with
our terminology, has a different theme, depending on the concrete historical
situation ('historical' here in microscopic dimensions) during which it is
enunciated and of which, in essence, it is a part.
"It follows, then, that the
theme of an utterance is determined not only by the linguistic forms that
comprise it -- words, morphological and syntactic structures, sounds, and
intonation -- but also by extraverbal factors of the situation. Should we miss
these situational factors, we would be as little able to understand an utterance
as if we were to miss its most important words. The theme of an utterance is
concrete -- as concrete as the historical instant to which the utterance belongs.
Only an utterance taken in its full, concrete scope as an historical
phenomenon possesses a theme. That is what is meant by the theme of an
utterance." [Voloshinov (1973), pp.99-100. Bold emphases alone added.]
And yet, Voloshinov situates understanding
in
the psyche, which means the "signs" comprising "inner speech" must have a
"theme" of their own:
"In the first instance,
to
understand means to refer a particular inner sign to a unity consisting of
other inner signs, to perceive it in the context of a particular psyche....
"Now in what form do we
receive the psyche, receive inner signs, for observation and study? In its pure
form, the inner sign, i.e., experience, is receivable only by self-observation
(introspection)....
"The fact is, after all, that
inner sign is the object of introspection and inner sign, as such, can also be
outer sign. Inner speech could indeed be given voice....
"Self-observation
(introspection) is the understanding of one's own inner sign.... We do not see or
feel an experience we understand it. This means that in the process of
introspection we engage our experience into a context made up of other signs we
understand. A sign can only be illuminated with the help of another sign."
[Ibid., p.36. Italic emphases in the original.]
But, as noted above, if the "signs" comprising "inner
speech" attract a "theme" of their own, understanding must fail:
Hence, even if we could comprehend the nature of "inner
speech", it wouldn't establish
that a successful translation had been accomplished by means of it. It
would count as a translation of the former only if the words they contained had the same meaning (and
presumably the same "theme") as those which they sought to translate, but according to this theory,
no two utterances can have the same "theme"
(and thus not even the same meaning), so these annoying 'inner voices' would
be no use at all (even if they existed!). Just like the antics of an incompetent
translator, this ghostly charade would get no translation right, since there is,
on this view, no such thing to be had!
[This passage has been quoted from
here.]
The traditional
account (i.e., one that holds that 'thinking' (etc.) takes place 'inside the
head') is in fact derived from the mystical idea that
'consciousness'/the 'soul' can be likened to
a sort of internal viewer of, or
listener to, the contents of the 'mind'/brain -- somewhat similar to the way a
cinemagoer watches a film in an auditorium, only this is supposed to be far more intimate. This metaphor
implies that 'consciousness' operates like a sort of linguistically-challenged,
sub-, or quasi-human
'entity', a social atom located somewhere in the cerebral, psychological or verbal universe. In Voloshinov's work,
as we have just seen, this 'inner auditorium' re-surfaces as the "psyche",
which is a sort of semi-passive,
mute 'inner couch potato', whose only job, it seems, is juggling with, and comparing,
whatever "signs" manage to drift its way.
It is "semi-passive" since
it looks like this ghostly
head-lodger isn't
permitted to translate "inner speech" into speech that is even more inner
so that it can understand the original "inner speech", in order to
forestall the infinite regress alluded to above. This ethereal, internal
individual certainly doesn't seem to engage in any practical activity; no one imagines it jogs about inside
the skull, finds employment in a
mitochondrial power plant, or agitates neurons into working-to-rule. It
certainly enjoys no social connections of any sort.
Alternatively, this trope might suggest that
whether or not inner 'consciousness' possesses its own 'inner, inner
language', it need never use it because plain and simple "inner speech" can be
understood directly with no need of further acts of intercession. In that case, this 'inner spectator' would be a sort
of taciturn but highly intuitive (if not magically gifted) 'inner couch
potato', since it wouldn't need to translate "inner speech" into something even more
'inner' in order to comprehend it.
Indeed, it would appear to be an 'inner
projection' of ourselves, just as 'god' is an 'outer projection'.
Small wonder, then, since this view originated in Christian and Platonic
Mysticism, and was given its modern shape by Descartes. Voloshinov's "psyche" is
clearly the correlate of the Cartesian 'soul'
One or other of these alternatives would have to be the
case if translation is to stop at some point, and this semi-passive
'cranium squatter' is to 'understand' things directly (and then 'explain' them
somehow to us -- perhaps we have 'inner ears', too?) without the need for still
more 'inner, inner, inner...' intercessors.43b
On either account, the connection between the
use of language and understanding has been severed -- which result seems to be
contrary to Voloshinov's own stated aims. That is because, even given this approach,
language drops out of the picture, since, at some point, "translation" must be effected in a non-linguistic
form or medium, 'intuitively', as it were. Understanding has in the end to be divorced from
the use of language to avoid the infinite regress of ever 'inner, inner couch
potatoes', required by this theory in order to facilitate the entire process. Comprehension thus becomes a non-linguistic, sui generis,
feature of our private 'mental' lives. But, if comprehension works like this (i.e., if
it is in the end 'inner', direct and immediate), then the motivation to provide an
explanation for it by postulating such 'inner processes' vanishes. If we all
understand one another in such a direct way at some point, why postulate
the need for "inner speech" to assist us? Other than serving to confuse
the easily confused, what possible role can it play?
So, even
on this account, "inner speech" does no work; at some point we all seem
just to understand
one another.43c
[Any who
reckon this misrepresents Voloshinov
should consult Note 23, and then
perhaps think again.]
Once more, if 'inner understanding' is itself
sui generis and spontaneous,
needing no further acts of intercession, why can't everyday 'outer' understanding work
in the same way? What possible reason could there be for an internal device of
this sort to provide an inner sanctum where language is finally processed.
Indeed, what is the point if, in the end, we end up with an explanation of
understanding that simply reduplicates the 'problems' associated with whatever
it supposedly replaced, and which mystifies the
phenomenon into the bargain (by locating it in a hidden and inaccessible realm)? What is gained by an appeal to an 'inner' process
that works just like the outer one for which it was supposed to provide some
sort of explanation? If the immediate understanding of one human being by
another is indeed a 'problem' (which requires a philosophical
and/or scientific
'solution'), why is the reduplication of that very same 'problem' in
an occult, 'inner' sanctum deemed a significant advance? If in the end
understanding is something we just do (if it is a basic fact about all of
us), then why do we need to burrow away inside our heads to find a more basic
process that merely reproduces the very thing that needed 'explaining' in the
first place: the intelligent use of language by humans who in the end typically understand
one another
immediately?43d
At this point, and as noted in several other Essays (for example,
here and
here), the atomistic
nature of the traditional theory (that understanding, etc, takes place in our
heads) should be obvious for all to see, for the 'explanatory' core of this
approach to language presents us with what looks suspiciously like an isolated
individual -- beloved of bourgeois ideology -- lodged inside each head.
This oracular, cranial squatter -- who differs from the
Cartesian 'soul' in name
alone -- is, on this account (and not surprisingly), far removed from the
affairs of communal life. Such a speechless atom would have no need of a public
language -- nor would it require socialisation. Its 'discourse' (if such it may
be called) can't in fact be social, it is manifestly 'inner' and private.
Nevertheless,
private property in the means of speech production sits rather awkwardly
with what is supposed to be a Marxist account of language.
If we continue the above
theme (no pun intended), we encounter another, related, problem: even if the 'representational'
view of language were correct, how could language actually represent things to
this 'inner spectator'? Voloshinov talks
as if "signs" are able to do this all on
their own, perhaps as we internally compare
them
(which signs seem to carry their meaning on their faces, as it were). But, who
views these inner "signs" is left a complete mystery -- unless, of course, we
postulate an 'inner eye', or an 'inner observer', to fit the bill.
Anyway, how can sounds or words communicate anything to a mere
viewer of pictures (or, indeed, a hearer of sounds)? Surely, they could only do this
if this 'inner watcher' was already a language user, and possessed
'inner, inner eyes' or 'inner, inner ears' of her/his own -- along with an 'inner'
social life, whereby these skills were first acquired.44
If our 'outer' social life and our 'outer' eyes and ears aren't enough, then how can these 'inner' sense
organs take up the slack? In what way are they superior to the skills we have
all acquired as part of our socialisation?
In
fact, and to the contrary, an 'inner spectator' like this is nothing
more than a little man/woman "in the head", with no family, friends or
acquaintances -- entirely 'self-socialised' and 'self-educated'.45
Naturally, the metaphor used earlier (i.e.,
that cinema-going head-lodger) itself suggested this 'inner spectator' interpretation,
but even if this analogy were inapt, how else are we to make sense of these "inner
representations" to 'consciousness'? What is the point of using the word
"represent" (that is, if we interpret Voloshinov's
own words in
this way) if there is no
one to whom anything is represented? If this word means what we
ordinarily take it to mean (that is, if we don't misrepresent its meaning, or fail
to
regard it as the
transitive verb
that
it is, and which requires an object -- without an object, a viewer of
these 'representations', they would be like pictures hung by a robot in a
gallery that no one has ever visited, or can visit), then this account clearly depends on yet another homunculus theory of the mind.45a
Here is Voloshinov's metaphor (which suggests
he accepted his own version of this 'theory'):
"Individual consciousness is
not the architect of the ideological superstructure, but only a tenant lodging
in the social edifice of ideological signs."
[Voloshinov (1973), pp.12-13. Bold emphasis added.]
So,
according to Voloshinov, instead of having to endure an
interminable i-Pod, with no 'Off' button,
inside our skulls, we all seem to have an invisible internal friend who
sifts through the myriad of sensory inputs the
CNS sends
his/her/its
way, all of which are then 'represented' to this 'friend' so they can be communicated
somehow to
each of us. This inner invisible companion must, of course, explain everything to us --,
presumably by 'whispering' in our 'inner ears', making use of inner "inner speech"
-- since we seem incapable of understanding anything without him/her/it
intercessing on our behalf.46
Naturally, this means that there would have to be at least two of these 'cerebral
squatters' inside each skull: one to do the explaining and one the
listening. Worse still, each of these homunculi would themselves have to have
similar, but smaller 'friends' in their minds/brains/heads to 'whom' things are
likewise 'represented', and so on... We might then wonder how we ever manage to hear
anything above the din, as this potentially infinite body of jabbering
Russian Dolls went about their cacophonous daily business.47
Figure One: The Human
Psyche?
In July 2015,
Disney
Pixar released an animated, children's film,
Inside Out (and in 2024, Inside Out 2), which inadvertently exposed the absurdity of the 'Homunculus
Theory' of 'the mind':
Videos One, Two, And Three: Disney's
'Little People In The Head'
On the other hand, if understanding is made manifest by our
competent use of language (alongside associated skills and performances) in a public
domain, then an appeal to the intercession of "inner
speech" to facilitate it, is unnecessary. Indeed, we don't need to anthropomorphise
the brain/'mind'/CNS in this way in order to account for our ability to
comprehend one another --, since, of course, there is nothing here that needs accounting
for.48
The
contrary supposition (i.e., that "inner speech" is essential to understanding) is clearly motivated by a
powerful set of ideological illusions, chief among which is the belief that
unless something is internalised it can't be understood. This by-now-familiar representational view of
language and thought is itself based on the idea that it is mere proximity and
internal immediacy
that renders "inner speech" directly comprehensible to 'consciousness'.
That is, it is the inner manipulation of signs
and/or symbols (or their physical or psychical correlates) that constitutes
understanding, as opposed to 'outer' communication, behavioural competence and
social interaction that does. [On
this, see below.]
It is also plain that the traditional
picture is itself motivated by yet another set of inappropriate
nominalisations
and
reifications
of everyday words -- terms that ordinarily express or exhibit our intellectual
and
linguistic skills, dispositions and states --, a wrong turn that is compounded by their consequent
fetishisation.48a
This traditional approach runs along
the following (highly truncated) lines: if 'consciousness', 'language' and
'the understanding' are in fact objects or inner processes (and who can
possibly doubt that if they have been given names?), or if they are based on these
inner objects and processes, a successful theory (especially if it hopes to be
'scientific' and 'philosophical') must account for their
inter-relationship.
However, these 'inner entities' have
been conjured into existence by the simple expedient of 'naming' them --
which plainly divides and then separates one from another by objectifying, or
reifying, them.
Because of such moves these separated 'items' now
require a 'theory' to re-connect them! Enter Traditional Philosophy and
contemporary Cognitive 'Science'.49
But, this is an attempt to find a 'solution' to a bogus problem. Bogus, because
the original distinction between these 'internal objects and processes' was motivated by these inappropriate linguistic moves, and nothing more.
Attempt because it is impossible to complete the task this pseudo-problem
presents those who invented it, or who now try to wrestle with it, since these entities (i.e.,
'consciousness', 'language' and 'the understanding', etc.) are
figments of the imagination, motivated by the
reification and fetishisation of a handful of concepts.50
As any competent user of the language may
readily confirm, this isn't how we already use words like "understand",
"think" and "to be aware"; we don't employ them to name inner objects and
processes. This is revealed by the further fact that we ordinarily decide, for instance,
whether someone has understood what is said to them by an appeal to outer criteria.
We don't examine the contents of their heads, or try to access their mental
imagery. If this is what we mean by "understanding" (that is, if we apply this
word successfully on the basis of outer criteria like this, which cri8teria are associated
with publicly checkable performances, skills and achievements (as opposed to hidden and
mysterious inner 'events'),
then the employment of this word to depict what goes on inside our heads
will be seen for what it is -- the
Platonic-Christian-Cartesian Paradigm in
all but name.
Naturally, this last set of bald assertions needs some defending --
but, fortunately, no much.
Undeniably, language has developed and grown
as result of the material interaction between human beings and the world.
Manifestly, this didn't
take place as a result of the occult deliberations of an obscure, inner ethereal
entity (i.e., "consciousness", or "thought") beloved of tradition.
That
observation isn't just consonant with a Marxist view of the social nature of language
and human beings, it agrees with everyday
linguistic and social practice. When studying the social and intellectual development
of humanity, for example,
archaeologists and historians would make no progress at all if they attempted to
consider the machinations of these mythical inner objects and processes.51
What they do (what we all do), of course, is examine the conditions under
which our ancestors lived -- the social and political forms they assumed --,
their struggles, writings, inter-relationships, means of production,
relations of exploitation, etc., etc. In addition to this, the study of artefacts, inscriptions, buildings,
coffins, possessions, property relations, class structures, and so on, would add
detail, where necessary. This is
what constitutes an HM study of the past (and of the present, for that matter). If language is
intimately connected with humanity's social development, then a materialist account of
discourse and comprehension need
take no heed of these hidden, 'inner objects and processes', even if sense
could be made of them.
'Inner processes' like these aren't hidden from us because they are
especially well-concealed, difficult to locate or
inspect;
there is in fact nothing there to study -- or, rather, it makes no
sense to suppose there is -- and this is so for reasons given above (which are further elaborated upon below).
The contrary supposition that there are such
occult (i.e., hidden) goings-on is often motivated by yet another inappropriate use of language,
itself a result of the influence of an archaic tradition, the aforementioned Platonic-Christian-Cartesian
Paradigm -- and nothing more. Apart from a crass misuse of words,
allied with
this mystical tradition, there is nothing to suggest that
such 'inner processes' exist. Indeed, that is why it was asserted
above that these mysterious 'inner objects and processes' are immaterial (in both
senses of that word); they couldn't feature in a materialist account of
anything since they don't exist (or rather, once again, no sense can be
made of the supposition that they do). In our practice we take no heed of them; our
material use of language and our shared behaviour show that such
'objects and processes' are chimerical.51a
The
social nature of language implies that individuals aren't free to attach their
own private meaning to words so that they become the meaning of those
words -- least of all a meaning that runs counter to the open, public application of terms like "understand",
"thought", and
"to be aware". This is partly because whatever personal gloss might be put on
any such words -- as is the case with other social products, such as commodities --, their meaning
or 'value' is fixed
by outer, not 'inner', material conditions. [This topic will be examined in more detail
presently.]
Hence, despite his disclaimers, Voloshinov's
theory not only depends on just such a
reification of language, it relies on an
anthropomorphisation of the mind or brain. That is, it depends on a
inner projection of outer social categories
onto the aforementioned fictional, 'inner couch potato'
-- i.e., onto what is, in all but name, theCartesian Soul.
These seemingly dogmatic assertions will now
be defended.
To summarise:
In connection with Voloshinov's
claims about translation and "orienting" ourselves to another's speech, it is worth noting that unless
listeners could confirm that they had translated their interlocutor's words into
their 'own language' correctly they would be in no position to say
whether or not they had successfully "oriented" themselves toward that speaker,
to begin with. But, how
could they do that without alreadyhavingunderstood
what was said? Otherwise, any translation is going to seem right -- in which
case we can't talk about "right" (to paraphrase Wittgenstein).51b
Hence, the 'theory' of understanding being
examined here implies that there must be a correct pre-translation
of a speaker's words into the "inner speech" of his/her hearers if
they are to "orient" themselves to that speaker correctly. If so, the claim that speakers have to "orientate" themselves to one
another if they are to understand what is said is the reverse of the truth. They would in fact have to understand the words spoken to them before
orientation could even begin (otherwise, on this 'theory' the supposed
translation would itself be incomprehensible). In that case, the appeal to
translation and orientation to account for understanding is an empty gesture,
since it, too, would require the pre-existence of the very thing they had been
introduced to explain -- i.e., the inter-subjective understanding of language.
Once again we see that
the idea that understanding is a mysterious 'inner process' in need of
scientific 'explanation' underlies this traditional
approach to language, and because of that our capacity to understand one another
is turned into a 'philosophical problem'. But, there could be no philosophical
problem concerning 'the understanding' that required for its resolution the
application
of a some sort of linguistic or psychological Superscience.51c That is because we
should already have to be expert in the use of the word "understanding"
even to be
able to comprehend the formulation of the 'problem', let alone grasp its
supposed 'solution'.
Naturally, this isn't to suggest that
most scientists and philosophers don't find 'understanding' problematic, but
that 'difficulty' is a direct result of conceptual confusion.
This can be seen from the fact that if
scientists, for example, didn't already
comprehend the word "understanding", they would be in no position to
put together a single coherent sentence that expressed even the suspicion that
there was a problem concerning 'understanding', to begin with -- nor would they be able to
comprehend any of the proposed 'solutions'.
And that goes for
Philosophers, too.
That is why the difficulties theorists claim
to find with the use of words like "understanding" (and a host of other related
terms) can be attributed to conceptual confusion; if they weren't already masters of this
word, its application and associated
vocabulary, they couldn't function as educated or competent adults. The
fact that they find these concepts 'problematic' when they theorise about
them -- as opposed to when they use this word along with its associated termscorrectly every day -- shows they
are conceptually confused. 'Problems' only arise when an attempt is made to
interpret these terms theoretically (i.e., 'philosophically'), as the supposed names of
these mysterious 'inner processes' (etc.).
Either that, or they are being deliberately disingenuous.
It could be objected here that the mere fact
that we are competent users of certain words doesn't mean that understanding
and communication aren't problematic. Human beings, for example, used words like
"water" successfully for centuries, but it is ridiculous to suppose they
understood its nature (i.e, its chemical structure or why it behaved the way it
does) just because of that. In this case, the difficulty for scientists is
to give a scientific account of how human understanding works; this task
is therefore one of providing a scientific and in some cases a materialist,
theory of the way we internalise, or make sense of what is said to us (etc.). To
give an analogy: able-bodied people can walk, but that doesn't mean that they
know how they manage to do that (i.e., what muscles they use, etc.). And yet that doesn't prevent scientists
from studying the physiology of walking in order to discover its underlying
mechanisms, etc.
Or, so an objection might go.
Alas, the above analogy is lame. First of
all, our capacity to walk is plain for all to see, as is the existence of water.
That capacity, and that substance, weren't conjured into existence by
inappropriate nominalisations and reifications, as is the case with the internal
processes assumed to be identical with, or constitutive of, understanding.
Second, we don't use
walking in order to comprehend our ambulatory skills, but we have to
understand something before it can become part of theexplanation of anything
-- and that includes the supposed 'inner processes' associated with understanding
itself, as well as the nature of water. That requirement can't be bypassed or circumvented. In short, we have to be experts at using
language intelligently in order to grasp the supposed 'problem', in the first
place,
let alone its alleged
'solution'. But, in this respect we already know all there is to know
about the use of the word "understanding", and its related vocabulary.
If we didn't, we would certainly fail to comprehend the original
'problem' as well as any proposed 'solution'. This isn't the case with walking or with
the study of the chemistry of water. [Naturally, that fact doesn't
prevent anyone studying the physical concomitants of walking, or the chemistry
of water.]
So, the insistence that we need a theory
of how 'understanding' works (as part of a scientific
or philosophical account of 'the mind') was, as a matter of fact, first motivated by a series of linguistic
false moves, and nothing more. As noted above, traditionally, the phrase
"the understanding" (and related concepts and expressions) were interpreted as the names of
inner processes (as part of the
Platonic-Christian-Cartesian tradition), and that was in turn motivated by the
nominalisation of verbs
like "to understand", "to think", "to imagine" (etc.).
This means that the onlyevidence that there are such 'inner
objects and processes' is a series of spurious nominalisations and incautious
reifications!52
It is worth emphasising this point since
most theorists either ignore it or fail to recognise its significance: the
only 'evidence' that there are any
wild geese here that need chasing is
a
spurious set of nominalisations. Hence, this phoney 'chase' depends
solely on the idea that if there are names for these inner
'objects' and 'processes', there must be
objects and processes (in nature, or in our heads) which answer to them. [We have met
'word-magic'
like this several times in other Essays at this site -- particularly
here, and
here.]
In fact, to call a philosophical investigation of this sort a "wild goose chase"
would be to slander wild geese -- at least they had the decency to exist before
they were chased!
Again, it could be objected that somethingphysical
must be responsible for our understanding if we are to base it on real
material processes. In that case, as materialists we have little choice but to attribute the capacity to
form thoughts (etc.) to processes at work in the
CNS -- mediated by practice, subjectivity and ideology, etc.,
etc.
Such thoughts and processes are emergent properties of complex structures,
and have evolved as result of our intelligent use of language -- which were
themselves materially-, and dialectically-grounded in our social and economic
development. In that case, the nominalisations referred to above needn't imply that a single 'entity' answering to a given
Proper Name is responsible for all, or even most, of our psychological abilities,
qualities, processes, states, skills and affectations. It could be the case that a series of
(suitably complex, dialectical) processes in the brain (mediated by the other
features mentioned above) underpin the original emergence of
'consciousness', and thus of understanding (etc.), from its
material base.52a
Or, so it
might be maintained...
Allied with this is the claim that we use our brains to think. In that case, it might
be a good idea to examine that particular claim in more detail. So, for the
purposes of argument let usassume we do use our brains to think.
Well, we
certainly use books, pens, paper, computers -- even our hands and feet (and much
else besides) -- to do whatever we choose to do with them, but without a brain we
couldn't use or do anything at all.
I presume all
are agreed on that.
However,
if we did use our brains -- and we now agree we need a brain to use anything
-- then that must mean we would have to use our brains in order to
use our brains! But, if that were the case, we would soon have an infinite regress, since
we would have to use our brains in order to use our brains, in order to use our brains, in order
to...
If so, it makes no sense to suppose we use our
brains to do anything.
[That
shouldn't be taken to mean that the present author thinks the brain is a redundant organ! After all,
it has just been asserted that without a brain we wouldn't be able to do anything at
all.]
Admittedly,
the above conclusion isn't just controversial, it seems to be both counter-intuitive and
un-scientific -- hence, preposterous. In which case, it might be worth
developing the above argument more slowly, carefully and in more detail, to see if a mistake has been made
somewhere and to answer any objections that could be raised against it.
We may
perhaps begin with this uncontroversial assumption (to which few, if any, would
demur):
P1:Having a brain is a necessary
condition for being able to doanything at all.
From which
we may perhaps argue as follows:
P2: If P1 were the case,
we would need another brain to use our brains.
P3:
But, we
don't have a spare brain; we have only
one brain each.
P4:
Therefore, we don't use our brains.
P5: If
we don't use our brains, we certainly don't use our brains to think.
Of course,
that doesn't imply thinking doesn't take place in the brain, only that if
it does take place there, we wouldn't be in control of our thinking,
since we don't use our brains.
In response,
it might be countered that we think withour brains, but
it is unclear what that means. If it means that having a brain is a necessary
condition for us to be able to do anything at all (i.e., P1), then, once
again, that is uncontroversial. On the other hand, if it means we use our brains
to think (which is its most natural interpretation), then no sense can be
made of it (and for the above reasons).
Furthermore,
any suggestion to the contrary (i.e., that we do in fact use our brains to
think) clearly implies a modified form of Cartesianism. That is because it
means there is 'something'
over-and-above the brain that uses the brain to think. But, what
can this
'something' be other than
the 'soul', or maybe a disembodied or non-material 'mind'? Either that, or it implies that each of us
has an homunculus (a
little man) in our
heads that uses our brains (or even employs 'his' own brain) to do the thinking for us. That
would be rather like the
set-up suggested in the
Disney film,
Inside Out (and in 2024, Inside Out 2). But, does this 'little man in the head' also have a brain?
If so, the same problems would applies to 'him'. 'He' can't use 'his' brain to think,
either! On the other hand, if 'he' has
no brain -- and
having a brain is a necessary condition for anyone
to be able to do anything at all (i.e., P1) --,
then this 'little man' wouldn't be able to do anything, let alone do any thinking for
us.
P1:Having a brain is a necessary
condition for being able to doanything at all.
Some
readers might counter with a claim that that the "I" here refers to the individual concerned.
However, there are good reasons to think otherwise (on that, see Anscombe (1975), and Hacker (1993a), pp.207-28).
But,
even if "I" were a referring expression that designated the
individual concerned, its use here would still suggest that there was something
identifiable and 'internal' to each individual that was separate from the brain
(Another brain?), which uses the brain to do the thinking. But, as
pointed out above, if having a brain is a necessary
condition for being able to doanything at all (P1, again!), then this "I" would also need a brain to use its brain, and so on...
It could be objected that the brain uses
itself, or even that a part of the brain -- a module, perhaps -- uses other parts to do the
thinking. However, if having a brain is a necessary
condition for being able to doanything at all (P1), and
the brain used itself to think, then the brain
would
need a second brain in order to do that --, or even cause a given module to think. Down that route I fear lies yet another
infinite regress -- as in: the brain needs a brain, which needs a brain, which
needs a brain, which needs..., to think.
Is this perhaps still
being a little too fast?
In that
case, we can slow down the argument even more.
To that end, let us suppose that apart
of the brain uses another part to think. These 'parts' could be:
(i) Entirely
separate from one another;
(ii) Partially interlinked
among themselves; or,
(iii) Completely
interconnected with every other part/module of/in the brain.
Let us call
the part of the brain that does any of the above, "B(1)", and the part that does the thinking as
a result -- or which is
controlled by B(1) -- "B(2)". In that case, B(1)
does no thinking itself, B(2) does it all under the control of B(1).
[B(1) can be as simple or as complex and multi-layered as any given
theory needs it to be.]
But, if that were so, we wouldn't
need a brain to do, or to use, anything, as we supposed earlier (P1),
we
would
just need a part of the brain to do it. So, in order to do
anything
we would only need B(1). In that case, B(1) would either (i) be
under our control or it would (ii) autonomously control the rest of the
brain.
If (i) were
the case, we'd be back where we were a
few paragraphs ago (which would imply the existence of a 'non-material
something' separate from the brain (a soul?) that controls B(1)).
On th other
hand if (ii) were the case that would just replace "brain" with "B(1)"
in P1-P5, yielding P6-P10:
P6:
Having a working B(1) is a necessary condition for being able to do anything
at all.
P7:
Hence, if we usedB(1), we would need another
B(1) in order to do so.
P8: We have only one B(1) each.
P9:
Therefore, we
don't use B(1).
P10:
If we don't use B(1), we certainly don't use it
to make B(2) think.
[P1:Having a brain is a necessary
condition for being able to doanything at all.]
It could be
objected that the above reasoning is biased and prejudicial. While we might need B(1) to make, or enable, other parts
to think, it doesn't follow that we need it in order to do everything, or
even anything.
Different modules take on different tasks. We might even argue that thinking is
a capacity, state or activity that is distributed across an entire brain.
In that
case, let us suppose that module
M(1) does task T(1), module M(2) does task T(2),
module M(3) does task T(3),..., and
module M(n) does task T(n) -- the use of one or more of which
enables us to think. We could even suppose that
these tasks overlap or ('dialectically') interlink in some way.
Given this
revised account, one or more of the above modules must be identical with B(1), from earlier. But, that just replaces
"B(1)", or even "brain", with one or more of the following:
(i) A
conjunction
of the elements/modules, M(1), M(2), M(3),..., M(n);
(ii) A
disjunction
of the elements/modules, M(1), M(2), M(3),..., M(n); or,
(iii) A
sub-set of those elements/modules.
Hence,
B(1)
would now be the equivalent of, or function in the same way as, one or more of
the above.
Taking each in turn -- suppose (i) were the case:
[α]
If so, call the set comprising M(1), M(2), M(3),..., M(n),
"C".
Alternatively, suppose
(ii) were the case:
[β]
If so, call this set "D".
[Where "C" stands for "conjunction of elements/modules M(1), M(2), M(3),..., M(n)", "D" for "disjunction
of elements/modules M(1), M(2), M(3),..., M(n)".]
Finally, suppose (iii) were the case:
[γ]
If so, call this option, "S".
[Where "S" stands for "sub-set
of elements/modules M(1), M(2), M(3),..., M(n)".]
Clearly, one of
[α], [β]
or [γ] must be the
right choice. That is, either C, D, or S must be the correct option.
[δ] Call that
disjunction itself "M(Ω)".
So, M(Ω) isa disjunction ofmembers ofthe set: {C, D, S}.
In
that case, we can replace the word "brain" in P1, or "B(1)"
in P6, with "M(Ω)", as follows:
P11: Having a
working M(Ω) is a necessary condition for being able to do anything
at all. [That must be so since M(Ω) is in effect every module
comprising the brain.]
P12:
Hence, if we used M(Ω), we would need another M(Ω)
in order to do so.
P13:
We have only one M(Ω) each.
P14:
Therefore, we don't
use M(Ω).
P15:
If we don't use M(Ω), we certainly don't use it
to think.
[P6:
Having a working B(1) is a necessary condition for being able to do anything
at all.]
[The same result
will emerge if we take each
element of M(Ω) severally or collectively, but that task has been left to the reader
(but not her brain) to complete.]
If we
now argue that we don't useM(Ω),
or any element comprising it, to control anything, then it must beautonomous of our
will.
But, if
that were so, the result is conceded:
we don't use our brains, after
all. If M(Ω) is autonomous of our
will, we have no control over it, and hence do not use it.
That disposes of the response that thinking is a capacity, state or activity
which is distributed across an entire brain (that we met earlier). If thinking is
distributed in this way across an entire brain, but isn't controlled by
anything, then thinking must be 'independent of the will'. In that
eventuality, it isn't easy to see how it could be attributed to the individual
concerned, how any thought would be that person's thought.
The problem is that if the above were so, something elsemust control what we do
or what we think (since we plainly don't act or think randomly or
capriciously). What that 'something else'
might be is now entirely
mysterious, but it looks suspiciously like an autonomous Cartesian 'soul', or
a 'non-physical mind', once more.
It could be
objected that the argument presented above itself leads to a modified form of
Cartesianism, for if we don't use our brains, or we don't use M(Ω)
(or any sub-set of it), then it/they must indeed be independent of our
brains, and hence must be controlled by something immaterial.
But that
isn't so. Cartesianism implies there is something over and above the brain that
controls the brain. Full-blooded Cartesianism posits an immaterial soul that
does this, and which also thinks independently of the brain. Modified,
more contemporary versions of Cartesianism delegate this to the entire brain,
or to some module or modules of the brain, which do the thinking, etc. But, this
is just Cartesianism Lite; it still locates thought
in our heads. The argument presented above (i.e., P1-P15) shows
that no sense can be made even of this modified version. The implication
is that if we want to continue arguing that thought takes place in our heads,
that would automatically commit us to full-blooded Cartesianism.
By way of
stark contrast, the approach adopted at this site rejects
(in its entirety) any such metaphysical use of
language; it redirects our attention to how we actually use
words to speak about our psychological make-up and our cognitive states and
abilities (which doesn't even remotely suggest we think with our brains
or even that we do so 'in our heads'). I have referenced dozens of books and
articles (written from a Wittgensteinian perspective) that argue in support of
that approach,
here
and here, for example.
This means
that the approach adopted at this site is neither committed to the doctrine that
thought takes place 'in our heads' -- nor even its opposite, its 'negation'
-- just that no
sense can be made of eitheroption. The same applies to the claim that the brain
acts independently of our will, or that it is even controlled by our will.
Every single oneof those options is rejected, and for the same reason. P1-P15-type
reductios can be used to show that no sense can be made of each of these
metaphysical (or 'scientific') alternatives. Again, that is left to the reader to
complete for herself. [Or she can check out the literature mentioned in the
previous paragraph.]
Now, we
might be tempted to replace the above modules with a computer programme of some
description -- or something analogous to one --, claiming perhaps that we use some sort of 'software' to do all our
thinking, or, indeed, anything at all. In that case, just replace "M(Ω)" above with "software of
some sort". Nothing changes. No sense can be made of the idea that
we use our brains, part of our brains,
or even 'software', to do anything.
It could be
countered that computers actually use software to do various tasks. If so, why can't
we use
something analogous to software to do whatever we do, just like computers? However, as the above shows,
whatever it is that computers actually do, we can make no sense of the
supposition that we do likewise. Of course, computers don't actually use
anything, they just operate in the way they were intended by their designers and programmers. In which
case, a human being is necessary for computers to do anything, and if that is
so, we are back
where we were a dozen or so paragraphs ago.
Someone
might wonder what would happen to the above considerations if one day
computers became autonomous, or even 'self-aware', and then maybe they were
capable of designing and building other computers of the same sort. In that
case, no human being would be required for these computers to do
anything.
But that
is just science fiction, and when we descend into realms of fantasy like
this, where we have
no rules to guide us, no rules to determine what we can legitimately assume or
infer, and, indeed, what we
may not assume or infer,we can plainly assume or inferanything we like.
Hence, if someone were allowed to assume computers might become autonomous or
even 'self-aware', what is to stop someone else from assuming that an 'evil
genius' (analogous to the 'evil demon' Descartes introduced into his argument to
test the veracity of his own thoughts) has installed a (hidden) programme (or
'back door') in such computers that mimics autonomy and self-awareness, when the computer
concerned isn't
actually autonomous or self-aware, after all. And, in response to every attempt
to circumvent that counter-response, we assume this 'evil genius' is even more cunning than we thought and has
designed a programme to circumvent each and every such attempt. And if we now try to argue
that computer experts would surely be able to locate or detect any such hidden programmes/'back doors' in the
code, that
would simply concede the point at issue since it would amount to admitting that
this 'evil genius' had succeeded in
mimicking human autonomy, self-awareness and thought.
It would then be a short step to supposing that this 'evil genius' had fooled
every computer expert into thinking they had found this hidden code. [This what
happens when fantasy is allowed to dominate theory!]
Suppose we
now attempt to argue that other computers might still become autonomous and
self-aware
where it could be shown by computer experts that no secret programmes/'back doors' had been
installed to mimic autonomy. But, once again, what is to stop someone else arguing that
those who have checked to see if there are no secret programmes (etc.) were all
mistaken, or that they had all been bribed, conned, drugged or hypnotised to find no
such secret programmes/'back doors'? Or, indeed, that these experts are
themselves androids programmed by this 'evil genius' to come to that
very conclusion. It would be no good objecting that such a scenario is implausible since it is no less implausible to suggest that computers might become
'self-aware'. If one implausibility is permitted it is impossible to see how
the number of allowable implausibilities may be restricted to just
this one.
Indeed, in science fiction
we
aresupposed tosuspend our usual notions of plausibility.
The bottom
line here is that when we introduce science fiction
and fantasy
into philosophy, all bets are off.
Of course, there are colloquialisms -- such
as "Use your brain!" -- that seem to suggest otherwise, that we do
use our brains, but we should no more
want to take that colloquialism literally than we would want to take: "I have half a mind to enter
politics, so I will" literally, either.
However, let
us suppose some way can be found to circumvent or neutralise the above
conclusions; even then our difficulties would only just be beginning. In
that eventuality, it would be worth asking the following: If we do indeed use our brains, or we
run software that enables us to think "in our heads", or which in the end does
all our calculations/thinking for us, how do we know that any of our
thoughts and calculations are valid or correct? How might we check what our brains
supposedly conclude? How might we validate the results, or the output, of this 'internal processor'?
Especially if it controls everything we do and think? If we can't
check these results, except we use our brains and/or this 'software' to check itself
-- which would be like someone checking their own height by placingthe
palm of their hand on the top of their head -- how
do we know that anything our brain (or this 'software') has produced is correct,
even about itself and
what it supposedly does or doesn't do, does or doesn't conclude?
Again, it could be
countered that we must
be able to arrive at correct conclusions, at the truth, some, or maybe even most, of the time,
otherwise our species wouldn't have survived. But, if all we have available to
us is the output of this 'internal processor', how do we know that anything is
the case? How do we know that we have even evolved? That we have brains to begin
with? That we are
human beings? That anything else exists? No good appealing to 'evidence', since all
such
'evidence' is processed by this 'Internal Big Brother' [IBB],
which has yet to be judged trustworthy. And, of course, it will always try to
exonerate itself.
It doesn't take much
navel-gazing to see that
we are now facing exactly the same sort of problems that haunted Descartes
(indeed, this is just a variation of the science fiction option we met earlier). He
'extricated' himself from this bottomless pit of scepticism (concluding that he had neutralised the
'evil demon' he imagined might be screwing with his thoughts) by appealing to a
beneficent 'God' -- whom he imagined wouldn't allow him to be so deceived -- in
order to validate his conclusions. Here he is:
"For since
God has endowed each of us with some light of reason by which to distinguish
truth from error,
I could not have believed that I ought for a single moment to rest satisfied
with the opinions of another, unless I had resolved to exercise my own judgment
in examining these whenever I should be duly qualified for the task...."
[Descartes (1997b),
Part 3,
p.88.
Bold emphasis added.]
Must we
do likewise?
But, even
that would be no help. The question would then be: who
guarantees 'God's' thoughts?
[That is just
the theological version of this unanswerable question:
"Who guarantees the results generated by our very own IBB?"]
Of
course, Descartes's 'solution' was no solution, since everything we think we know about 'God'
has been fed to us by our very own IBB, which has yet to be judged reliable.
So, theological fiction turns out to be
no more help than science fiction was to begin with!
That is why it was asserted in Essay Three
Part Four (not yet published) that every Traditional and Modern 'Theory of Mind' collapses into
some form of
Solipsism-- via Descartes.
[On whether
we think with our
brains, see Geach (1969b). See also Note 2. Added on
Edit: Details concerning the references cited can be found in the
Bibliography. However,
no one should conclude from the above that I am a sceptic or that I accept any
of the
sceptical conclusions I have just drawn. They are only being aired in order to underline
the sceptical implications of the idea that we 'think with our brains' --
or, indeed, that 'thinking' takes place 'in the head'. In other words, it was
meant to highlight the sceptical implications both of Cartesianism and neo-Cartesianism
(i.e., Cartesianism Lite -- contemporary Cognitive Science). I have
said much more about that in Essay Thirteen Part
One, and, as noted earlier, I will say even more in Essay Three Part Four.]
Despite this, it is worth re-iterating the fact that the only
'evidence' to suggest that there are 'processes' at work in the brain/CNS
(etc.), which underlie 'consciousness' and 'the understanding' (etc.), are
the aforementioned nominalisations and reifications. In fact,
apart from tradition (i.e., the dominant Platonic-Christian-Cartesian
Paradigm), the idea held by revolutionaries that there must be such
'processes' is itself motivated by the misguided belief that materialism somehow
requires it.
It could be objected that evidence substantiating a dialectical approach
to 'consciousness' does in fact exist; indeed it is now
possible to correlate certain mental processes with neurological states (using
brain
scans, etc.). [Even so, the correlations themselves are in fact rather suspect; on that,
see Note 53.]53
However, these correlations don't prove that
an identity exists between neurological events and 'mental states', any more
than a wet pavement proves rain is identical with wet pavements, or a headache
is identical with a blow to the head!
It could be countered that rain causes wet
pavements. In like manner, certain states and process in the
CNS cause specific
psychological or mental traits -- or, at least, the latter emerge from the former.
However, as pointed out in this Essay, the belief that there are
such "mental" states and processes is motivated solely by the
nominalisation and then the
reification of verbs associated with our psychological makeup/lives. That in
turn is the result of the fact that
theorists still haven't broken with the dominant Cartesian Paradigm.
What, for instance, does the phrase "mental
state" mean? As it is ordinarily used, "mental" often means that the
relevant subject is mad or deranged; the phrase itself is connected with
psychological well-being in general -- or, indeed, its opposite. Hence we have
"mental patients", "mental hospital" and "mental health". To be sure, we all know
that "mental state" doesn't mean a
secret state or process 'in the head', of suspicious provenance and even more dubious
nature. How do we know this? Well, as Norman Malcolm points out:
"The causal theory of mind
defines the concept of a mental state as the concept of a state that has a
certain causal role. The advocates of this theory employ the term 'mental state'
in an uninhibited way. Any belief, or desire, or pain, is called a 'mental
state'. This is not the way that this expression is used in everyday life. A
twinge of pain in the shoulder cannot be called a 'mental state' -- nor wanting
a second cup of coffee, nor the belief that one left one's gloves in the car. In
everyday language a long-term anxiety or depression is called a 'mental state':
in regard to a person whom you knew to have suffered from a persisting
depression you might ask 'What is his present mental state: is he still
depressed?'
"Some of the causal theorists
are aware that the expression 'mental state' is unsuited for many of the
phenomena they want the causal theory of mind to cover. For example,
Armstrong says 'I attach no special importance to the word "state". For
instance, it is not meant to rule out "process" or "event".' This concession is
not much help, since a twinge in the shoulder is no more called a 'mental
process' or a 'mental event' than it is called a 'mental state'." [Malcolm
(1984), pp.75-76. Malcolm is referring to Armstrong's book: A Materialist
Theory of the Mind, p.82. Quotation marks altered to conform with the conventions
adopted at this site.]
Some readers might be tempted to object that
what we ordinarily say has no place here. But, that isn't so, as Button, et
al, explain (although their comments were largely aimed at the analogy
Cognitive Scientists draw between computers and human psychological
attributes, states, and processes, they nevertheless apply in general here):
"As
to the widespread disparagement of attempts to
resolve philosophical problems by way of appeals to 'what we would ordinarily
say', we would proffer the following comment. It often appears that those who
engage in such disparaging nonetheless themselves often do what they
programmatically disparage, for it seems to us at least arguable that many of
the central philosophical questions are in fact, and despite protestations to
the contrary, being argued about in terms of appeals (albeit often inept) to
'what we would ordinarily say...'. That the main issues of contemporary
philosophy of mind are essentially about language (in the sense that they
arise from and struggle with confusions over the meanings of ordinary words) is
a position which, we insist, can still reasonably be proposed and defended. We
shall claim here that most, if not all, of the conundrums, controversies and
challenges of the philosophy of mind in the late twentieth century consist in a
collectively assertive, although bewildered, attitude toward such ordinary
linguistic terms as 'mind' itself, 'consciousness', 'thought', 'belief',
'intention' and so on, and that the problems which are posed are ones which
characteristically are of the form which ask what we should say if
confronted with certain facts, as described....
"We have absolutely nothing
against the coining of new, technical uses [of words], as we have said. Rather,
the issue is that many of those who insist upon speaking of machines' 'thinking'
and 'understanding' do not intend in the least to be coining new, restrictively
technical, uses for these terms. It is not, for example, that they have decided
to call a new kind of machine an 'understanding machine', where the word
'understanding' now means something different from what we ordinarily mean by
that word. On the contrary, the philosophical cachet derives entirely from their
insisting that they are using the words 'thinking' and 'understanding' in the
same sense that we ordinarily use them. The aim is quite
characteristically to provoke, challenge and confront the rest of us. Their
objective is to contradict something that the rest of us believe. What the 'rest
of us' believe is simply this: thinking and understanding is something
distinctive to human beings..., and that these capacities set us apart from the
merely mechanical.... The argument that a machine can think or understand,
therefore, is of interest precisely because it features a use of the words
'think' and 'understand' which is intendedly the same as the ordinary use.
Otherwise, the sense of challenge and, consequently, of interest would
evaporate.... If engineers were to make 'understand' and 'think' into technical
terms, ones with special, technical meanings different and distinct from
those we ordinarily take them to have, then, of course, their claims to have
built machines which think or understand would have no bearing whatsoever upon
our inclination ordinarily to say that, in the ordinary sense, machines do not
think or understand." [Button, et al (1995), pp.12, 20-21. Italic
emphases in the original. Quotation marks altered to conform with the conventions adopted at this site.]
[On this also see Hallett (2008), particularly pp.91-99. There are dozens of
examples of the above sort of equivocation and confusion, listed and analysed in Bennett and Hacker (2021).]
Hence, if Cognitive Scientists want to explain things like "belief", "emotion"
and "thought", they will either have to use these words as they are
ordinarily employed, or they will fail to explain what they imagine they are
referring to, having replaced them with a handful of technical, but
typographically identical, alternatives. Of
course, in that case belief, emotion and thought, as we ordinarily
understand them, will thereby remain unexplained. That being so,
the idea that computers can be used to explain or help understand human
psychological attributes, states and processes
is entirely misguided. [On this, see Robinson (2003b).]
Returning
to an earlier point: neurological correlations can't show that
'mental events' are brain processes, or even that they are caused by brain
processes; that is because the term "mental event" is, at best, hopelessly
obscure, at worst, a complete misnomer. More specifically: as we have seen, the terms
employed by
theorists (e.g., "thought", "consciousness", etc.) are
either (a) the same as those we ordinarily use or (b) they aren't related to the ordinary
employment of typographically similar words. If (a) were the case, then
theorists will be misusing such terms, and what they have to say by means of
them would be both misguided and confused. Alternatively, if (b) were the case, then
any alleged correlations can't illuminate thought or consciousness, but 'thought' and 'consciousness', words that have a technical, and as-yet-unexplained,
meaning, and we would be none the wiser.
Indeed, as Malcolm points out,
correlations can't possibly
capture what we mean when we use ordinary psychological predicates (etc.); in
which case these correlations are philosophically uninteresting -- howsoever
scientifically illuminating, or otherwise, they might be. [In connection with this,
however, see the
suggestions made in
Note 37.]
Naturally, the above comments aren't correct simply because of the present
author's diktat; Note 53 and the
references given in Note 86, are aimed at
substantiating these seemingly perverse (if not apparently anti-scientific)
assertions. [On this, see also Hanson
(1971b).]
A
comparison might help here. Let us suppose that there turned out to be a
verified correlation between the exchange values
of certain commodities and specific 'brain', or 'mental states', of those who
had, say, either manufactured or purchased those commodities -- assuming
for the moment that such phrases had a clear meaning and weren't dependent on the
aforementioned Cartesian Paradigm. The question is: Would
we abandon the labour theory of value as a result of this hypothetical discovery? The
answer, I trust, is pretty clear; if a link like this were demonstrable, we
should rather say that the value of the commodities in question (or some other
'objective' feature of the world) was responsible for producing the said
'brain'/'mental states', not the other way around (or something along those
lines), and advise further research. We wouldn't want to abandon our commitment to the truth of
propositions ascribing value to commodities if neurological concomitants were
one day unearthed by scientists -- no matter how well confirmed they might prove to be.
Of course, the obvious objection to the
above would be to counter-claim that no such correlations between commodity
values and brain/mental states exist, nor are they ever likely to. But, that
riposte would miss the point; what goes on 'in the head' can't provide the basis for an
ascription of social concepts to anything that a Marxist should rightly regard
as significant -- in this instance, 'objective' economic factors, but in the case under
consideration, the products of social relations among human beings expressed by our ordinary use of language in connection with familiar psychological
attributes (etc.). [And this is a view shared by Voloshinov, no less --, that is,
as far as can be
ascertained. On that, see Note 11
and Note 23.]
Hence, the position defended here isn't that
psychological and neurological research into the brain is misguided, but that reference to what goes on in the
CNS (etc.) can provide
no insight into linguistic meaning -- even if one were needed. Just as
the value of a commodity is a consequence of complex social structures,
processes and inter-relationships (and isn't dependent on what transpires inside the heads
of individuals collectively or severally -- the latter is indeed part of
the folly behind
Rational Choice Marxism), our use of language is also the product of collective
action and interaction (and is similarly irreducible to individual 'mental acts' -- even if
there were such a thing).
[This
topic overlaps with a recent controversy in Analytic Philosophy between
"Internalism" and "Externalism". Since this doesn't impact on
ideas of central concern to Marxists, no more will be said about it here. However, since "Externalism" is relevant to some of the ideas developed in
Essay Twelve
Part One, a section on this topic
will be added there at later date.]
On
these alleged correlations, and the so-called "modularity of mind", see
Uttal
(2001). Here is Professor Uttal's summary of some of the main points of his
recent book
(however, readers mustn't
assume that I agree with everything he says -- he, too, is clearly in the grip
of the aforementioned Cartesian Paradigm, evidenced by his talk about the relation between 'mind' and 'brain', and his reference
to "mental processes", for instance
--; these comments are only being quoted here for his view of these alleged correlations):
"On the Limits of Localization of Cognitive Processes in the Brain.
"Psychology has always been in search of metaphors and explanatory theories.
Earlier we had to do with hydraulic, mechanical, electrical, and eventually
computer models to serve as
heuristics
to help guide our thinking about the
nature of cognition. In this century a new science --
neurophysiology
-- and a
remarkable collection of new
physiological
recording tools have become available
as an alternative to these older metaphors. We have gone through a series of
physiological measures including, the
galvanic skin response, the
electroencephalograph, and the
evoked brain potential, each of which promised to
provide a material key to understanding mental activity. All of these methods
were especially exciting for psychologists because they promised to provide a non-invasive means of correlating brain activity with mental actions. In the
main, however, none of these methods has been successful in answering even the
most basic questions of how the brain produces or encodes mental activity. The
main reason for this failure has been the fact that these measures are asking
questions as [at? -- RL] the wrong level. The ultimate basis of mental activity must be the
informational state of a huge collection of
neurons
interacting, not en masse,
but as an intricate web, a network in which the details of the intercommunicated
information are salient. Measures of integrated activity such as the
EEG
or the
EVBP
simply do not assay the essence of the relationship between mind and brain.
"The latest
'new' methodology
"Now there is another entry in the search for a metaphorical model. The
availability of the
PET
and
fMRI
scanning procedures in the last decade has once
again excited psychologists. Indeed, it has more than just excited them. Entire
sections of experimental psychology in some of our most prestigious university
departments have abandoned purely cognitive studies in favour of correlative
studies of these images and behavioural tests. Furthermore, some departments have
frighteningly over committed their resources to this single line of research. I
believe this to be a programmatic error that is based upon an inadequate
consideration of the basic assumptions and logic of the research that is
emerging willy-nilly from this breathless attack on one of the most fundamental
questions of
psychobiology
-- the issue of whether or not mental processes can be
localized in particular regions of the brain. It seems to me that there should
be a cooling off period before we charge ahead into a research paradigm that has
many unanswered questions and faces many conceptual, technical, and logical
problems.
"In the following paragraphs, my goal is to raise some cautions and to
stimulate a bit of reflection about what is currently going on in many
neuroscience
laboratories. Some of the cautions are age-old ones, but some are
associated with the most modern technical matters.
"Six suggestions
"First, perhaps the most difficult challenge that has to be faced by those who
are comparing brain images and cognitive processes is the uncertainty involved
in precisely defining the components of mental activity. Throughout the history
of psychology, we have tried to define mental activity in an enormous number of
different ways. Other than the antique and persisting trichotomy of
'input-central-output', efforts to develop sharp definitions of mental modules
have been notoriously unsuccessful. Every century defines their own mental
components and few of these definitions are perpetuated into the next. A few
very general terms persist -- memory, emotions, percepts, etc. -- but even these
are fraught with lexicographic difficulties. Arguably, the mental modules that
psychology currently uses are either a priori or ad hoc hypothetical constructs
or are operationally defined by the experiments we use to study mental activity.
At least one survey (Grafman, Partiot, and Hollnagel, 1995) goes on for seven
pages listing the variety of cognitive processes that have been associated with
the
frontal cortex
in particular! Clearly, an adequate classification of mental
processes is not yet at hand.
"Second, the findings that have emerged from the scanning-cognitive
laboratories are not yet stable. Pulvermüller (1999) has pointed out that the
cognitive processing of word meanings has been 'located' in all of the major
lobes of the brain! Few studies are replicated under the same conditions, and
often those that are do not support each other.
"Third, there is ample evidence, especially that emerging from some of the
newer event-related scanning procedures that the cognitive processes are not
localized but the result of widely distributed action in the brain.
"Fourth, there is a host of technical uncertainties and a highly fragile
logical chain between neural activity and the scanned outputs from fMRI and PET
systems and even more concern about what these signals mean. Experts in the
field are well aware of these difficulties, but often we psychologists take at
face value some highly dubious steps in the logic. At the very least, it must be
appreciated that it is a mathematical truism that any bounded
field
will exhibit
a maximum. This means that there will always be a peak of activity someplace in,
for example, a fMRI image. Correlations between behaviour and cognitive activity
are, therefore, guaranteed regardless of the actual biology of the situation.
The emphasis on 'hot spots' incorrectly directs attention away from critical
changes of activity in other regions -- both increases and decreases.
"Fifth, The statistical and experimental design aspects of the scanning
procedures are also matters of deep concern. Small shifts in criterion levels
can force drastically different interpretations of data. Normalization and
averaging procedures may produce spurious conclusions concerning localization.
The frailty of the subtraction and double dissociation methods, and the
elaborate processing necessary to see anything at all raise serious concerns
about whether this new approach will fail in the same way that the older methods
did to answer the most basic questions faced by cognitive neuroscience.
"Finally, despite its implicit acceptance by many researchers in this field,
the localization versus distribution issue remains unresolved. There is a
theoretical bias toward 'localization' abroad in cognitive neuroscience these
days that may be totally unjustified. The entire scanning-cognition effort is
based upon the assumption that mental processes or modules are actually
localized in particular regions of the brain. However, there is abundant
evidence that this may be a misreading of the data. The brain is a highly
interconnected, redundant, and nonlinear system that is more likely to use a
distributed representation scheme than a highly localized one. Localization is
an easy way out for experimental design, but it may be fundamentally incorrect
in principle. Not in the sense of any obsolescent idea of 'mass action' but,
rather, in terms of a complex network of interacting parts. There is, in this
regard, a great confusion in this field over such a simple matter as the
necessity versus the sufficiency of a brain region's role in a cognitive
process. Experiments may quite properly show that one region of the brain is
necessary to carry out some mental task, but that does not rule out the
possibility that many other regions are also required for the process to occur.
The 'necessary' region may not be 'sufficient' to encode the cognitive act. The
emphasis on associating one or a few regions with some cognitive task may thus
produce an illusion of localization where none, in fact, exists.
"Conclusion: The Challenge
"I hope that my readers will not do the field of cognitive neuroscience the
disservice of dismissing this essay as just a 'pessimistic' view. Given the
state of the science, it may be more realistic than pessimistic. At the very
least, it seems to me that we should be considering these issues rather than
plunging ahead into what may be an enormous waste of resources and time.
Whether...my point of view is correct or not, there is an obligation to at least
consider the questions that are raised here.
"In this brief opinion piece, it is not possible for me to provide the
scientific citations to support the assertions that I make. A much more complete
rendition of the argument against an assumption of brain localization, and,
thus, the importance of a considered evaluation of what psychologists are doing
in scanning laboratories is presented in my forthcoming book -- The New
Phrenology: The Limits of Localizing Cognitive Processes in the Brain (MIT
Press. 2001).
"References:
"Grafman, J., Partiot, A., & Hollnagel, C. (1995).
'Fables in the prefrontal
cortex'. In Behavioral and Brain Sciences18, 349-358.
"Pulvermüller, F. (1999). 'Words in the brain's language'. In Behavioral and Brain
Sciences22, 253-336.
"William R. Uttal
"Professor Emeritus (Psychology) University of Michigan
"Professor Emeritus (Engineering) Arizona State University" [Partially
quoted from
here
(the old link to the full passage is now dead). Quotation marks altered to conform with the conventions adopted at this site;
spelling modified to agree with UK English.]
[Professor Uttal has updated and augmented his criticisms in Uttal (2008, 2011).
There is a Podcast of some of his ideas available
here (it would be wise to open this in a new window or tab, and there is a
slight delay
before the Podcast begins), as well as a PDF transcript of an interview with Professor
Uttal,
here.]
Again,
it is worth pointing out yet again that like so many working in this area, Professor Uttal
is still labouring under illusions generated by the Cartesian Paradigm.
Uttal (2004) tries to break free from its spell, but it is nevertheless clear
that he has failed in that endeavour. That can be seen, for example,
from the following comment found in the above PDF:
"Our best guess now --
without a lot of supportive evidence -- is that the brain is best studied at the
level of the neuronal nets (the network of cells that make up the brain tissue);
that that is where the mind resides, in terms of the information processes at
this very fine microscopic network."
There are numerous
other comments like this scattered throughout the interview mentioned above, the Professor's
books, and the summary quoted earlier.
To be sure, some might want to argue that if
mental events had no material/dialectical base, that would leave it open for others to postulate a
non-material platform for human 'consciousness', which would clearly possess unwelcome Idealist implications.
Of course, Idealists are going to argue for
an immaterial mind whatever we say. But, in order to avoid the implication that
their theory leaves room for immaterialism, materialists have simply
assumed they have no choice but to postulatejust sucha 'material foundation'
for thought in hypothetical processes in the CNS (but which aren't reducible to
them), etc. However, what this material base could possibly be is seldom
spelled-out in any detail. And no wonder! Given Lenin's 'definition' of
matter, it isn't at all clear
whether any such account could be given.54
Be this as it may,
HMdoesn't in fact
require such a (metaphysical) theory of 'consciousness'.55
That, of course, hasn't stopped dialecticians from wanting to impose
their version of this Ancient
Paradigm on hypothetical processes and structures in the brain, 'the mind',
or the CNS -- and despite
the constant refrain that this is something they
never do. Nevertheless, this imposition involves them reading into the phenomena the view
that 'consciousness' is an "emergent property" of the CNS --
dialectically linked to increased neural complexity, social development and practice -- in order to support
the idea that there is (or could be) a 'scientific', or a 'dialectical', theory of 'mental
phenomena' not reducible to the "crude" operation of mere "matter and motion".55a
There is no little irony here: in order to avoid
Idealism, DM-theorists have spirited ("divined") into existence several
highly obscure 'concepts' (which, by shear coincidence, I am sure, turn out to have been borrowed from
Traditional Metaphysics and
Neo-Platonic
Mysticism -- thus imbuing their ideas
-- even
here -- with an impeccable ruling-class pedigree). But, these 'concepts' readily collapse
into LIE since they, too, are dependent on the derivation of a set of
psychological truths from the (altered and
distorted) meaning of certain words.
In trying to avoid Idealism, DM-theorists have simply slipped right back into
it!
[This
slippage was explained in detail in Essay Twelve Part One,
as well as here.]
Unfortunately, few branches of science are as
suffused with conceptual confusion as Psychology (and that comment doesn't just apply to
DM-theories of 'cognition'). It is highly doubtful, therefore, whether Voloshinov's attempts to clarify
matters will greatly alter this sorry state of affairs (despite
the plaudits of his epigones).56
It could be objected here that while
Voloshinov himself provides no evidence in support of his claims, there is
evidence that substantiates what he says, and John Parrington's article, for
instance, summarised some of it for us.
However, when we examine the assembled
'evidence' we find it is indirect and
allusive,
at best. In
fact, Parrington's 'evidence' is perhaps more accurately to be described as mere
supposition. As he himself admits (but note once again the
neo-Cartesian language!):
"The problem with studying
inner speech is that it is impossible to observedirectly using
objective scientific methods, hidden as it is within the mind of the
individual. However, much valuable information about [inner speech's -- RL]
character has emerged by using some ingenious indirect methods….
"…An excellent attempt at
describing what inner speech would sound like if wecould actually
hear it is
James Joyce'sUlysses….
"…A study of literature may
seem a strange way to investigate the workings of the mind, but
Bakhtin
believed
that novels 'permit readers to see things that are obscured by the
restraints on expression in other applications of language.'…
"…[M]atters become
necessarily more speculative once we start to consider some of the
possible concrete mechanisms whereby social change is translated into change in
consciousness itself. Part of the problem will always be our inability toaccess inner speech directly. However, it should be clear from
Voloshinov's work, as well as that of Bakhtin and
Vygotsky, that a wide range of
sources can be used successfully as an indirect source of information
about our thought processes…." [Parrington (1997),
pp.134-35, 141, 143. Bold emphases added. Quotation marks altered to conform
with the
conventions adopted at this site.]
But, if we can't access "inner speech"
directly, how do we know it even exists? Worse still: Do we have any idea what the phrase "inner speech" could possibly mean?
Are we to imagine that cells or neurons in our heads hold conversations with
each other? Must we suppose that certain "modules" in the brain chat amongst themselves over
the synaptic fence? Does the brain give lectures to the spinal chord?
Is
Parrington perhaps referring to sub-vocal movements of the larynx? That is, 'outer speech' with the volume turned way
down?56a0
In fact, Parrington finds he has to refer to
James Joyce's Ulysses to provide his readers with a vague sort of idea what "inner
speech" might be. But, if we all know (from
introspection?) what "inner speech"
is,
we should hardly have to be told. Anyway, Parrington's own phrasing
indicates that few of us (if any) know what this mysterious 'inner process'
actually is.
This can be seen from
his use of the prefixing clause: "if we could actually hear" "inner
speech". He wouldn't have needed to add that rider if we all knew what to
listen out for, or what he was talking about.
In addition, it is also worth asking how Parrington
knows that certain novels reveal "inner speech" to us. All he says
in support of this odd idea is
this:
"An excellent attempt at
describing what inner speech would sound like if wecould actually
hear it is James Joyce's Ulysses…." [Ibid., p.135. Bold emphasis
added.]
But, if he has never heard "inner speech", and no one
else has, how could he possibly know that certain novels are capable of reveal anything to
us in this respect? Apparently, Parrington's onlyevidence that the
above novel (along with its many cryptic allusions to everything under the
Sun?) can
serve as an accurate (scientific!?) source is based on abelief expressed by
Bakhtin(a
non-scientist!) about novels in general! Exactly how Bakhtin himself knew
that certain novels record "inner speech" is no less
mysterious. No doubt, as Lenin said of Hegel,
Bakhtin must have "divined" it.
[Several detailed comments
about Vygotsky's work will be added at
the end of this Essay
at a later date. Until then the reader is directed to Chapter Ten of Williams
(1999) -- 'Vygotsky's Social Theory of Mind'.]
Perhaps then, Parrington is alluding to soliloquy,
or to other vocal/sub-vocal ramblings? If he is, then we already understand what these
are, and we have direct access to them (unlike "inner speech").
Moreover, and better, we don't need to be referred to passages lifted from
obscure "stream of consciousness" novels to tell us any more about
these phenomena, nor need we appeal to
indirect evidence to identify them (again, unlike "inner speech").
If so, Parrington
can't be referring to soliloquy (etc.). Perhaps he is alluding to the thinking we all
supposedly engage in while awake, or while reading (say)? If so, why call this
"inner speech"?
Maybe, then, Parrington referring to the low
mumblings that certain individuals produce as they read? But, not everyone
mumbles sub-vocally to themselves when reading. In fact, speed readers don't mumble at all. Even so,
and once again, we already know what this is, and we have direct
access to it (unlike "inner speech").
Nevertheless, this 'inner dialogue', about which
we are all supposedly aware -- or half aware --, which allegedly accompanies
our waking moments, is perhaps Parrington's best candidate so far. But, we can
all hear (in the sense of "attend to") this commentary, so it can't be "inner
speech", either, or Parrington wouldn't have had to labour this point quite
so much. We would all know what he was talking about if this were the
correct
candidate.
More pressing, however, is this question:
How
does this approach to 'inner speech' manage to avoid undermining belief in
the social nature of language?
Gilbert Ryle's comments are oddly apposite
here (since they were based on a firm and unequivocal commitment to the social nature of
language):
"This trick of talking to
oneself in silence is acquired neither quickly nor without effort; and it is a
necessary condition of our acquiring it that we should have previously
learned to talk intelligently aloud and have heard and understood other people
doing so. Keeping our thoughts to ourselves is a sophisticated accomplishment.
It was not until the Middle Ages that people learned to read without reading
aloud. Similarly a boy has to learn to read aloud before he learns to read under
his breath, and prattle aloud before he prattles to himself. Yet many
theorists have supposed that the silence in which most of us have learned to
think is a defining property of thought. Plato said that in thinking the soul is
talking to itself. But silence, though often convenient, is inessential, as is
the restriction of the audience to one recipient.
"The combination of the two
assumptions that theorizing is the primary activity of minds and that theorizing
is intrinsically a private, silent, or internal operation remains one of the
main supports of the
dogma of the ghost in the machine. People tend to identify their minds with
the 'place' where they conduct their secret thoughts. They even come to suppose
that there is a special mystery about how we publish our thoughts instead of
realizing that we employ a special artifice to keep them to ourselves." [Ryle
(1949a), p.28. (This links to a PDF.) Bold emphasis and link added.]
Although
Voloshinov gives lip service to the social nature of language, his
commitment to 'inner speech' only succeeds in undermining that commitment (as
readers will see for themselves if they follow the above link). It seems that Parrington's approach does likewise.
Be this as it may, Parrington then proceeds
to tell us that one of the major features of "inner speech" is:
"...its predicative
(subject-less) character. A child talking to itself 'already knows' what he or
she is talking about and therefore there is no need for naming the subject….
Inner speech must be even more telegraphic and abbreviated, and probably uses
words that are highly personally coded -- that is they have a private meaning
for the person who is using them, which may be different from their accepted
social meaning." [Parrington (1997), p.135.]
Here, he is drawing on a limited contrast between "inner
speech", soliloquy and the ramblings of children. While all of these are
"predicative" to some extent, "inner speech" is perhaps even more so
-- or, so he tells us. But, the
subjectless nature of such 'language' here isn't unique to any of the latter. Conversations
between two or more people often take much for granted, including the subject of
discussion. Many are "predicative", and many are coded (think of
Cockney
Rhyming
Slang, 'rap', etc.). So, it seems that "inner speech" is much like "outer speech" in this
regard, too.56a
All this is quite apart from the fact that
Parrington can't possibly know that 'inner speech' is how he says it is, since
we have already been told that:
"The problem with studying
inner speech is that it is impossible to observedirectly using
objective scientific methods, hidden as it is within the mind of the
individual. However, much valuable information about [inner speech's -- RL]
character has emerged by using some ingenious indirect methods….
"An excellent attempt at
describing what inner speech would sound like if wecould actually
hear it is James Joyce's Ulysses…." [Ibid., pp.134-5. Bold emphases
added.]
If 'inner speech' can't be observed directly
(but can it even be observed indirectly?), and if we can't actually hear
it, how does Parrington know it is "predicative"? I rather suspect
Parrington has confused his own (or someone else's) guess with evidence,
here.
This, of course, only deepens the mystery; if "inner" and "outer" speech
are no different, at least in this respect, why all the fuss? And how is it possible for
the allegedly "private meanings" of certain words to engage with, or be
captured by, "inner speech"
and thus be of any use? If certain meanings and certain words are private, even though no one can
hear the 'speech' that is supposed to encapsulate them (otherwise Parrington would have appealed to that
phenomenon as a fact, and thus wouldn't have bothered with all that 'indirect' evidence), what linguistic function could they possibly serve? How could
'private meanings' even be recorded (imprinted) if no one has access to
them, and no one can hear them?56b
Anyway, it is reasonably clear from other things that Parrington says
that he is alluding to something much deeper than mere soliloquy,
as he himself notes:
"…[I]nner speech is the link
between thought and language…. [T]here is a gap between thought and words…inner
speech is the fluid interphase where meaning can start to be formed and
shaped…." [Ibid., p.135.]
If "inner speech" stands between "thought" and
language, then it can't be identified
with any meaningful use of language, let alone any that is exhibited in soliloquy.
However, Parrington is rather coy about what this 'something'
actually is. Even so, he does refer his readers to studies that
Vygotsky completed several generations ago (pp.133ff.), but he failed to direct
them to more recent research carried out into this obscure phenomenon (it if is
one!) -- for example,
that supposedly relating to children.57
Even if Parrington had done this, we would still be no further forward, for
we still have absolutely no idea what "inner speech" is; until
that daunting
problem -- in fact it is a pseudo-problem, as we
have seen,
and will see -- is resolved we are in no position to decide what would even count
as evidence for or against 'its' existence. If we haven't a clue what we are
looking for, any evidence gathered could, for all anyone knows, relate to
something else, or, indeed, to nothing at all. As noted in Essay Six: you can
look for your keys if you don't know where they are but not if you don't know what
they are.
Moreover, it is unclear how Vygotsky himself
was able to study something that Parrington elsewhere declares no one can
directly experience. If Vygotsky had in fact succeeded in listening to the
"inner speech" of children as they spoke to him, as they conversed with others.
or as they babbled to themselves, then this would make "inner speech" look
pretty
direct, and hence unproblematic, since it would be a clear example of these
children learning to soliloquise, or, indeed, to ramble aloud. Such phenomena then,
if that is what Vygotsky observed (or was referring to), wouldn't
count as "inner speech" -- at least, not as Parrington seems to understand the term.58
All this is, of course, in addition to the serious philosophical difficulties
(outlined
earlier) associated with "inner speech".
Putting these annoying quibbles to one side
for now, Parrington clearly wants to read more into "inner speech" than even the dearth of 'evidence'
he presents permits, for, as we saw above, a few pages later we are
informed that:
"…[I]nner speech is the link
between thought and language…. [T]here is a gap between thought and words…inner
speech is the fluid interphase where meaning can start to be formed and
shaped…." [Ibid., p.135.]
How Parrington knows that "inner speech" is such a
link he once again failed to say. [It is to be hoped he isn't trying to impose yet another thesis on the brain/'mind'!]
The question now is: Is "inner speech" even a
linguistic phenomenon? If it is, how could it be an interface
between language and 'thought'? If 'thought' and language absolutely require
just such an intermediary, and "inner speech" is indeed a linguistic
phenomenon, then there would have to be an analogous link between
'thought' and "inner speech", too. On the other hand, if "inner speech" doesn't
itself need such an interface with 'thought', why then is one needed between
ordinary 'outer' language and 'thought'?
Alternatively, if "inner speech"
isn't a linguistic
phenomenon, why call it "speech" and credit it with other linguistic
features, such as meaning and predicativity, in the first place?
More problematic, however, is the fact that
the occurrence of episodic bouts of "inner speech" -- if they
aren't examples of
soliloquy, etc. -- would normally be regarded as clear evidence of a psychotic
personality disorder
in the one so afflicted. Given what little we are told, such inner voices
would be a sure sign, not of a fluid interface between 'thought' and language,
but of a deranged or split personality. Small wonder then that Ulysses
seemed to some to be so apposite. What next? The 'memoirs' of
Charles Manson
or Peter
Sutcliffe?
Nevertheless, if we reconsider the following
words, they might help us understand what Parrington really meant:
"…[I]nner speech is the fluid
interphase where meaning can start to be formed and shaped, based on the
emotional, practical and social experience of the individual…." [Ibid.,
pp.135-36.]
But, what sense of "meaning" is this? Is
Parrington
speaking about linguistic meaning? If so, it would be of little use in
helping us understand Voloshinov, for according to him:
"Meaning does not reside
in the word or in the soul of the speaker or in the soul of the listener.
Meaning is the effect of interaction between speaker and listener produced
via the material of a particular sound complex." [Voloshinov (1973),
pp.102-03. Bold emphases added.]
Plainly, such an "interaction" can't reside
in the head of either interactor. Hence, if Parrington is trying to make
Voloshinov's ideas clear, contradicting him is hardly a good place to begin!
[However, as we have seen (here
and here), the source of this difficulty
lies in the fact that Voloshinov can't make his own mind up whether
meaning is a social phenomenon, or whether it is a private, 'internal' affair. Parrington has obviously inherited
this confusion.]
Of course, part of the problem here is the fact that the word
"meaning" itself has many different meanings;
here are just a few:
(1) Personal Significance: as in
"His Teddy Bear means a lot to him."
(2) Evaluative Import: as in "May Day
means different things to different classes."
(3) Point or Purpose: as in "Life has
no meaning."
(4) Linguistic Meaning, or
Synonymy: as
in "'Vixen' means 'female fox'", "'Chien' means 'dog'",
"Comment vous appelez-vous?" means "What's your name?", or
"Recidivist" means someone who has resumed their criminal career.
(5) Aim or Intention: as in "They mean
to win this strike."
(6) Implication: as in "Winning
this dispute means that management won't try another wage cut again in a hurry."
(7) Indicate, Point to, or Presage:
as in "Those clouds mean rain", "Those spots mean you have measles", or
"That expression means she's angry".
(8) Reference: as in "I mean him over
there", or "'The current president of the USA' means somebody different at
most
once every
eight years."
(9) Artistic or Literary Import:
as in "The meaning of this novel is to highlight the steep decline in political integrity."
(10) Conversational Focus: as in "I
mean, why do we have to accept a measly 1% offer in the first place?"
(11) Expression of
Sincerity or
Determination: as in "I mean it, I do want to go on the
march!", or "The demonstrators really mean to stop this war."
(12) Content of a Message, or the
Import of a Sign: as in "It means the strike starts on Monday", or "It
means you have to queue here."
(13) Interpretation: as in "You
will need to read the author's novels if you want to give new meaning to her
latest play", or "That gesture means those pickets think you are a
scab."
(14) Import or Significance:
as in "Part of the meaning of this play is to change our view of drama",
or "The real meaning of this agreement is that the bosses have at last learnt their
lesson."
(15) Speaker's Meaning: as in "When
you trod on her foot and she said 'Well done!' she in fact meant the exact
opposite".
(16) Communicative Meaning: as in "You
get my meaning", or "My last letter should tell you what I meant", or "We have
just broken the code, hence the last message meant this...."
(17) Explanation: as in "When
the comrade said the strike isn't over what she meant was that we can still
win!", or "What is the meaning of this? Explain yourself!"
(18) Translation, or a Request for
Translation -- as in "What does 'Il pleut' mean in German?"59
This isn't to suggest that these are the only meanings of "meaning", or that
several of the examples listed don't overlap. [For example, items (4) and (17)
intersect, as do (5) and (11), and (9) and (14), as well as (4) and (18). For
more on this and the distinction between "natural" and "non-natural" meanings
see
Grice (1957) -- this links to a PDF -- and Stainton (1996), pp.103-10,
although I don't necessarily agree with everything Grice had to say.]
From what little Parrington says, it looks as if he might have meant (i.e.,
"intended") senses (1), (2), and, of course, (15).
Nevertheless, it seems reasonably clear that
many of the problems confronting Parrington, Holborow and Voloshinov's accounts
of language arise from their failure to notice that this apparently simple
word (i.e., "meaning", and its cognates) is in fact highly complex. Because
they have conflated several different connotations of this word, their ideas
naturally create confusion instead of dispelling it --indeed, as we have seen.
However, and once more: In this they are in
good company: most Traditional Philosophers have done (and still do) the very
same thing.60
Earlier we had occasion to quote the
following passage from Parrington's article:
"…[I]nner speech is the link
between thought and language…. [T]here is a gap between thought and words…inner
speech is the fluid interphase where meaning can start to be formed and
shaped…." [Parrington
(1997), p.135.]
The first thing that strikes one about the above
comment is that Parrington appears to think that thought and language are
distinct, so that the former can exist without the latter. This might be to
misinterpret him, but he does invite misunderstanding when he says such
things. [On this, see here.]
Now, although Parrington asserts that there is
here a "gap",
he neglected to show that there is indeed one. Worse still, he failed to explain what a supposition
like this could possibly mean. For example, might such a "gap" be measured in centimetres, seconds,
or missing teeth? If not, what sort of "gap" is this? Is it a literal "gap"
(like the space that exists between the platform and the trains in certain
underground
stations), or is it metaphorical (like a gap in someone's memory)?
Well, perhaps he is alluding to an explanatory
"gap"? But, if so, there is no such thing. If, per impossible, there were, it would 'close'
even
before it 'opened'. That is because the supposition that there is such a
"gap" would have to be expressed in the same medium either side of the supposed
divide -- in thought and in language --, thereby 'closing' the alleged
"gap". Plainly, the thought that there might be such a "gap" and its
linguistic expression are one and the same.61
To some, these claims might seem somewhat dogmatic, if not
perverse and wrong-headed. Hence, it
could be argued that if there is an objective gap between thought
and language, the above constraints on its explanation, even if correct,
are surely irrelevant. The gap either exists or it doesn't -- or so the argument might
go.
However, any
who doubt the claims made in the last but one paragraph are invited to say to themselves: "There is a gap between thought
and language" and then repeat the same 'thought' without using any words
at all! Upon doing that (or, in fact, uponfailingto do it!), they
will soon see there is no such "gap".61a
Indeed, it is worth reminding
ourselves that the deflationary argument
presented above was originally aimed at countering the idea that there is an explanatory
gap in our knowledge, and it sought to establish (indirectly) that our mastery
of language shows that no such "gap" exists -- in the sense that the supposition
itself made no sense, not that it was empirically false.
This is largely because the formulation of the thought that there is such a
"gap" and its linguistic expression are one and the same, as we have
seen.62
All of this is quite apart from the fact that
the supposition that there is a "gap" is itself based on the idea that the words
used to describe 'either side' of it are the names of 'internal objects and processes'.
In turn, the existence of these'internal objects and processes' is
based solely on this
nominalisation! Because of that, 'language' and 'thought' have now been
separated, thus producing a spurious "gap" by the simple expedient of inventing artificial
names like these, and nothing more! Hence, the "gap" Parrington
refers to is a consequence of this linguistic false step.
Or, so things at least seemed to the
Traditional Theorists who invented this way of depicting the 'mind' (even if
they might not have put it this way!). Plainly, to them, this meant that these spurious
entities ('thought' and 'language') needed to be 're-connected'.
[However, this is just as empty a supposition as thinking
that the word "God" and the word "Satan" imply there is a gap between these two! What two?!]
From this
tightly knotted web of confusion out
popped the 'philosophical problem' of the 'gap' between 'language' and
'thought' -- and, indeed, between 'mind' and
brain!
In that case, all we have here is yet another
spurious 'problem' that has arisen from a crass
use/distortion of ordinary
language -- and nothing
more.
Again, as Marx noted:
"The philosophers have only
to dissolve their language into the ordinary language, from which it is
abstracted, in order to recognise it, as the distorted language of the actual
world, and to realise that neither thoughts nor language in themselves form a
realm of their own, that they are only manifestations of actual life." [Marx
and Engels (1970), p.118. Bold emphasis added.]
On
the other hand, if the words "language" and "thought", for example, don't actually name, or refer to, any such objects
and
processes, then
the assumption that there is a "gap" between the things they supposedly denote
is baseless.
[On this, see Note 61. The idea that all words are
names was debunked in Essay Three
Part One.]
In which case, not only is there no "gap", there are no (named) objects or processes here to form one,
in the first place!
As should now seem clear, here as elsewhere, this "gap" has only opened up because of the
literal interpretation of
an inapt metaphor
-- compounded by yet another distortion of language.62a0
Nevertheless, it could be objected that this
doesn't even begin to deal with the scientific problem of the relation
between language and thought. Hence, it could be pointed out that Parrington might
have meant that there is a gap in the current scientific explanation of
the connection between thought and language, one that Voloshinov's ideas helped
close. That appears to be why Parrington said the following:
"However, this still leaves
us with the question of the concrete mechanisms whereby this process takes
place." [Parrington
(1997), p.136.]
From this it seems that the supposed "gap"
might be a euphemism for our
profound lack of knowledge of the physical, mental or
psychological mechanisms or causal links, or, indeed, "mediations", that connect 'thought' and
'language'.
But, a few lines earlier Parrington had already declared that:
"Language, therefore, is not
just an expression of otherwise independent and fully formed thought, but rather
is a necessary form of the thought's realisation." [Ibid., p.135.]
This appears to mean that our only handle on
thought is purely linguistic. This further seems to suggest that
Parrington himself half accepts the view that (i) thought isn't in fact an aspect of the
'mind' that can be isolated independently of its linguistic expression, and
that (ii) there is
thus no "gap" between thought and language, after all.
[However, things aren't quite this simple. On
that, see Note 61, again.]
Nevertheless, there are other things that
Parrington says which suggest he failed to appreciate the significance of the
above
admission. As we found with "meaning", the word "thought" (and its cognates) is
far more complex than most theorists seem to acknowledge. Again, it is only when
philosophers try to theorise about this 'concept' (and thus restrict the
meaning of the word "thought" to what goes on in our
heads) -- as opposed to when they
use language normally to express their thoughts and to understand the thoughts of
others -- that confusion arises.62a
If so, there is no object or
'mental process' here called "thought";63any supposition to the contrary
can
only have been prompted by yet another inept linguistic
reification. Moreover, what is true of "thought" is also true of "language" (and for the same
reason). In which case, there aren't two objects or processes here (inside or
outside the head) for there to be a "gap" between, or for
science to study.64
The fact that Parrington has been misled --
as have so many others -- by a series of spurious reifications like this is
confirmed by the way he poses the problem: it is
only if thought and language are understood as literally two sorts of
objects or processes that a "gap" could emerge between them (even if this is
just an explanatory "gap"). Otherwise, his use of this word (i.e.,
"gap") is surely
metaphorical.
Be this as it may, traditionally, several competing media have
been proposed as bearers of thought, or which are capable of bridging the alleged 'gap' between these two nominalised
'entities' ('thought' and 'language'). For example, (a) Some hold that one side
of this 'divide' consists of material processes and events, while the
other side comprises the mental or psychological concomitants of language/thought.
(b) Another view
sees mental processes linked to words
(or proto-words, or semantic structures) physically represented
(somehow) in the brain (perhaps as 'concepts', or "signs" (Voloshinov's
view)), or in some other
inchoate
form. There are of course many other possibilities, here. Indeed, while there are
parts of Parrington's article that suggest he might have favoured the second
of these options, others indicate that he might in fact have preferred the first.
Whichever set of functional inter-connections Parrington accepts, both of those
mentioned above are motivated by the inappropriate metaphors already highlighted
--, i.e., those that represent the contents of our heads as 'objects' or
'processes' of some sort, which stand in specific (if changing) relationships
with one another. So, on one side of the "gap" we might have physical
processes; on the other, 'mental' or 'psychological events'. Alternatively, one
side might consist of 'mental events' ('thoughts'), while the other is comprised
of 'internal
representations' of linguistic expressions (in the 'mind', or in 'consciousness')
etc., etc.65
Even so, wherever the boundary between these disparate entities is imagined to
lie,
and whatever supposedly falls either side of it, Parrington seems to believe
that "inner speech" can be slotted neatly into the resulting "gap".
"…[I]nner speech is the link
between thought and language…. [T]here is a gap between thought and words…inner
speech is the fluid interphase where meaning can start to be formed and
shaped…." [Ibid., p.135.]
But, the
"gap" itself appeared out of nowhere as a result of the reification of a
metaphor (which pictures thoughts as objects or
processes), and nothing more.
Well, is there anything to recommend this
(traditional) view over
and above the (inapt) metaphors and
nominalisations from which it emerged? It seems
not --. or if there is, Parrington was remarkably quiet about it. In that case, if there is
nothing to recommend this picture other than the linguistic distortions and misapplications outlined above,
maybe we should re-direct our attention to the motives of those materialists
who think there is, or should be, something that fills this "gap".
Perhaps these motives arise from a genuine desire to find a
materialist-sounding explanation for 'consciousness'? This might involve, inter alia, an attempt to go behind
the social conventions that already exist for expressing our thoughts or talking about them in order to trace their material
roots in the CNS (etc.). But, why would anyone
want to go behind social convention in order to account for human thought? It
would seem that only
the politically naive, or those with overtly anti-socialist aims and intentions, would want to do that. In
which case, have those Marxists who have toyed with these ideas been duped once again
into accepting an alien-class agenda, and the adoption of a ruling-class view,
not just of nature,
but now of the 'Mind'? It isn't easy to resist that conclusion in view of the
Idealist implications of this approach --
outlined earlier, and
again below.66
However, the problem with attempts to
go behind convention lies not so much with the ideological compromise this
introduces (on that, see Essay Twelve
Part One, and here), but with the fact that those who
venture down that path are forced to
employ words they already comprehend (as competent language-users) as
if they didn't! Or, they find they have to use words which now have to be interpreted in odd ways in order to convince themselves that
there is a
'problem' here, and hence that there is such a "gap", to
begin with.
But, this worry (and this alleged
'problem') has only arisen because of the misuse
of these very same ordinary words, again, as Marx hinted:
"The philosophers have only
to dissolve their language into the ordinary language, from which it is
abstracted, in order to recognise it, as the distorted language of the actual
world, and to realise that neither thoughts nor language in themselves form a
realm of their own, that they are only manifestations of actual life." [Marx
and Engels (1970), p.118. Bold emphases added.]
Indeed, those so minded have to persuade
themselves that there is a 'scientific problem' about the reference of
these transmogrified 'concepts', when the ordinary terms used to set the
'problem' up weren't referential to begin with. That is because these
ordinary words don't
in fact
represent anything, they merely facilitate description and communication
(based on criteria available and applicable in a public domain). In effect,
as has been pointed out in many of the Essays at this site (and as Marx
indicated), such theorists fall prey to the idea that scientific-sounding
philosophical 'problems' can be
manufactured to order by the simple expedient of
misusing ordinary language.
The upshot is that this
set of linguistic moves involves a distortion of the very medium it had been
intended to explain (i.e., language), turning the word "thought", for instance,
into a name, when it functions typically as a verb (or,
predicatively, as a
descriptive, not a referential, term). We have encountered
this dodge several times before; this is just the latest
unfortunate (but almost universally misconstrued) example.67
Hence, in order to motivate this 'scientific'
enquiry, a 'problem' had first to be created where none before existed. [And, it
is worth recalling that it
had originally been motivated by assorted
mystics and priests.] In order to do this, it had
to be shown (or, rather, it was merely hinted at and insinuated -- it was never
demonstrated) that
our ordinary words and phrases relating to our psychological lives (e.g., verbs/compound
verbs,
like "to think", "to be aware", "to understand", etc.)
were
limited, defective,
contradictory or misleading -- or, at least, that they were superficial,
non-'philosophical', pre-scientific, or they reflected 'folk
psychology'. To that end, these perfectly ordinary expressions were
torn from their usual contexts and turned into the names of metaphysical
objects, containers or processes, so that these (now) 'private objects and
processes' could be
re-located
inside our heads -- opening up the very "gap" that Traditional Theorists,
and others, then spent the next two thousand four hundred years trying
(unsuccessfully) to close, or bridge!
An alternative strategy turned ordinary words
into the names of the 'inner' psychological attributes of that all-wise, all-knowing, constantly
jabbering, surrogate in-house theatre critic (i.e., "inner speech", mentioned
earlier, which serially
explains the stream of life to each of us in terms we instantly understand,
since we are apparently too dim to comprehend such things for ourselves, unaided) -- a source we
implicitly trust, that has our inner ear at all times, and to whose own "speech" we must, and always do, attend if we are to grasp what others say to us -- but whose own
"speech" is readily comprehensible and has no need for its own 'inner, inner' intercessor.
[This is, of course, the Platonic-Christian-Cartesian
'soul' in all but name.]
In all this, it was just assumed that because we are all
familiar with these ordinary psychological words in their normal everyday contexts,
a radical change of use wouldn't affect their meaning. Either that, or it is
simply taken for granted that part of the meaning of these ordinary words -- i.e.,
whatever it is that helps us use them in normal contexts -- could be transposed
without alteration into entirely novel contexts. Plainly, the intention here was to
investigate perfectly normal phenomena
(like our ability to think) when no literal sense can be made of the novel use of languagenecessary to usher in these 'problems' -- without, of course, conjuring into existence
that
super-loquacious, 'inner' invisible
intercessor again.
In effect, language was taken on a trip -- it
went "on holiday", to paraphrase Wittgenstein.68 Words
were uprooted
and flown off on a mystery tour, dressed in outlandish -- nay garish -- clothing
so that they looked entirely alien and 'problematic'. While away on
this merry jaunt they are encouraged to do outlandish things; they became
'metaphysically drunk', as it were, which motivated all manner of strange ideas
and goings-on, and they (these words) found themselves in totally alien surroundings. This
linguistic vacation
'allowed' theorists to derive immaterialist, if not
ghostly
Idealist, conclusions
from them --
theories which (unsurprisingly) turned out to be highly conducive to a ruling-class,
and mystical,
view of the human 'condition' or 'soul'.68a
Even so, never questioned (and seldom
justified) is the spurious legitimacy that this 'linguistic mystery tour'
conferred on the metaphysical and
super-scientific
theories that were 'derived' from itover the centuries. Indeed, this is just one
more reason why ruling-class ideas become the rulingideas --
and they rule,
alas, Marxist thought, too. Few comrades even so much as question this picture! In fact, this is one
area where right-wing ideologues and Marxists share common ground (even if they
paint the same picture using different metaphors, and then re-employ them
in different ways, to different ends).
Worse still: not only did
a more recent
'bourgeois twist' give life to the 'rational economic mind', it opened up the
metaphysical space for ruling-class ideas about humanity and nature to dominate
and distort our idea of ourselves and hence our relation to the 'status quo' -- receiving spurious
'scientific' support dressed-up
these days in Neo-Darwinian finery.68b
Unfortunately, this 'virus of the
mind' -- the metaphysical approach to language
and 'mind' -- seduces far too many erstwhile materialists into thinking boss-class, immaterialist, individualistic thoughts
of their own. Not so much divide and rule, then, as: duped and ruled.69
Whatever merits the
philosophical-sounding arguments in favour of such an alien-class approach to language
and 'mind' possess, the results seem clear enough: This distortion of language
(i.e., the transmogrification of all words into names designating abstract
entities, concepts, ideas, essences, substances, processes or events) amounts to its
fetishisation -- i.e.,
the distortion of social forms of communication,
so that what had once been a result of the relation between human beings
(language) is transformed into the
real relations between things, or into those things themselves -- or,
indeed, into linguistic forms that represent these things 'in our heads'.69a
And, this is
all the more unimpressive when Marxists are co-opted to this end.70
Apart from this, there is little to
recommend the traditional approach.71
On the other hand, perhaps Parrington meant by "thought"
something pre-linguistic. [Indeed, it seems he did.] Alternatively, maybe he holds that the
"mind" is a sort of 'container' in which "mental entities" mill around, occasionally
bobbing to the surface from time to time in 'conscious' linguistic form --, or
even that they do so disguised as "images", "feelings" or "emotions"?72
Parrington's reference to
Freud's
"unconscious" (pp.139-40)
suggests that he does indeed incline in this direction. If so, that is bad news.73 The nature of the
"unconscious" is obscure, at best -- despite the totallyundeserved
fame Freud enjoys for having 'discovered' 'it'. In fact, it is now clear that
other ruling-class theorists had already concocted this notion decades, if not centuries, earlier.74
It is highly doubtful whether scientific knowledge
will be advanced much by the use of fictional notions like this -- which are themselves the
product of further linguistic distortion and fetishisation.
However, since Parrington didn't outline his ideas in this area in any great detail, little more can be said
about them.
Despite the above dead end, what sense can be
made of the following claim?
"Inner speech is the fluid
interphase where meaning can start to be formed and shaped, based on the
emotional, practical and social experience of the individual." [Parrington
(1997), pp.135-36.]
This idea is connected with
Parrington's earlier
comments on something else he found in Voloshinov:
"A related feature of inner
speech that Voloshinov pointed to was that it is more concerned with 'sense'
rather than 'meaning'. In this definition, meaning is the dictionary definition
of a word, for instance cat: 'a furry domestic quadruped'. Sense, on the other
hand, refers to the whole set of psychological events aroused by a word, such as
the personal memories of your own pet and its mannerisms, the feel of its fur
and so on. It contains activities, impressions and personal meanings, not just
accepted social definitions. A word acquires its sense from the context in which
it appears; in different contexts, it changes its sense." [Ibid.,
p.135.]
From this it looks like Parrington is
interpreting, or translating, Voloshinov's word "theme" as "sense".75
[Incidentally, in what follows, since Parrington is using "sense" in a different
way to my use of typographically the same word, in order to prevent confusion I will use
"senseRL"
to distinguish the latter from the former, but only in this section. (My use of
"sense" is explained in Essay Twelve
Part One.) I use the senseRL
here
in the following way: it expresses what we understand to be the case for the
proposition (or
indicative
sentence) in question to be true or what we understand to be the case for the
proposition in question to be false, even if we don't know whether it is
actually true or whether it is actually false, and may never do so, or wish
to do so -- the
comprehension of which allows us to understand that proposition before we know whether
it is true or whether it is false. More on that
here, too.]
Unfortunately, the above passage is rather
unclear. If the "sense" of a linguistic expression relates to aspects of an
individual's personal recollections, intimate feelings and idiosyncratic
associations, which each user brings to language, then how are we to understand
the following?
"A word acquires its sensefrom the
context in which it appears; in different contexts, it changes its sense."
[Ibid., p.135. Bold emphasis added.]
This passage tells us that a word "acquires its
sense from the context in which it appears" (emphasis added), not from
the contingent associations a speaker attaches to it.
Perhaps Parrington meant by
"context" the immediate framework of a speaker's life, her memories,
associations and 'values', as well as the social situation in which she
might find herself -- or even a 'dialectical' combination of these and other
related considerations?76
But, how does any of this relate to the public use of language, which is its
primary function? Private associations may add flavour (or "tone", as certain
Analytic Philosophers
have called it) to some of our words, but they can't affect their
linguistic meaning, as we have already
seen. Or, rather, they can no more do so than, say, a
person's idiosyncratic view of money can affect its public,
economic value.
Even so, it is reasonably clear from what
Parrington does say that for him the linguistic meaning of words is (at least
implicitly) the dominant factor here. For instance, unless the word "cat" meant
(linguistically) what it actually does mean, the correct images,
feelings, associations (etc.) wouldn't be prompted in the 'mind' of the
individual using, or hearing, that word (that is, if any are so prompted). If everyone associated what they liked
(or what their psychological make-up causes in them) with any of their words, and this was the deciding factor influencing
linguistic meaning, then the word "cat" could conjure up a dislike for fish
fingers, fond memories of the last time the individual concerned joined a strike, their hatred of
Norman Tebbit,
or, indeed, anything whatsoever.
Of course, Parrington doesn't deny this; in
fact, he says:
"A related feature of inner speech that Voloshinov pointed to was
that it is more concerned with 'sense' rather than 'meaning'. In this
definition, meaning is the dictionary definition of a word, for instance cat: 'a
furry domestic quadruped'. Sense, on the other hand, refers to the whole set
of psychological events aroused by a word, such as the personal memories of your
own pet and its mannerisms, the feel of its fur and so on. It contains
activities, impressions and personal meanings, not just accepted social
definitions. A word acquires its sense from the context in which it appears;
in different contexts, it changes its sense." [Ibid., p.135. Bold emphasis
added.]77
But, he couldn't have argued this way --
i.e., that the word "cat" conjured up the sorts of associations he mentions -- if
a particular user
failed to employ the word as the rest of us do, that is, to talk
(typically) about cats. If so, the public meaning (use) of any word must be
primary, even for Parrington (and Voloshinov). With that
observation, Parrington's entire case is completely undermined;
if public meaning (use) in fact governs 'outer' and "inner speech" (if,
that is,
the latter exists!), then contingent idiosyncratic associations must drop out of
the picture as far as linguistic and communicative meaning are concerned. The
contingent associations Parrington lists are parasitic -- or
derivative --, at best. Naturally, that is why when someone talks about
cats, for instance, what they say will readily be comprehended by anyone who
knows how to use the word "cat" without having the associations Parrington
mentions, or without ever having owned a cat.
Of course, the sheer ordinariness
of the word "cat" obscures this point. Anyone who remains unconvinced should try
arguing as Parrington does with far less common words and phrases -- such as "eggplant", or "oxbow lake" -- or,
indeed, attempt to spell out the "sense" of any
verb, preposition or conjunction in this way (and good luck with that one!).
In fact, and on the contrary, provided that prospective users understand the (linguistic) meaning of
these words (i.e., provided they know how to use them properly), no personal associations would be needed in order
to employ them successfully, or, indeed, grasp what is communicated by
means of them.
Despite this, it isn't difficult to show
that "sense"/"tone" [henceforth, S/T]
can't attach to all words, or even to words in
general -- as they appear in the public domain --, and for this to function as a
primary determinant of meaning in the way that Parrington and Voloshinov seem
to think. Here, for instance, are several words that don't possess an S/T: "and", "if", "but", "was",
"inadvertently", "sense", "tone", "word", "idiosyncratic", "theme", "meaning".
[The list is, of course, endless.]
Perhaps someone might object that
such words do possess an S/T for them; their very
mention conjures up all manner of associations and feelings. Naturally, there is
no way of refuting this contention -- or, indeed, of confirming it. And,
there is no way to determine whether or not the 'same' S/T occurs each time
they are employed by the same user, even when they appear in unusual
sentences, clauses or phrases -- like the one in the last paragraph, namely:
"Here, for instance, are several words that don't possess an S/T: 'and', 'if', 'but',
'was',
'inadvertently', 'sense', 'tone', 'word', 'idiosyncratic', 'theme', 'meaning'".
But, even if each of these words did possess an
S/T for such an
objector, the images, feelings and associations they conjured would be a result
of that objector already having understood them -- and with their usual
import --
otherwise they would fail to prompt the correct images, feelings and
associations. Indeed, if this weren't so, they could in fact induce the wrong
images, feelings and associations (if, that is, any
sense can be made of the use of "wrong" in such a context).
[Anyone
who thinks this misrepresents what Parrington is trying to say should check out
Note 78.]78
Of course, two different words would be
synonymous if they engendered the same associations. If this were possible,
"Marx" and "Hitler" could mean the same! But, would any of this affect who it
was that those two words named? "Socialism" and "fascism" could be
synonymous in the same way, too. Would a single Marxist accept that equation?
Hardly. This shows that the public, linguistic meaning of words isn't affected by the
idiosyncratic associations anyone brings to language. And that is because the contingent feelings or associations
an individual attaches to words depend on
those words being used to identify the alleged object of those feelings correctly, and
they can only do that if
they are employed in the same way that the rest of us use them.78a
Again, this isn't to deny that idiosyncratic
S/Ts might be associated with many (perhaps all of) the above words by
someone, only that this feature of our allegedly 'private' lives can't affect the public meaning of words.
Nevertheless, with respect to the idea that
there might be a 'dialectical' interplay between public meaning and private
S/Ts,
which determines the import of the words we use, consider the following
sentences:
C1: I inadvertently killed your cat.
C2: London is the Capital of
the United Kingdom.
If Parrington and Voloshinov are right, then
whatever images, feelings and associations C1
conjures up, they would clearly be specific to the present circumstances of this
Essay. That is, they would be connected with the reason why C1 was chosen -- which
was, in turn, for it to serve as an illustrative example
criticising this aspect of dialectics! But, that fact doesn't alter C1's
senseRL,
or the meaning of the words it uses. That is why we would all be able to understand C1 before we knew whether or not
it was true (or before anyone knew what the present author was or wasn't seeking to do or
accomplish by means of C1).
Of course, C1 contains several terms whose
reference is indeterminate: Whose cat? Who is the one claiming to have killed
the said cat? Which cat? When? And so on. [Naturally, C2 doesn't face any of
these problems.] Clearly, although C1 itself may well be understood, its
precise import would have to wait on the clarification of indeterminacies like
these. But, one thing it won't have to wait on is the pooled S/Ts of anyone
hearing it or reading it. The components of C1's actual senseRL are clearly unrelated to the pooled
S/Ts of its
constituent words. The S/Ts of individual words drops out of the
picture if C1 is to be understood by both originator and recipient. The senseRL
of C1 depends on the reference or use of the indeterminate terms it contains
-- and the latter are surely independent of anyone's 'feelings', 'associations'
and 'values'. So, despite what anyone feels about cats, which cat
is being referred to in C1 is independent of theirs or anyone else's
feelings. That being so, there doesn't
appear to be a hook here for any sort of 'dialectical' interplay to latch onto.
Nevertheless, even if it were still
maintained that all the words in C1 possessed their own individual, or idiosyncratic, S/Ts (which contributed 'dialectically' -- perhaps, orchestra-like
--, to give the S/T of the whole) for whomever it is that might be still
be objecting along these lines, this would still be irrelevant to the
content expressed by C1. If each speaker associated a content of their own to
each utterance (and the latter were linked to the S/Ts idiosyncratically
connected with the words used), then it plainly wouldn't be the same content
that was being entertained by their interlocutors (sheer coincidence to one side). Each would have their own set of
S/Ts which would be different from anyone else's. Including their own on
each occasion!
In which case, no shared content could
ever
be conveyed or received, and that would completely undermine the idea that
language is a social phenomenon, acting primarily as a means of communication. The fact that we
do succeed in communicating countless times each day shows that S/Ts have little,
or no, linguistic, communicative role to play.78b
Now, readers of the above words may or may
not disagree with their import, and some may continue to maintain that S/Ts (as
understood by Parrington) are central to their comprehension and use of language.
However, such individuals may do so only after acknowledging that they
will have succeeded in understanding the above contentious thoughts without having a clue what
S/Ts
their author -- RL -- either attributed, or did not attribute, to, or associated with, any or all of them. Upon doing
that, of course, such
erstwhile contrarians
would then be disagreeing with themselves, for then it would be plain
that they had grasped those words -- even while dissenting from their content --
when that act itself could only have succeeded because the meaning of the
author's words isn't
dependent on a single S/T being attributable to any or all of them.
In addition, die hard S/T
fans (if such there be) would also need to explain, for example, what Parrington himself meant by
S/T without access to his emotional state, biography or
predilections.
Of course, no one else would be able to comprehend
even that long overdue explanation
without performing the same miraculously psychic trick on the words of these die-hards contrarians themselves.
Oddly
enough, Parrington and Holborow both quote a
passage from Voloshinov's other work (on Freud) which they seem to think provides an important
insight into the entire nature of language and communication:
"How does verbal discourse in life relate to
the extraverbal situation that has engendered it? Let us analyse this matter,
using an intentionally simplified example for the purpose.
"Two people are sitting in a
room. They are both silent. Then one of them says, 'Well!' The other does not
respond.
"For us outsiders this entire
'conversation' is utterly incomprehensible. Taken in isolation, the utterance…is
empty and unintelligible. Nevertheless, this…colloquy of two persons…does make
perfect sense…." [Voloshinov (1987), p.99; quoted in Holborow (1999), p.29, and
Parrington (1997),
p.127.
Quotation marks altered to conform with the conventions adopted at this site.]79
Parrington then points out that while this communicative episode might be given
an indefinite (potentially infinite?) number of interpretations, Voloshinov was
able to reduce them to manageable
proportions by arguing that:
"…[they] must take place within the
particular space where differences in a word's meaning can be registered, namely
between two speakers in a particular social context." [Parrington (1997),
p.127.]
Undoubtedly, "Well!" could (and probably does) have a different (speaker's)
meaning to users and hearers alike on different occasions of use. But,
Parrington's response won't do. In any
social context, the word "Well!" could mean practically anything. If
we are to zero in on only one of them out of the many, we would already have to
understand our interlocutors, or risk a high probability of guessing wrong.
As we have seen above, it is central to
Voloshinov's theory that "theme" is (radically) different in each
social situation for any given utterance, as is meaning, and hence that they vary
from one
conversational interaction to the next. If we estimate that the entire human population
capable of uttering this word in their own tongues is approximately five
billion, the number of unique pairs of conversationalists selectable from this
set is roughly 1.25x1019.
[This is 125 followed by seventeen zeros!] Hence, in any given
'social situation' comprising just two people, we would have the
potential for at least that number of different meanings as either one of them uttered this word. If we now generalise across
all actual and/or conceivable 'social situations', and expand the scope this
scenario to include the many different audience sizes there are (ranging from
one to many millions) this already huge number would escalate beyond all
comprehension. Finally, if we add to this all the different words that could be
uttered in all languages, in all circumstances, the
resulting numbers would soon become unmanageable.
Unfortunately, the magnitude of even that
astronomical set of diverse 'meanings' would itself become insignificantly small if we re-introduce Voloshinov's
other vague notion (i.e., "theme") -- which is, so we are told, unique to each moment (let
alone each "social context") -- and which, according to him, supposedly determines
meaning (as far as can be ascertained, that is!).
Moreover, if we now assume that the
average conversationalist lives for approximately forty speaking years (averaged
across all populations, reduced to account for sleep, etc.), and that each 'theme-instant' lasts just one
second, then any one utterance of "Well!" by each speaker (and in the ear of
each hearer) could take on approximately 1.3x109
[13 followed by eight zeros] different additional
meanings, if said at any one of those instants. If we now recall that for
Voloshinov the microscopic details surrounding any utterance affect its "theme",
then one second would probably be far too long. Consequently, the number of
'meanings' available to the average speaker in a lifetime, while not infinite,
would be, on this view, excessively large. Naturally, this would mean that the
chances of any speaker accessing the 'correct' meaning of any of their
interlocutors' words would be vanishingly small. Voloshinov's disarming
reassurance that speakers and hearers lock onto each other's meanings isn't at
all convincing, especially since his theory leads one to suppose that no one
would ever manage to do this because of the baleful influence of
Occasionalism and "theme".80
Quite apart from this, the idea
that the consideration of one-word sentences like this warrants
conclusions about the general use of language across an entire population (and
throughout all of human history) is bizarre in the extreme! Indeed, the
fact that the above comrades based their scientific-sounding conclusions on
this one example (which is itself a laughably weak evidential base (i.e.,
it is aone-word sentence!)) is as astounding as it is alarming.80a
Nevertheless, Parrington and Holborow
failed to consider perhaps more revealing scenarios for the use of single-word
sentences, such as the following:
M1: Several comrades are on a picket
line. The Police fire tear gas. A canister is heading toward a group of pickets.
Comrade NN spots it and shouts (for the first time in his life):
"Incoming!"
Are we really expected to believe that
"incoming" is only comprehensible to one or two in this group -- maybe
only those who know the
biography, likes, dislikes and preferences of the one who shouted the warning? Or, that
only those with the requisite 'associations' will dive for cover? Do we really
have to appeal to "private meanings" to explain the subsequent scattering of
these individuals? Do these pickets have to sift through the countless likely social
settings they might or might not have encountered in the past before they hit on the correct reading of this warning, and then
proceed to act?
If the answer to these is "No", as surely it
must be, then it is safe to conclude that just as one militant doesn't make
a movement, one conversation doesn't make a theory.
Indeed, if we were to consider more complex conversations, the
completely bizarre nature of the idea under consideration
here would become even more apparent.
Now, it could be argued that this is
grossly unfair to Parrington and Holborow in that the argument above (i.e., that which depends on those
unmanageably large numbers, etc.) ignores what Parrington himself says:
"…[they] must take place within the
particular space where differences in a word's meaning can be registered, namely
between two speakers in a particular social context." [Ibid.,
p.127.]
Hence, it could be pointed out that an
interlocutor's
knowledge of the social circumstances -- these perhaps including conversational
and situational assumptions or implications about which only those party to this
conversation are aware, or which
form part of the tacit knowledge each speaker brings to any setting --
would reduce the possible interpretations of this word to manageable
proportions.
Admittedly, the fact that people do utter
one-word sentences and succeed in communicating in such circumstances does seem
to support Parrington and Holborow's case. However, since it isn't in
dispute here that acts of communication do indeed take place, this is of little
help. It is how and why conversationalists manage to do this that is
still up for
grabs.
Despite this, it
might be felt that
Voloshinov does in fact narrow down the options when he argues that a close,
even microscopic scrutiny of the word "Well!" won't help us understand this
"conversation". Voloshinov then goes on to say:
"Let us suppose that the
intonation with which this word was pronounced is known to us: indignation and
reproach moderated with a certain amount of humour. This intonation somewhat
fills the semantic void of the adverb well but still does not reveal the
meaning of the whole.
"What is it we lack, then? We
lack the 'extraverbal context' that made the word well a meaningful
locution for the listener. This extraverbal context of the utterance is
comprised of three factors: (1) the common spatial purview of the interlocutors
(the unity of the visible -- in this case, the room, the window, and so on), (2)
the interlocutors' common knowledge and understanding of the situation,
and (3) their common evaluation of that situation.
"At the time the colloquy
took place, both interlocutors looked up at the window and saw
that it had begun to snow; both knew that it was already May and that it
was high time for spring to come; finally, both were sick and tired
of the protracted winter -- they were both looking forward to spring and
both were bitterly disappointed by the late snowfall. On this 'jointly
seen' (snowflakes outside the window), 'jointly known' (the time of the year --
May) and 'unanimously evaluated' (winter wearied of, spring looked forward to)
-- on all this the utterance directly depends, all this is seized in its
actual, living import -- is its very sustenance. And yet all this remains
without verbal specification or articulation. The snowflakes remain outside the
window; the date, on the page of a calendar; the evaluation, in the psyche of
the speaker; and, nevertheless, all this is assumed in the word well."
[Voloshinov (1987), p.99. Italic emphases in the original. Quotation
marks altered to conform with the conventions adopted at this site.]
There are several points in Voloshinov's argument
that are worthy of comment:
(1)
In the above, any words that might have been spoken in the build-up to this
'conversation' were omitted. And yet, it is only on the basis of such shared
words (had he heard them) that Voloshinov would feel confident enough to tell
us that these two conversationalists had over-lapping knowledge, beliefs,
evaluations and
attitudes, as well as a joint appreciation of the surrounding circumstances of this
conversation and knowledge of one another. Now, if these two had been total strangers, the whole scene could, and
probably would have taken on a completely different complexion. Hence, Voloshinov is acting like
the author of a novel; he is supplying the reader with an almost
'god'-like view of the recent biography, thoughts, beliefs and intentions of his characters.
So, all the 'shared background details' are in
fact part of Voloshinov's imputations and assumptions, not those of these two
fictional individuals. No wonder then that he can pack so much into this one
word sentence, and into this one scene; it is his word, his scene, and his
understanding of both that is on show
here, not theirs.
(2) The above scenario was clearly tailored
to fit the purposes Voloshinov intended for it, where two interlocutors
shared much in their current surroundings and background knowledge. But, this
isn't always the case. Many conversations are between total strangers, and yet
communication is, nonetheless, typically successful. Even those that take place
between friends and acquaintances aren't always so well coordinated or tightly
constrained. In that case, very
little of substance can be inferred from this special case.
(3) Even so, given the circumstances depicted
by Voloshinov, this one word might still mean many things. All we have to do is
introduce a few more details, and what might seem to Voloshinov to be the
obvious and clear-cut implications of the use of this one word will soon become
its
opposite. So, let us assume that (a) Speaker A had heard
earlier that morning that her
daughter was going to visit that afternoon, and that (b) Speaker B was
planning on going
to the beach -- both, weather permitting. In that case, "Well!" said by
A could
(speaker's meaning) mean any one of the following: (i) "That's torn it!" My daughter will have to
cancel!", (ii) "Oh dear! I do hope my daughter will be safe driving!", (iii) "Drat!
I was so looking forward to seeing her!", (iv) "The weather forecaster is an
idiot! He predicted sunshine today!" (v) "Darn it! This means I can't do any
gardening this afternoon!", and a host of other things. All the while B could
take it to mean (i) "That's our plans out of the window!", (ii) "She
[i.e., A] means I
can't go to the beach. I'll show her!", (iii) "She wants my opinion, but what do I
know...?", (iv) "She keeps saying that! What the hell does she mean!",
and a host of other things besides.
So, apposite though Voloshinov's comments are
with respect to the surrounding circumstances in which conversationalists can be
imputed to hold some things in common -- in this special case --, this
thought experiment isn't much use in helping us understand meaning in language
in general. Naturally, that is because one word can't on its own tell us much,
whereas full sentences can. And that is partly why Frege's context principle
(introduced earlier -- on that,
see Note 24) is far more fruitful
in this respect than anything Voloshinov committed to paper.
It is also worth pointing out that Voloshinov wrote in sentences, not single
words. So, even he had little faith in his own theory!
Another dubious notion that makes its appearance in
Parrington and Holborow's work (which they both appropriated from
Voloshinov) is that all words are somehow ideologically
coloured, or compromised.81 As
Holborow puts it:
"Voloshinov's starting point is the
ideological nature of all signs, including language. He defines a sign as
that which 'represents depicts or stands for something outside itself'
(Voloshinov 1973:9). This correspondence is an essential feature of all signs….
Sign systems exist side by side with material reality, not independently of it.
"'A sign does not simply
exist as a part of reality -- it reflects and refracts another reality.
Therefore it may distort that reality or be true to it, or it may perceive it
from a special point of view…every sign is subject to the criteria of
ideological evaluation…. The domain of ideology coincides with the domain of
signs. They equate with one another. Wherever a sign is present, ideology is
present too….' (1973:10…)
"The quality of signs to represent, to
'reflect and refract
another reality', to interpret, is what gives them their conceptual potency and
makes words the very stuff of ideology…." [Holborow (1999), p.25; quoting
Voloshinov (1973), pp.9-10.]82
However, when Voloshinov says that:
"A sign does not simply exist
as a part of reality -- it reflects and refracts another reality. Therefore it
may distort that reality or be true to it, or it may perceive it from a special
point of view…", [Voloshinov (1973),
p.10.]
it isn't too clear whether he means that there
are several of these 'realities' which are "reflected or refracted" [henceforth
R/R] -- or only one 'reality' that is R/R-ed in different ways.83
But, if he were right, Voloshinov's own use of
signs could have (must have?) distorted things, too, so what his words have R/R-ed might not
be (can't be?) really real
-- if we must use this unfortunate way of expressing things.
In that case, it would seem that Voloshinov himself might have misrepresented
and/or distorted the subject of his own thesis by the use of
yet more of these inherently unreliable 'signs'. If so, Voloshinov's own words
can't be trusted to tell us the truth!
If exception is taken to this, then how might we decide
whether or not we can trust Voloshinov's words? It seems we can't, since, in
order to do so, we too
will have to use yet more of these dubious 'signs'!
And yet, if there is no way of deciding, what
sense is there to the claim that 'signs' might be misleading? Isn't that
very thought (expressed in 'signs') itself misleading? If it is, then
there is no good reason to accept
it. If it isn't, then Voloshinov is wrong anyway.
Either way, the rational thing
to do is reject this gratuitous slur on innocent 'signs'.
Alternatively, if Voloshinov is saying that
everything (both 'sign' and the things allegedly 'signified') is
capable of distortion, or of causing it, then the conclusion that there is no such thing as reality is
no less suspect, itself,
in view of the fact that we have to use yet more 'signs' to R/R that particular conclusion. But, in that
case, exactly what are our 'signs' R/R-ing? Even worse, how do we even know
there are any 'signs', to begin with? Any attempt to reassure us that
'signs' exist must itself
be expressed in 'signs', and hence must, of necessity, distort things! Indeed,
isn't the inherently unreliable 'sign' "reality" itself entirely bogus when used
in such circumstances? So, given what he says, the contrast Voloshinov wished to
draw between reality and our distorted images of it can't
in fact be drawn, for on his own account Voloshinov had to use several
untrustworthy 'signs' to make that very point. That being so, there is (for him)
no 'reality' against which anyone could compare or contrast even the mildest of
distortions with what 'signs' supposedly R/R -- nor could he truly report on them
even if there were any!
Naturally, this means no distortion can
have taken place -- at which point this theory self-destructs, once more.84
Unfortunately, Parrington and Holborow are silent on this issue. This isn't all
that surprising since any comment they might wished
to have made about
what Voloshinov could have meant (this side of their using some
form of telepathy that doesn't employ 'signs') will have
distorted what he actually had to say -- always assuming, of course, there is
such a thing as "What Voloshinov actually had to say", to begin with!
It could be objected that Voloshinov in fact
said the following:
"A sign does not simply exist
as a part of reality -- it reflects and refracts another reality. Therefore it
may distort that reality or be true to it, or it may perceive it from a special
point of view…." [Voloshinov (1973),
p.10. Bold emphasis added.]
Hence, not all 'signs' distort reality, and
that might still be true of the words, or 'signs', employed by anyone trying to report what
Voloshinov really said. So, the above comments are completely misguided.
But, how are we to tell which signs do the
one and which do the other? Which are distorting and which aren't? It seems we
can't possibly decide, since we are forced to use yet more suspect 'signs'
while trying to exonerate only some of them.
Ignoring whatever else we might think of the
fetishisation of 'signs' explicit in the above
quoted passage
(which seems to suggest that 'signs' are agents that control us!), it is
worth pursuing the above difficulties a little further.
It isn't entirely clear
what it means to suggest that every use of 'signs' is ideological -- if
this is what Voloshinov indeed meant (and our attempt to depict what he said
doesn't
thereby distort the "reality" he sought to depict about ideology
itself!).84a
Once more, if every use of 'signs' is potentially distorting, it is
difficult to see how any 'signs' could be employed to R/R reality -- or,
indeed, what the word "reality" itself could possibly mean --
or, worse, how it would be possible even to report this 'fact' accurately. If words --
operating as 'signs' -- are irredeemably ideological, then how might they be used
correctly to refer to anything at all? Presumably, this would only happen if some uses of
'signs' weren't ideological; that is, if in some
circumstances they did indeed truly R/R reality -- as opposed to merely
expressing ideological or class interests -- so that we might give some sort of
content to the supposition that on other occasions they do in fact
distort reality. But, if every use of words is ideological, then,
naturally, we can't appeal to this contrast (as already noted) -- and neither
could we trust even that assertion, for it, too, would be ideological, and
hence of suspect import (and so on).
Of course, it could be argued that ideology
doesn't distort reality, it merely inverts it. However, Voloshinov's own
metaphor (i.e., R/R itself) implies distortion, whatever else Marx meant by his
use of this word. [However, on that see Note 83
and Note 84.]
Moreover, Voloshinov himself declared that signs can distort "reality":
"A sign does not simply exist
as a part of reality -- it reflects and refracts another reality. Therefore it
may distort that reality or be true to it, or it may perceive it from a special
point of view…." [Voloshinov (1973),
p.10. Bold emphasis added.]
Finally, the idea that words are
'signs' is itself rather odd. Certainly, words can appear in signs
(for example, on a placard or badge saying "Victory to the Miners!"), and in
certain circumstances they can feature as signs themselves (as when the word
"red" might be coloured red to make a point in, say, a psychological
experiment), but words can't be signs. The
reasons for saying this are rather complex, and are outlined in Note 85.85
In addition to the above, it is also worth pointing out that language
itself can't be ideological. As I argued in Essay Four
Part One(see also
Essay Three Part Two):
Admittedly, ordinary
language may be used to express patent of falsehoods, as well as offensive,
reactionary and regressive of ideas, but it can't itself be affected by "false
consciousness" (and that isn't just because the latter notion was
foreign to Marx; on that see
here), nor can it be
"ideological".
Without doubt, everyday
sentences can express all manner of backward, racist, sexist and
ideologically-compromised notions, but this isn't
the fault of the medium in which these are expressed, any more than it is the
fault of, say, a computer if it is used to post racist bile on a web page.
Ideologically-contaminated ideas expressed in ordinary language result either
from its misuse or from the employment of specialised vocabularies borrowed from
religious dogma, sexist beliefs, reactionary ideology, homophobic bigotry, racist theories
or
superstitious ideas. This isn't to suggest that ordinary humans don't,
or can't, speak in such backward ways; but this is dependent on the latter
being expressed in ordinary language, while it isn't dependent on that language as
such. That particular claim might sound paradoxical, so I will attempt to
clarify what is meant by it.
First of all, this defence of
ordinary language isn't being advanced dogmatically. Every user of the
vernacular knows it to be true since they know that for each and every sexist,
racist and ideologically-compromised sentence expressible in ordinary language
there exists its negation.
This is why socialists can
assert such things as: "Blacks aren't inferior"; "Human
beings aren't selfish"; "Wages
aren't fair", "Women aren't sex objects",
"Belief in the after-life is baseless", "LGBTQ individuals aren't
perverts" -- and still be understood,
even by those still in thrall to these ideas but who might hold the
opposite view. If ordinary language were identical with 'commonsense' --
and if it were ideological (per se) in the way that some imagine -- you
just couldn't say such things. We all know this to be true -- certainly,
socialists should know this --, because in our practice we manage to deny such things every day.
So, as noted above, while
ordinary language might be used to express patent of falsehoods, as well as
offensive, reactionary and regressive of ideas -- and, in order to express such
ideas, reactionary, racist, sexist or homophobic individuals might depend on
ordinary language in order to give voice to their vile, or their anti-socialist,
opinions, the fact that socialist can reject all such ideas, using the very
same medium, means that the vernacular as such can't itself be associated with those
ideas.
In which case, it is odd that
socialists don't advance the opposite claim: because we can with
relative ease explain socialist ideas in the vernacular -- just as we can
challenge the regressive ideas mentioned above -- ordinary language is
inherently progressive. Now, I'm not promoting that idea myself,
merely asking why socialists are quite so quick to malign, or depreciate, the language of the
working class, and assume that because there are regressive ideas expressible in
the vernacular that this automatically condemns it, while at the same time they
ignore their own use of the vernacular to propagandise and agitate the working
class. [On this, see
Grant (n.d).]
In this regard, it is as
ironic as it is inexcusable that there are revolutionaries who, while they are
only too ready to regale us with the alleged limitations of ordinary language --
on the grounds that it reflects "commodity fetishism", "false consciousness" or
"formal/static thinking" --, are quite happy to accept (in whole or in part)
impenetrably obscure ideas lifted from the work of a card-carrying, ruling-class
hack like Hegel. Not only are
his theories based on alienated thought-forms (i.e., mystical Christianity and
Hermeticism),
his Absolute Idealism was a direct result of the systematic fetishisation of language -- indeed,
as Marx noted:
"Feuerbach's
great achievement is.... The proof that philosophy is nothing else but
religion rendered into thought and expounded by thought, i.e., another form
and manner of existence of the estrangement [alienation -- RL] of the essence of man; hence equally
to be condemned...." [Marx
(1975b), p.381. I have used the on-line version, here. Bold emphases
and link added.]
Finally, despite
what Voloshinov says about those who "ignore
theme", they are probably well advised to continue doing just that -- that
is, should we
ever be told what wtf "theme" actually is!
If the (linguistic) meanings of words were in general
dependent upon their intended sentential use we wouldn't be able to
comprehend the use of ostensibly familiar words in new surroundings,
until, that is, we had first apprehended the intentions of those who uttered
them, or the use to which they wanted to put that sentence. A recent and rather
comical example might help illustrate this point. A few months ago, The
Sex-Pest-In-Chief, Donald Trump,
had this to say on Twitter:
Despite the constant negative covfefe
This
nearly broke Twitter and the Internet as hundreds of thousands of
individuals tried to figure out what "covfefe" meant, as well as providing
ample ammunition for countless
Internet wags to poke fun at this easy target.
Here are just a few:
Figures Two, Three, Four,
And Five: Covfefe Fun
Figures Six, Seven, Eight
And Nine: More Covfefe Fun
Thousands of people tried to work out what
Trumplestiltskin meant, but all of them concentrated on what he might have
intended to type in place of "covfefe", the most likely candidate being
"coverage". No one to my knowledge argued that once they had worked out The
Liar-in-Chief's intentions, "covfefe" could be counted as a genuine word in
the language. If intentions decided meaning, they would surely have argued this.
No, they suggested an everyday substitute word that would have allowed The
Racist-in-Chief to express his intentions, and that guess was based on the
orthographic similarities between "coverage" and "covfefe",
as well as the fact that
"coverage" fitted the drift of the sentence itself and the things this
Coiffured Carpetbagger had tweeted before:
Despite the constant negative coverage
Since
The Minority President tweets at night, and he is now clearly well in his dotage, he obviously fell asleep in
mid sentence.
What speakers say (coupled perhaps with their overt behaviour) is often our best guide to what they intend
to say, not the other way round. Naturally, their intentions can affect the way
we then try to make sense of why they might have said what they did
(i.e., their speaker's meaning), but they can't in general alter what the words
they employ mean in the language. This doesn't of course
imply that the intentions of speakers are unrelated to their choice of words, or
how they use them (manifestly, their aims and intentions will largely govern
why they chose the
words they did, and what they wanted to achieve by using them, etc.). It does
mean, however, that the sense of any sentences they utter and the meaning of the
words they employ aren't in general so dependent. In fact, far from intention
determining (linguistic) meaning, or even sentential sense, the reverse is if anything the case; the latter
shapes the former not the other way round. We can only form the intentions we do because of our
socialisation and our facility with language. This can be seen by the way we
actually use language to speak about intentions (etc.). [On this see Anscombe
(2000), Hursthouse (2000), Kenny (1973a, 1975, 1992, 2003), and Teichmann (2008).]
Indeed, if intentions
could affect sense and meaning in general, then, in such circumstances,
a speaker's words would in fact be unrelated to his/her past use of 'similar' words,
or even to their accepted meanings (except fortuitously). A new extra-linguistic
context would thus define a new 'meaning'. In that case, any words at all
would suffice (and we would all use 'words' like "covfefe"). If context, intention and the use of sentences (and only
these) determined meaning, then a 'word' could be given any (linguistic) meaning whatsoever
by an intention to so use it on just that occasion. But, in such circumstances
even the meaning of the word "word" would begin to lose its grip. Naturally, this would rule out all
communication between speakers, since there would be nothing in the past use of
a word (or sentence) that a hearer could latch on to, to assist in the
comprehension of what was being said on this new occasion. Interlocutors
would confront each other like speakers of foreign languages -- only in this
case, they would be employing similar sounding words that now possessed
unknown meanings. [Except that, in this case, we couldn't even say this much!]
Moreover, if the use of whole sentences
determined the meaning of their constituent words, then the inner structure of sentences would be irrelevant; word order and grammar would be unnecessary.
Furthermore, inferences drawn between sentences would become problematic -- unless,
that is, the unit of meaning was extended to sets of sentences. But even
then, this would only compound the problem.
In addition, the concept we now have of the
misuse of a word would be undermined, since on this basis no one could
misuse a word if there was an intention of some sort underlying its actual
employment (and that in turn would compromise one of Wittgenstein's major criticisms of
Metaphysics). There would then be no such thins as a
malapropism or even a
Spoonerism.
In fact, it is impossible
even to spell-out the
details of this idea (i.e., that sentence use and intention can affect the
meaning of words). The reasons for asserting this are pretty clear. Suppose someone
were to say:
C1: "The leaflets advertising
the meeting have arrived."
But, 'intended' to mean:
C2: "I think you should start
handing them out."
If this were a general feature of
the use of language, then, as noted
above, the intended 'meaning' itself (i.e., C2) would also be
impossible to state, for it too would 'mean' something else. On the other
hand, if it
didn't 'mean' something else, this theory would fall at the first hurdle,
since there would be at least one sentence whose meaning wasn't sensitive to the intentions of its
user, namely C2. Naturally, this wouldn't be difficult to generalise across all
such sentences, which,
once done,
would mean that if, per impossible, this theory were true, it wouldn't
be possible to say what it implied -- since the linguistic expression of any
supposed intention would be subject to the very same equivocation. [I am, of
course, blurring the distinction between the sense of indicative
sentences/propositions and the meaning of words, here. That is because those
who adopt the Occasionalist view of language blur it, too. As do many other
Analytic Philosophers! In fact, many reject
this distinction out-of-hand. This adds further complications to their theory that I won't enter
into here.]
Conversely, if in general the words found in C2 meant exactly what they say
(and they did so without the need for any further paraphrase, translation or
consideration of the supposed intentions involved) there would be no good reason
why the same couldn't be true of C1.
Compare that with the following:
C3: "The arrived have
leaflets the for meeting advertised."
If the 'intended meaning' of C3 was still:
C2: "I think you should start
handing them out",
then C3 could just as well be uttered instead
of C1, and no one would be puzzled or would scratch their heads.
And, if we are fully consistent, there is no
reason why C3 itself couldn't be seen as 'intending' the following
jumbled up sentence:
C4: "I out think handing you
them start should."
Or, indeed, anything whatsoever.
If sentence structure were susceptible to the sort of radical reconstructive
surgery we see in C3 (because of the
doctrine that intentions determine meaning, which would obviate the need for any
sort of syntax or grammar), then this would also apply to
intentions themselves -- or at least to their linguistic expression -- as illustrated in C4.
If intentions were sufficient to determine meaning, who needs a settled grammar
or syntax? On the other hand, if grammar and syntax are integral to our use of
language, what space is there left for intentions to decide on anything other
than speaker's
meaning?
Of
course, where people utter odd sentences (or, the language in question isn't
their first language and they are struggling to make themselves understood), an
educated guess concerning the intentions of the speaker (what those are or might be) will assist in making sense of what they are
trying to say. But, even here, intentions won't affect what any of the
words they use actually mean.
The situation is even worse if word misuse is now thrown in:
C5: The boy stood between the
lamppost (sic).
C6: It square-rooted each
other's onion slippers (sic).
C7: Many years ago, my cat
is flying off-side in tennis tomorrow (sic).
Putting to one side whether or not these
could be
coded messages of some sort, it isn't easy to connect the radical misuse of words these sentences
exhibit with any
set of conceivable intentions (except, perhaps, those displayed by individuals
who might want to create confusion,
surprise or consternation, etc.) -- the linguistic expression of which
intentions couldn't itself contain an equally bizarre use of words (like those found in
C4). On the contrary, the linguistic expression of an intention would have to be
impeccable in all respects if we were to grasp, or even express, that intention --
unlike C4. And, if that is so, why can't it be true of C1?
C4: "I out think handing you
them start should."
C1: "The leaflets advertising
the meeting have arrived."
Coded messages to one side again, in order to
see this more clearly the reader is invited to try to use any one of C5-C7 above
to 'intend' to mean either or both of the following:
C8: The strike begins on
Monday week.
C9: Orange is darker than
black but not as sweet as vinegar.
[C5: The boy stood between the
lamppost (sic).
C6: It square-rooted each
other's onion slippers (sic).
C7: Many years ago, my cat
is flying off-side in tennis tomorrow (sic).]
As will readily be apparent, this isn't
possible. That alone shows intentions take their lead from our
(rule-governed) use of language, not the other way round.
Indeed, if language were in general dependent
on such bizarre 'translations', words like "intention", "occasion", "meaning"
and "word" would begin to lose their grip -- again, as noted above --
which is just another reason why this theory can't consistently be stated.
The obvious objection here would be to point
out that no one would use such odd sentences. But why not? What constraints are
there on the sorts of intentions we can form if not those already dictated to us by our
socialisation and the rule-governed use of language, by means of which this is
accomplished?
To be sure, deliberate errors over
syntax and grammar -- in addition to the innovative use of vocabulary and sentence structure
-- can extend a language. Manifestly, this often happens
in literature and in the course of the social evolution of discourse, etc. --
in fact, Voloshinov himself gives a rather good example of this on pp.55-56 of
his book. But, this can only take place if it is based on currently shared
word use, and on a locally universal syntax/grammar. This must be so if total
incomprehension is to be avoided. That is partly why my use of "BBB XXX ZZZ QQQ TTT" (in V7,
from
earlier) didn't
amount to a literary event of any great moment -- whatever intentions
lay behind it.
This point can be illustrated by considering
an example of the idiosyncratic use of language that occurs annually in a
school where a supporter of this site teaches, which I am sure happens
elsewhere. There are regular charitable events all year; one of these involves
the pupils each paying £1 so that they may attend school casually dressed --
i.e., without their uniforms for that day, but in clothes of their own
choosing (this sort of thing is called 'Dress-Down/Casual
Friday' in the USA). The staff can also pay whatever they feel is
appropriate so that they, too, can wear whatever clothes they want on the same
day. A collection box in the staff room has a notice on it that reads: "Put your
own clothes money in here", which on the face of it doesn't
seem to make much sense. Whether you pay or not you are presumably going to wear
your own clothes! And no one is going to put the money they have already used to
buy those clothes or which they intend to use to buy clothes in the future (i.e., the money they used weeks or months ago,
or might use next week, to purchase their "own clothes") in the box --
"their own clothes money".
Plainly, the
circumstances surrounding the use of this sentence changes the meaning (or
interpretation) of some or all of the words it contains. Here: "own clothes" is short
for something like: "your own choice of clothes" (but even this is misleading,
since whatever is selected will represent a choice of clothes by the wearer),
or, perhaps, "clothes not required by your contract of employment or code of
conduct" (etc.). Now, this is a clear case of the use of a sentence in
particular surroundings where the ordinary meaning of some of the words it
contains isn't ascertainable
solely from their past employment. Having said that, the novel use of a
sentence like this is stillrelated to the established meaning of
its constituent terms -- otherwise the notice on the box could have read "Place
a dead kipper on the computer's maiden aunt" (with the meaning of each of these
words having no connection with their past use, either!), while supposedly
'meaning' the 'same' as the actual sentence used -- or even "Covfefe your
clothes money!".
Consider another example: in football/soccer games, commentators will often say things like "Saved by the woodwork"
when a ball hits a goal post or cross bar, and bounces out. But, they can't mean this, since if
there had been no post or cross bar in the way of the ball, there couldn't have been a goal (or a game),
to begin with! What these commentators clearly mean is something like "That was
a near miss!", or "What a stroke of luck!" So here, the usual
meaning of these words coupled with the rules of football/soccer aren't a sure
guide to what was intended. However, having said that, these commentators
plainly wouldn't say in these circumstances "Sliced into the rough,
again!" or even "Who put that post there!?" The irony of "Saved by the
woodwork!" is understood by all who know football/soccer, which is why these
words seem so apposite. This is despite the fact thatthe speaker's intended meaning
fails to align with what his/her words actually say, and yet, no one supposes
that this odd use of language changes what words like "woodwork" and "saved" in fact mean
-- otherwise they wouldn't have been used.
Update 07/04/2020: During the ongoing
Coronavirus pandemic we have begun to hear things like this (about cancelled
football/soccer games, for example):
"No ball has been kicked in anger in the
football-mad country since March 9 because of an outbreak that bore down on
Italy from China last month." [Quoted from
here; accessed 07/04/2020.]
No
one understands this literally, either -- i.e., that football can only be
played, or is normally engaged in by footballers who are fuming or who are resentful.
Plainly, the above individual meant that no ball had been kicked in a
competitive game for a month. Again, having said that, no one at all would
understand this odd use of language if, instead of the above, we had read the
following (assuming the intentions of the individual concerned were the same and
no coded messages were implied):
"No
narwhal has
been filtered in vinegar in this crocus-mad bottle of beer...."
But,
if intentions determined meaning the individual who came out with the first of
above could just as well have used the second, and we would all know what he was
trying to say.
Finally, sometimes we will hear one individual say to another "You know where
the door is", and we all know what they mean -- i.e., "Please leave!" No one
takes this to be an attempt to find out if the second individual does in fact
know where that door is (even though that presumed fact can be inferred from the use of
this
sentence, otherwise it wouldn't have been uttered. The first person assumes the
second isn't new to the building). If that were the case (i.e., if this were an
enquiry into the second person's knowledge of the whereabouts of the said door),
the sentence would have been a question "You know where the door is?", signalled
by a slight raise in the voice at the end of uttering it (called "uptalk").
But, it is the use of just these words (in what is now a clichéd
sentence) that allows us to infer the first speaker's intentions, not the
other way round -- otherwise she could have said "You know where my cat is",
instead.86
While Voloshinov's book is ostensively about
the Philosophy of Language, its main aim seems to be to advance the Science of Linguistics.87 However, by pitching his work in both
camps, Voloshinov managed to replicate many of the metaphysical confusions that
have crippled 'Materialist Dialectics' -- chief among which is the complete absence of any
evidence substantiating his many bold claims. Indeed, Voloshinov's book has
little else in it other than a priori assertions and dogmatic theories about meaning,
understanding, evaluative import, "theme", "consciousness", and so on.87a
Hence, far from resembling a work of science, Voloshinov's book reads more like
Traditional
Philosophical Dogma. For instance, there are none of the following (which one
would expect
to find in a scientific report, book or paper): experimental detail coupled with graphs, tables, diagrams, photographs, charts,
spread sheets, primary data, mathematical analysis, etc., etc.
Naturally, this means that Voloshinov's work
succeeds in doing what Traditional Philosophy has always done -- that is, it
confuses
a priori thesis-mongering with science itself.
[I examine some of the 'evidence' others have offered
in support of Voloshinov's theory, below.]
On a related topic, despite the fact that
most of what Parrington and Holborow say undermines the role that
language plays in communication -- reinforcing the view that language serves to 'represent' things
to us in our heads (even if this process is filtered through our own
idiosyncrasies, attitudes, social situation, prevailing ideologies, etc., etc.) --, they
appear to believe that human beings developed language because of a "need to
communicate". This is how Holborow puts it:
"The genesis of language is in human labour….
Communication is not therefore just one of the functions of language; on the
contrary, language presupposes both logically and de facto the
interaction among people. Language only arises from the need to communicate with
other humans. It is quintessentially social." [Holborow (1999), p.20.]
Parrington concurs:
"Crucially labour...developed within a co-operative and social
context. It was this that led, through the need to communicate while engaging in
co-operative labour, to the rise of the second specifically human attribute --
language." [Parrington (1997),
p.122.]88
As far as I am
aware, this was an idea first explored by Rousseau:
"As soon as one man was recognised by
another as a sentient, thinking being similar to himself, the desire or
need to communicate his sentiments and thoughts made him seek the means to do so.
These means can only be drawn from the senses, the only instruments by which one
man can act upon another. Hence the institution of sensible signs to express
thought. The inventors of language did not make this argument, but instinct
suggested its conclusion to them." [Rousseau, Essay On The Origin Of
Languages, quoted in
Knight (2010a). Bold emphasis added. Readers will no doubt note that
Rousseau clearly thinks that human beings could think before they could talk.]
While
I don't wish to question the role that co-operative labour
has played in the development of language and thought (quite the
opposite, in fact), several aspects of these two quotations seem highly dubious, especially the idea that human beings
invented language because of a "need to communicate". To be sure, we use
language to communicate, but the claim that it arose because of a
specific need to communicate is highly questionable -- that is, except forLamarckians.
Of
course, the word "need" is ambiguous itself. We use it in a variety of different
ways. Consider just a few:
N1: That cake needs more sugar.
N2: This strike needs to be widened.
N3: You need to put
oil in your engine.
N4: We need a pay rise.
N5: The giraffe needs a long neck
to browse tall trees.
N6: That drunk needs to go home.
N7: Plants need water.
N8: The state needs to
be smashed and the ruling-class needs overthrowing.
N9: Tony Blair and
George W Bush need
prosecuting as war criminals.
N10: Comrades need to shout louder
on paper sales.89
Precisely which of the above senses of "need"
these two comrades intended is unclear -- several of them relate to what can only be called felt needs, or conscious needs (e.g.,
N4, and possibly N2), expressed perhaps as part of an agent's aims, goals and
intentions. Others refer to the causal concomitants or prerequisites of a
flourishing organism, successful revolution, strike, comeuppance for Bush and Blair,
a paper sale, or a well-run engine -- all of which are largely, if not totally,
unfelt. Some, of course, can't be felt.
Nevertheless, it is patently obvious that
human beings couldn't have invented language as a result of a felt
"need to communicate" (unless, that is, we assume they could think before they developed language -- which
idea would naturally imply that thought
isn't a
social phenomenon, dependent on collective labour), since such a need would presuppose the very thing it
was
aimed at explaining. The idea that this type of necessity mothered
that sort of invention would imply that the first human beings to talk had
earlier formed the thought: "I/We need to communicate" (or something
equivalent in their
assumed
proto-language). Clearly, such a felt need to
communicate could only be expressed if language already existed. On the
other hand, if the thought (or its equivalent) that supposedly
motivated the "need to communicate" wasn't in fact linguistic, then little
content can be given to the notion
that human beings once possessed such a need without being able to give voice to it.
Indeed, how would it be possible to form the thought "We need to communicate" if
the individual, or individuals, concerned had no idea (as yet) what communication
was. That would be like arguing that we can (now) form the thought "We need to
schommunicate" when none of has a clue what "schommunicate" means. In fact, it is
worse than this, since we are already sophisticated language users and can
not only conceive of certain possibilities we can give expression to them.
They weren't.
It could be objected that
such a need might be a biological one (analogous to that which is expressed in,
say, N5). There are two problems with this response. First, reference to
the biological needs of organisms to explain the origin of adaptation is Lamarckian,
not Darwinian. Secondly, and far worse, this alternative completely
undermines the view that language is a social phenomenon.89a
In reply to this it could be argued that
revolutionaries have in fact given a Darwinian (but not Lamarckian)
explanation of the origin of language. A relatively recent article written by
Chris Harman, for
example,
demonstrated that such an account of human development -- augmented with ideas
drawn from Engels's work -- provides Marxists with an adequate, materialist
theory of the origin of language and culture, and one that was founded on a
"need to communicate".89b
Unfortunately, there are serious
problems
with Harman's explanation of the origins of speech. For example, after outlining
the increasing dependence that our human ancestors had on social organisation
and the use of tools, he argued as follows:
"Natural selection would bring about…evolution
in the direction of ever larger, denser and more complex neural networks,
capable of directing and learning from intricate motor functions of the hand and
of using minute changes in gesture or voice to communicate….
"A cumulative process would soon have been
underway in which survival depended on culture, and the ability to partake in
culture [based?] upon a genetic endowment that encouraged the combination of
sociability, communication, dexterity and reasoning power….
"The development of labour and the
development of communication thus, necessarily, go hand in hand. And as they
both develop, they both encourage the selection of those new genes
which made people more adept at both: the more agile the hand, the larger
the brain, the [larger the] larynx that made a wider range of sounds.
"Such developments do not involve just
quantitative changes. As the growth of labour, the growth of sociability and
the growth of language reinforced each other, encouraging the
selection of a whole range of new genes, new networks of nerve cells would
emerge in the brain, making possible whole new ranges of interaction
between people and the world around them….
"So there has to be a recognition of how
quantity turns into quality, of how through successive changes animal life
gave birth to that new form of life we call 'human', which has a dynamic of its
own, shaped by its labour and its culture not by its genes…." [Harman (1994),
pp.100-02. Bold emphases added.]
There are several highly dubious things Harman says
here about which I will comment later, but for present purposes I will simply draw
attention to his use of DM to defend,
or buttress, his argument. It is abundantly clear that Harman relies heavily on Engels's first 'Law' (i.e., Q«Q)
to plug a gap he thinks he has spotted in standard Neo-Darwinian theory in this
area.89c This allows him to
smuggle into his own account an inappropriate but revealing teleological
slant (indicated by most of the words and phrases highlighted in bold in the
above passage). Of course, it could be argued in his defence that these
supposedly "teleological" expressions are metaphorical, or they are merely rhetorical flourishes;
but if that is the case, and they are replaced by more 'neutral' terms, Harman's
account
falls apart alarmingly quickly. Why that is so will now be explained.
The problem with the highlighted parts of the
above passage is that they suggest that evolution has a goal, one that has
already been decidedupon by the operation of Engels's first 'Law'
-- i.e., when
the latter has been coupled with natural selection. In order to see this, compare it with a similar
but far less complicated example: the formation of ice or steam. Just as
a sufficient quantity of heat will change water into steam, Harman's use of this 'Law' suggests that the
accumulation of small quantitative changes (in the genetic code, the
development of the CNS,
social organisation, etc., etc.) would
automatically produce 'consciousness' and culture.90
Given Q«Q,
and the water/steam analogy, the outcome of evolution appears to have been written
into the fabric of the universe from the very beginning (that is, it was stitched into the laws
that apparently govern everything -- laws which clearly
include Engels's Q«Q).
That is why Harman's account makes it seem as if this 'Law' -- which also
'determines' the inevitability of water boiling when heated sufficiently -- must have 'determined' the ineluctable development of language and thought.
Apparently, so this story seems to go, gradual increases in the complexity of the
CNS (etc.), linked to,
and emergent from, the development of
collective labour, guaranteed that 'thought' would 'emerge' at some point
in the life of our ancestors, as quantitative changes in their biological
makeup and social organisation "passed over" into qualitative changes in their 'minds' (etc.). This is
the only explanation there can be for Harman's cavalier use of words like: "natural selection
would bring about", and "they both encourage", which clearly suggest agency
in nature. Otherwise, why use such terms?91
It could be objected to this that: (i) DM-theorists
don't claim that such developments were/are "determined", or that there
is anything inevitable about the whole process, and that (ii) DM-theorists
insist there is a dialectical interplay
between an evolving organism (or population of organisms) and its environment --
leading, in this particular case, to the development of 'consciousness' and language.
The second of these volunteered responses
(i.e., (ii)) is by now
well-worn, even threadbare. When faced with what others see as a genuine problem in their
theory, dialecticians almost invariably refer us to the "dialectical interplay"
between this or that object or process, neglecting to give the details. Naturally, this works for
them as a handy 'get-out-of-a-theoretical-hole-free' card, in ways reminiscent
of the use of the word "miracle" in 'Creation
Science', which is no less dishonest for all that.92
In support of
(i) above it could be
argued that the development of
language was in fact dependent on countless contingent events, so it
can't have been inevitable, contrary to an earlier allegation. Now, this would have been an effective response had Harman himself not already holed it well below the waterline with the
following comment:
"…[I]n fact, everything is not
'contingent'.
In certain conditions, both in the biological world and in history, certain
things are likely to happen…." [Harman
(1994), p188, n.73.]
With the best will in the world, it isn't
easy to see how Harman's rejection of universal contingency could in any way
be supported by his claim that some events "are likely to
happen", since part of what we mean when we use the word "contingent" is
that the events so described are "likely to happen" (or otherwise, as
circumstances dictate), not that
they must occur and are hence "inevitable".
Contingent events are those that are neither necessary nor impossible; so
they range from the highly unlikely to the highly likely. In that case, anything that
is "likely to happen" already counts as contingent! Exactly how the
above words substantiate Harman's
rejection
of contingency is, therefore, something of a mystery. If so, his use of terms
that are practically synonymous with "contingency" doesn't in any way help the reader
understand how his
dismissal of contingency can succeed. Indeed, if "everything is not
contingent", then some things must be necessary, and therefore 'determined'!92a
In fact, Harman needs a "must happen" here to
counterpose his rejection of contingency -- i.e., something like the
following:
"...[E]verything is not 'contingent'. In
certain conditions, both in the biological world and in history, certain things
must happen." [Harman (1994), p.188, n.73, deliberately misquoted!]
Admittedly, the inclusion of that particular
phrase comes with a price tag attached: the use of "must" would expose the teleological
nature of the whole argument. This modal qualifier ("must") clearly implies the operation
of some sort of will, direction, intelligence, logical consequence, necessity or purpose in nature. Hence, the
open presence of this sort of claim:
C1: Natural selection must in
the long run bring about language and consciousness,
would be a dead give-away. Even though Harman
evidently requires (and clearly assumes) the truth of C1, it is obvious that he
couldn't risk using a "must" here for fear of undermining Darwinism -- which,
by the way, many still
think removed teleology from nature, or at least from our depiction of it,
when the opposite is the case.93 In the end, what
Harman actually opts for is much stronger
than a mere "will happen". However, the problem is that since he has
already denied contingency in the above passage, he, for one, can't assert that language developed as
the
result of a series of contingent events, without contradicting himself.
In fact, as was asserted above, Harman's
rejection of contingency harmonises almost seamlessly with the necessitating force
underlying Engels's first 'Law' (i.e., Q«Q). So, just as water
has no 'choice' but to turn into
steam at 100oC (this
change is entirely 'determined' by antecedent events, or so the story goes),
similar concomitants must have necessitated the origin of language (even if they are more 'dialectically' complex, in this case). Harman might not like to use such words
himself, but they are nonetheless an apt summary of his position.94
In order to confirm the accuracy of this
'revisionary' interpretation of Harman's reasoning, we need look no further than
several other things
he
says. Near the beginning of his article, we find this dismissal of the
many "just so"
stories that supposedly litter Neo-Darwinian writings (an epithet Harman clearly
borrowed from
Stephen Jay Gould):
"The sparsity of reliable information
makes it very easy for people to make elaborate, unsubstantiated conjectures
about what might have happened, with no facts to confirm or deny them -- the
modern version of the 'Just so' stories
Rudyard Kipling
wrote for children
nearly a century ago. All sorts of writers on human evolution make hypotheses of
the form, 'And, so, perhaps, we can explain the descent of certain apes from the
trees by their need to do X'. Within a couple of paragraphs, the 'perhaps' has
gone, and X becomes the origin of humanity.
"This method is the special hallmark of
sociobiologists…. It is a
method Marxists have to reject." [Harman (1994), p.87.]95
But, unless there has been a misprint here, Harman's rejection of
these "likely stories" (which were, and often still are, triggered by an appeal
to what researchers perceive to be the needs
of organisms, even if that is hidden beneath several layers of inappropriate
metaphor and analogy; more on this below)
-- his rejection must also count as a repudiation of Engels's appeal to the "need to
communicate". In a pre-linguistic group, the assertion that there was a felt
"need to communicate", which led to the development of language, would itself be
just another "just so story", but with added DM-spin. On the other
hand, if the said need weren't a felt need, it would represent a damaging concession to
Lamarckism.96
However, even if this weren't the
case, Harman is fooling himself if he thinks that sophisticated
Neo-Darwinians
and sociobiologists attempt to explain the development of life in such crude
quasi-Lamarckian terms. Little wonder then that he quoted so few references in support of this allegation.97
If Marxists are to confront successfully the
arguments of knowledgeable
Sociobiologists and
Evolutionary Psychologists,
something far less
insubstantial than Engels's first 'Law' will need to be wheeled out of the Dialectical
Dungeon. Unfortunately, this 'Law' seems to be the only
'solid' premise Harman has available to him to prevent his ideas sliding back
into (at least) this area of the crude sociobiologist camp (with its own "just so" tales
--
Gottlieb (2012)) -- the idea that distinctly human behaviour traits somehow
"emerged" against a background of increasing material complexity, but which can't be reduced to it,
being one such. [More on this in Essay Three Parts Three and Five (not yet published).]
Ironically then, because Harman buttressed
his account of human origins with his own "just-as-Hegel-and-Engels-say-so-story", he
ended up tail-ending the fabulous tales concocted by sociobiologists, which he
also rightly castigated. To
compound matters, Harman pointedly failed to substantiate this part of
his story with any evidence of note (nor did he address the fatal weaknesses
that afflict Engels's shaky Q«Q
'Law' -- detailed in Essay Seven
Part One).
We can see this more clearly if we examine
how Harman justified the following 'leap' in his argument:
"Only if you see things in this way
can you explain why our species was already endowed with the capacities 35,000
years ago to develop a whole new range of technologies." [Harman (1994),
p.100.
Bold emphasis added.]
Apart from a dire warning that the
consequence of not seeing "things in this way" risks slipping Marxist
theory back into an idealist, "postmodernist", mechanical-materialist
and/or
sociobiological
swamp, this is all Harman
had to offer in support of his own distinct DM-ideas.
Now, while the
Dennetts,
Cronins,
Pinkers, and
Dawkins
of this world mightn't object to much of the secondary evidence Harman
marshalled in support of his account of human origins, they would
surely take exception to the use to which he puts it. This evidence
is in fact consistent with much of what those theorists would argue anyway (particularly all
that material about genes). However, Harman presents us with no new facts from
Psychology, Anthropology, Anatomy, Physiology, Linguistics, or any other branch
of science, for that matter, to substantiate his own "just so story"
--
that 'consciousness' is an "emergent property" of matter.97a Or, indeed, that there
is such a thing as 'consciousness' to begin with (so that it is capable
of "emerging" from anywhere -- on that, see Note 65). Or, that there are
such things as "emergent properties" (that aren't themselves dependent on a quirky misuse of language, or,
which are based on speculative forms of science fiction). Or, that language itself is genetically-based.
Or, that Darwinian change can account for it. Or even that DM has
anything useful to add to our knowledge in this area -- or, indeed, anywhere else,
for that matter.
So, if Harman is to be believed, the only
thing that dialecticians can offer in order to counter theories that are
inimical to Dialectical Marxism (in this respect, at least) is a way of "seeing things" -- albeit augmented
by the convenient ability DM-fans have of being able to "grasp" a-contradiction-a-day
(mentioned earlier),
"emergent" no doubt from the quantitative repetition of Engels's rather
shaky first 'Law'.
Now, any response to the above that
is itself based on a
further appeal to Engels's first 'Law' would be to no avail. That 'Law'
can't bear the weight constantly put upon it by DM-theorists; it is certainly
incapable of countering the detailed arguments that sophisticated sociobiologists have
constructed in support their own ideas. Waving
it
about as some sort of talisman does Marxism no favours -- especially when we
discover that this 'Law' is fundamentally
flawed, to begin with.
The harsh words Harman reserved for
Chris Knight (a reference to whose work provided the only support
for his contention that sociobiologists depend on "just so stories") --
whether deserved or not -- might well now be flipped over and directed back at his own account: by
resting the whole credibility of this area of Marxism on such
wafer-thin foundations, he invites not just disbelief, but easy refutation.
If you are going to take on sophisticated
anti-Marxist theories with little more than an appeal to Engels's first 'Law',
and a hope others will "see things" your way, all the while lambasting
them for their reliance on myth, story and fable, excoriating them for their lack of
corroborating evidence, deprecating the supposed reactionary consequences of
their ideas, speculating about how they simply reflect the "mood of the times",
it isn't a good idea to do so with an account that is contradiction-friendly
itself, overtly Lamarckian, supported by little or no evidence, remarkably badly-stated and
reliant on a few fairy-tales of its own.
The wise course of action here would be
to admit that we just do not know how language developed, and neither does anyone
else -- and we will probably never know. But, that doesn't
mean we have to accept
Adaptationist or
Nativist
accounts of its origin just because the majority of theorists apparently
do, and neither should we make the slightest concession to their ideas.
In fact, it is disconcerting to
see how many of the latter Harman is prepared to take on board, adapting to
other reactionary Neo-Darwinian ideas in this area along the way.98
This is despite the fact that such theories have little to recommend them beyond
an excessive of
metaphor, tailor-made and ideologically-biased mathematical models, convoluted teleological
language, and wild extrapolations from an impoverished evidential base.99
"The genesis of language is in human labour….
Communication is not therefore just one of the functions of language; on the
contrary, language presupposes both logically and de facto the
interaction among people. Language only arises from the need to communicate with
other humans. It is quintessentially social." [Holborow (1999), p.20.]
"Crucially labour…developed within a
co-operative and social context. It was this that led, through the need to
communicate while engaging in co-operative labour, to the rise of the second
specifically human attribute -- language." [Parrington (1997),
p.122.]
Earlier, we saw that Parrington and Holborow
had simply reproduced Engels's comment about the origin of language, that it arose as a result of a
"need to communicate". Oddly enough, these two failed to quote the following words of
Engels's -- a
quirky passage that is often overlooked by those who regard him as a great
philosopher, or even a profound theorist of science:
"Comparison with animals proves that this
explanation of the origin of language from and in the labour process is the only
correct one. The little that even the most highly-developed animals need to
communicate to each other does not require articulate speech. In a state of
nature, no animal feels handicapped by its inability to speak or to understand
human speech. It is quite different when it has been tamed by man. The dog and
the horse, by association with man, have developed such a good ear for
articulate speech that they easily understand any language within their range of
concept (sic)…. Anyone who has had much to do with such animals will hardly be
able to escape the conviction that in many cases they now feel their
inability to speak as a defect…. Let no one object that the parrot does not
understand what it says…. [W]ithin the limits of its range of concepts it can
also learn to understand what it is saying. Teach a parrot swear words in such a
way that it gets an idea of their meaning…; tease it and you will soon discover
that it knows how to use its swear words just as correctly as a Berlin
costermonger. The same is true of begging for titbits." [Engels (1876),
pp.356-57.]
Contrary to what Engels asserts, we shouldn't want to concede that animals understand our use of language (or,
indeed, that they grasp the import of swear words, for instance) simply
because parrots, for example, can make certain sounds -- or, just because some humans
are a tad too sentimental and believe that their pet dog can "understand
every word they say". If understanding were attributable to animals solely
on the basis of vocalisation, then we might have to admit that, for example, the
ability most of us have of repeating foreign words upon hearing them means that
we too understand the language from whence they came, when quite often we don't. For example, although I can read both Hebrew and Greek, I actually
understand very few words of either language.
But, even
when we repeat foreign words we don't understand we would still
be viewing them from our standpoint as sophisticated users of our own
language, which means that the dice have already been heavily loaded (so to speak) in our
favour. Because of this, we often make an educated guess concerning the meaning
of any new (foreign) words we might encounter, based on knowledge of our own
language. Moreover, we do this against a background of shared behaviour and a common
culture that links us -- directly or indirectly, closely or remotely -- with all
other human beings. The same can't be said of parrots,
dogs and horses.
We should, I think, only want to count
someone (or something) as having understood what is said if it possessed a sufficiently detailed verbal and behavioural repertoire,
at the very least. If, for example, such a 'proto-linguist' couldn't form
new sentences from their 'vocabulary', if they were incapable of forming the
negation of any of their sentences (for example replying "No, it isn't raining"
when told it is raining), or couldn't cope with word-order change, if they
were unable to refer to anything proximate to, or remote from, their
immediate surroundings, if they couldn't identify or specify any of the
implications of what they had said, or of what was said to them, if they were
incapable of reasoning (hypothetically) both with truths and
falsehoods (in the latter case, for example, with an "If that's true, I'm a
monkey's uncle!"), failed to appreciate stories or fiction, if they couldn't
respond to humour, or engage in self-criticism, if they were regularly perplexed
by new sentences they had never encountered before (even those that contained
'words' drawn from their own repertoire), if they couldn't follow or give instructions, and so on, then I
think most of us would have serious doubts about their
capacity to understand the target language.100
On the other hand, had Engels said the
following to one of his parrots: "Swearing isn't allowed here because it
represents the language of oppression" (to paraphrase Trotsky) -- and the
parrot had stopped swearing as a result, or, maybe, had deliberately sworn even
more as a result! -- we might
be a little more
impressed with his claims.
Despite this, Engels's ideas don't seem to
hang together even on their own terms. If language and understanding
are the products of social development (augmented by co-operative
labour), then they most certainly do not hang together. Indeed, Engels even says:
"Comparison with animals
proves that this explanation of the origin of language from and in the labour
process is the only correct one.... First labour, after it and then with
it speech -- these were the two most essential stimuli under the influence of
which the brain of the ape gradually changed into that of man...."
[Engels (1876),
pp.356-57. Paragraphs merged.]
If so,at a minimum, how could an
animal comprehend our speech without also having gone through the same social
development and engaged in the same sort of collective labour with human beings?
It could be argued that animals have,
and still do work alongside human beings. Think of the phrase "work
horse", or the use to which dogs are put in guarding, sledging, hunting and
the herding of sheep, to say nothing of the work done by oxen, donkeys, camels
and pigeons, to name but a few. However, without wishing to minimise the use human beings have
made of animal labour (etc.), this hardly counts as collective labour (any more than the
use of some wood in a building counts as collective labour contributed by
a tree); it more closely resembles the use of living tools. The
differences between human and animal labour don't need to be itemised to see that
this line-of-defence won't work. Which Marxist wants to argue that an ox, for
example, shows any desire to communicate, or that a donkey or a pigeon shows any
sign it wants to verbalise its aims and intentions? But, if their efforts counted as
collective labour, we should be prepared to argue that these animals do
indeed show signs of a "need to communicate".
Moreover, Engels appears to think (somewhat
inconsistently) that mere proximity to human
beings is sufficient to engender (in certain animals) the "need to communicate".
If this were so, then manifestly an ability to use language can't have been the result of
collective labour. Surely, in humans (on Engels's own admission) the
"need to communicate" arose out of collective labour, not from mere
association. By way of contrast, in the passage above, Engels seems to think that this "need to
communicate" is a free-floating force when it comes to animal behaviour,
which can somehow be divorced from its connection with cooperative human labour,
and hence is capable of crossing the species boundary.
This explains
why he also appears to believe that mere association with human beings is
able to create
such a "need" in these animals, too. To be sure, the behaviour of domestic animals is
different from the behaviour of animals belonging to the same (or similar) species in
the wild, but if mere
proximity to human beings could account for language, then we should expect
cats, cows, donkeys, camels, oxen, sheep, goats, rats, mice, gerbils, fleas,
lice and bacteria to be able to
communicate with us, or with one another (to say nothing of viruses).101
Conversely, if animals were able to talk
or understand us then language can't be a
social phenomenon, or the result of co-operative labour.
It looks, therefore, that Parrington, Holborow and Harman (among others) have ignored this glaring
inconsistency in Engels's account.
However, in wanting to deny that there is a
significant gulf between humans and our closest relatives among the Apes, or
Ape-like ancestors, Engels
and Harman were clearly laying a foundation for their own theory of descent --
i.e., a theory based on the idea that a change in quantity leads to
a change in quality (and vice versa).101a
In order for this
'Law' to work, DM-theorists would have to argue that the important differences
between human beings and certain animals is merely quantitative -- even if it is finally
expressed qualitatively via this 'Law'. On this view, the gap between ancient humanoids and apes (or our common ancestors), say, would be somewhat analogous to that between two closely
related elements in the
Periodic Table
(except, of course, with respect to evolutionary descent, the situation is far more complex). So, given
this analogy, when one chemical element supposedly acquires a few more
elementary particles, "qualitatively" new properties automatically arise in the
elements so formed. The latter could then be said to "emerge" from the former as
the increased complexity exceeds a certain "nodal
point".
[However, as we saw in Essay Seven
Part One, this 'Law'
doesn't even apply to the elements in the Periodic Table, which removes one
of the best and most over-used 'illustrations' of this 'Law' that DM-fans' have
in their box of tricks.]
Indeed, a belief in the continuity of nature
seems to require a similar commitment to the idea that there is some sort of
'dialectical connection' between, say, our ape-like ancestor (or proto-human, or
humanoid, group, before the development of language, etc.) and modern human
beings (after
language had been acquired).
The idea appears to be that even though apes are biologically close to us, the
gradual increase in our ancestors' social and physical evolution in the end led to the
development of the profound
qualitative differences between humans and the aforementioned ancestors, culminating
in the 'emergence' of 'consciousness' and language, etc. Hence, Engels's claim
that certain animals are capable of understanding language looks as if it lends
support to the belief that some sort of continuity exists between modern humans
and our ape-like past, mediated by subsequent material and social
progress.
This seems to be the
only conceivable reason why Engels alluded to
parrots, dogs and horses in this way.
The only problem is that he left out
the Apes!
Plainly,
Engels chose the wrong animals to illustrate his point (if this was
his point). As should seem
obvious, no sane biologist would
want to argue that we are biologically closer to parrots, dogs and horses than we are to
the Apes.
Even worse, the latter aren't
widely known for their verbal skills (unlike parrots); in fact, they are
quite
incapable of vocalising
words. And yet, if the view outlined above were
correct, we should find Apes vastly exceeding the 'linguistic' production of parrots.
Hence, the "qualitative change" that is supposed to have "emerged" as a result
of increased "quantitative" evolutionary development must, it seems, have taken an unplanned "qualitative" detour
via birds, horses and dogs, outflanking our nearest relatives the Apes! Clearly,
this means that "quantity" doesn't in fact "pass over" into
"quality", but skips it sometimes, or shimmies past it. Either that, or it indicates that parrots and
other birds
(the Hill Myna, for example), as well as dogs and horses, somehow managed to defy this
dialectical 'Law' --, or, indeed, (if we absolutely insist on clinging to
this part of DM come what may), that these
species are evolutionarily closer to us than the Apes!
Comrades are oddly
silent on this issue.102 They are, however, free to
"grasp" what
little comfort they can from it.
Even though the nature of primate
'language' (and the question whether it is only those animals that possess 'language'
which are
capable of 'thought') will not be entered into here in any detail, a few comments
are clearly in order.103
A close reading of the writings of
those involved in research into animal 'language' reveals that rather too many authors
conflate several different senses of the word "communicate", and it is
this that
makes some of their conclusions seem initially plausible. Indeed, as seems
plain, these theorists find they have to
anthropomorphise the noises and signals animals make in order to get this part
of their story off the ground. The rest of the picture is then sold to us by the
by-now-familiar ploy of using ordinary words in odd ways in order to sanction
these seemingly innovative conclusions, the entire ensemble further motivated by the use
of inappropriate metaphors,
'educated' guesses and no little sentimentality.104
The word "communicate" can, of course,
mean anything from "to connect" (as in "communicating door"),
to
"to convey
information", "to achieve mutual understanding", "to share thoughts
and feelings", and "to bond socially" -- and, of course, simply
"to converse".
Naturally, certain senses of this word are closely linked to our nature as
social and political agents. While animals appear to be able to 'communicate' by means of
various calls, gestures, signs, smells and noises, calling this "communication" in
any of the above senses is clearly prejudicial. In fact, it is hardly more illuminating
than the claim that since certain rooms are connected by communicating doors,
rooms can signal to one another and share thoughts via the said door! Even less
persuasive is the idea that the
mere presence of signs indicates the presence of thought -- unless, that is, we
are also prepared to concede that the weather can think, too, since thunder is
often a sign of rain. Again, the careless (if not, thoughtless)
misinterpretation of the figurative use of language to depict animal
'communication' has misled many of those working in this area.105
Moreover, our comprehension of animal 'sign' systems is neither helped nor advanced
by an egregious distortion/misinterpretation of the conventional, logical and social features of our own language.
Nevertheless, one thing is reasonably
clear: in the absence of human intervention (that is, without socially-structured input from us), primate 'communication', for example, would be
seriously limited. Manifestly, it is we who train apes to respond to us, not the other
way round -- the same can be said of parrots and other
conversationally-challenged animals. Even this limited concession indicates that
linguistic ability is socially-, not genetically-based.106
Naturally, this brings us to a
contentious issue mentioned earlier: whether language is a social or a 'genetic' phenomenon. As pointed out
above, the danger
with the strategy Engels adopted is that it actually undermines
belief in the social nature of language.107
It also deflects, and even silences, the
hostility revolutionaries have generally shown toward biologically reductive
theories of humanity -- for instance,
those promoted by
Social Darwinists,
Ethologists,
Sociobiologists,
"Evolutionary Psychologists",
and, of course,
by
fascist/Nazi pseudo-scientists.107a
This might help explain why we
find prominent revolutionaries endorsing opinions that are
compatible with the very worst forms of biological and
genetic determinism. Hence, we have
comrades like
Alex Callinicos heaping praise on
Daniel Dennett's reactionary work (and in the pages of International Socialism, too!), joining forces with an attack on
Stephen Jay Gould,
all the while referring
to
Pinker's fashionable
(but nonetheless regressive) book on language -- and maintaining this
line even when such serious lapses were pointed out to him.108 In
addition, we find comrades like Chris Harman endorsing Chomsky's
Nativist theory of language -- also beloved of
characters like Pinker109
-- referring his readers to the work of a prominent sociobiologist to
support his case against Gould!110
With such a display of 'socialist thought',
what remains of the implacable opposition revolutionaries have generally
shown toward racism, sexism and elitism? To be sure, the above comrades rightly
abhor the use of Darwin's ideas to justify each and every such regressive doctrine
(as, for example, Harman does in
Harman (1994), pp.88-90, and n.39, p.186) -- and they constantly remind us
of their total opposition to such aberrations. However, other things they say
only succeed in undermining this otherwise laudable stance -- i.e., by adopting (or by being
sympathetic toward) the ideas of reactionary authors such as those mentioned
above.
Without doubt, ideas drawn
from DM have seriously compromised their commitment to HM.
Plainly, these comrades have made unwise concessions to ideologically-motivated
theories based on extreme forms of Neo-Darwinism. Once again, this is no
surprise since the concepts that litter DM enjoy an
impeccable ruling-class pedigree
themselves.
[HM = Historical
Materialism/Materialist; CNS = Central Nervous System.]
Ideological compromise
like this is the natural
bed-fellow of one or more of the following:
(1) Metaphysical theories of mind that
were aimed at 'solving' a series of pseudo-problems, which werethemselves motivated by the
ancient idea that
each individual (or, rather, each shadowy alter ego inside every skull --, i.e., the 'soul',
or, these days, 'intelligent' algorithms) can be viewed as a socially-isolated
unit that invents (or rather, re-invents) its own idiosyncratic language with its own private meanings,
acting just like a possessive individual
of bourgeois lore
-- a fable concocted by early modern philosophers
like Descartes, Hobbes
and Locke.
(2) The idea that language is based on an
internal grammar genetically-programmed into the CNS, which is independent of social forces,
social development or cooperative labour. [In its modern form, this theory was invented by bourgeois rationalists
(Descartes and
Leibniz),
but made 'respectable' these days by Noam Chomsky.]
(3) The
adoption of representational theories of mind, language and knowledge -- a family of doctrines that stretches
right back to Ancient Greece.
The acceptance of one or more of the
above anti-communitarian notions by DM-theorists means it is little wonder that
they can't explain language and 'consciousness' --, which means they have also
failed to account
for the connection either of these has with our social development. Because of this, DM-theorists find they have to refer to Engels's dubious
first 'Law' to account for the supposed "emergence" of this mysterious entity,
'consciousness',
from increasingly complex arrangements of matter -- not pausing for one second
to notice that this way of viewing things only becomes possible
if our individual and social characteristics are projected back onto nature -- and
then into each head -- in order to try to sell this idea both to themselves and the rest of us. This then succeeds in compromising their
theories of 'mind' and language, since they are now forced to concede ground to
reactionary Adaptationist and Geneticist models of the origins of one or both, as well as
give ground to ruling-class representational
theories of both, too. Once more, DM forces its
adherents to compromise and play into the hands of the ideologues of the class enemy.
This also explains why each of the comrades
mentioned above becomes uncharacteristically vague, evasive and defensive when
trying to explain how (i) social development could conceivably find
itself represented in, or be imprinted upon, our genes, or, conversely, how (ii) genetic change could possibly affect co-ordinated social behaviour across the
whole of humanity. But, if the first alternative is impossible to achieve
(as seems to be the case), then the development of language can't be accounted
for in Darwinian terms. If, on the other hand, the second option is also
non-viable (as it must be if language is, so we are told, "innate", and
genetically 'programmed' into each speaker), Linguistics can't be
reconciled with Marxism.
This main
sub-section is aimed at countering the responses DM-theorists tend to offer in
relation to (i) and (ii) above.
At this point it could be objected that
Harman did in fact outline a plausible, dialectical explanation of the origin
of language, culture and thought based on sound Darwinian
principles, and which is also consistent with Marxist materialism -- in the
article discussed earlier.
Nevertheless, as we will soon see, and despite what Harman says, his approach suggests that the origin of language was in fact an abrupt and semi-miraculous
event, somewhat
reminiscent of the way that ecstatic Pentecostals
suddenly begin "speaking in tongues" (i.e.,
Glossolalia).
[More on this presently.]
However, returning to Harman's theory, when the relevant sections of his
article are examined it soon becomes apparent that it contains what can only be
described as a mere gesture at giving a 'dialectical' account of the origin of language --
and,
once more, one that was 'substantiated' by several of his own "just so"
stories.
Harman's earlier theory -- as we
will see, he appears to have changed his mind -- was
based largely on Engels's The Part Played by Labour in the Transition from
Ape to Man, and The Origin of Private Property, the Family and the State
[i.e., Engels (1876, 1891)], updated with more recent work carried out in this
area. Engels saw the move made by our ape-like ancestors from an
arboreal to a more terrestrial
existence as a crucial development which allowed them to free their hands to
work with tools, thus facilitating their progression to more complex social
forms.
This was accompanied by parallel developments in tool-making, and therefore of the brain;
and that in turn
allowed proto-humans to invent language. Central to all this was
cooperative labour. Harman continues:
"It is this
which also explains the development of those most peculiarly human attributes,
language and consciousness. The distinctive feature about human language, as
opposed to the sounds and gestures made by other animals, is that we use words
to refer to things and situations that are not actually present in front of us.
We use them to abstract from the reality that confronts us and to describe other
realities. And once we can do this to others, we can also do it to ourselves,
using the 'inner speech' that goes on inside our heads to envisage new
situations and new goals. The ability to do these things cannot have arisen
at one go. It must have grown up over many generations as our remote ancestors
learnt in practice, through labour, to abstract from and to change immediate
reality -- as they began to use sounds and gestures not merely to indicate what
was immediately in front of them or what they immediately desired (which is what
some animals do) but to indicate how they wanted to change something and how
they wanted others to help them. In tool use we know there was a significant
change from the ape to the early humans: the ape picks up a stick or stone to
use as a tool; the early humans of 2 million years ago were already not only
shaping the stick or stone, but using other stones to do the shaping, and,
undoubtedly, learning from each other how to do this. This implies not merely
conceptions about immediate things (food stuffs), but about things once removed
from immediacy (the tool that can get the food stuff) and twice removed from
immediate reality (the tool that can shape the tool that gets the food stuff).
And it also implies communication, whether by gesture or sound, about things two
stages removed from immediate conditions -- in effect, the first use of abstract
nouns, adjectives and verbs. The development of labour and the development of
communication thus, necessarily, go hand in hand. And as they both develop, they
both encourage the selection of those new genes which made people more adept at
both: the more agile hand, the larger brain, the larynx that made a wider range
of sounds.
"Such
developments do not involve just quantitative changes. As the growth of labour,
the growth of sociability and the growth of language reinforced each other,
encouraging the selection of a whole range of new genes, new networks of nerve
cells would emerge in the brain, making possible whole new ranges of interaction
between people and the world around them. This may well explain why suddenly new
species of humans developed that lived alongside and then superseded those that
went before, as with the successive emergence of
homo habilis,
of
homo
erectus, of the various sorts of archaic human. Thus, it may well be the
case that modern humans eventually replaced the
Neanderthals
because they were able to communicate more quickly and clearly with each other
(although we will probably never know for certain if this was so).
"So there has
to be a recognition of how quantity turns into quality, of how through
successive changes animal life gave birth to that new form of life we call
'human', which had a dynamic of its own, shaped by its labour and its culture
not by its genes...." [Harman
(1994), pp.101-02. Bold emphases and links added.]
While
much of the above won't be
questioned here (or anywhere else, for that matter), we have already seen that
"inner speech" is far too vague and fanciful a notion to be of
any use.
Likewise, we have also seen (in Essay Three Parts
One and
Two) that the idea that
"abstraction" underpins language is no less misguided. In addition, we have
seen that Engels's first 'Law'
is no help at all in any attempt to understand evolution, or, indeed,
anything at all.
What will be questioned here,
however, is the theory, which Harman later endorsed, that language is
genetically-based [Harman (1999), p.621],
since that idea is
incompatible with his earlier stated commitment to the social nature
of language.110a
In this section, some attempt will be made to examine the
viability of Harman's 'later view' of the origin and nature of language:
"The ability to use language
is, according to the generally accepted theory of Noam Chomsky, a genetically
determined feature of all modern humans. The connection between language,
abstraction and human consciousness is spelt out in the books written by the
Russian Marxist Voloshinov during the 1920s, and in part two, Labour, of the
Ontology by the Hungarian Marxist
Georg Lukács."
[Harman (2008), p.621, note 6. Harman has made a slight error here; Labour is part three of Lukács's work.
Bold emphasis added, italic emphases in the original; link also added.]
At the outset it
should be stressed that the
comments below don't pretend to be an exhaustive critique of
the countless theories there are that attempt to explain the origin of language, nor are they aimed at
even a brief survey of current theory in this area.
[Indeed, that would require several books on its own!] This material is aimed
solely at Harman's comments, and those of Chomsky's, detailed later.111
This, I think, represents a first stab at
plausible scenario depicting the transition from a proto-linguistic community
that only used what we would now call nouns, to one that used nouns and what we
would now call verbs:
[1] Imagine, say, a
proto-linguist, NN,
at time, t1
(i.e., X thousand years ago), who has in her linguistic repertoire a limited set of 'words', W1. Suppose
this set contains only 'proto-nouns' used to 'name' things.112
Suppose also that at time, t2
(Y thousand years later) another
proto-linguist, MM (who perhaps belongs to the descendants of the group
to which NN belonged), has in her repertoire a wider set of 'words', W2.
Further, let W2 contain
'nouns'/'proto-nouns' and 'verbs'/'proto-verbs', which are used to say things. Finally, let MM
be the first to innovate in her group.113
A linguistic development like this
could
be the result of one or more of the following: (i) A series of genetic changes
in the individual concerned or in her group,
(ii) A pattern of cultural development, or (iii) A
'dialectical' combination of the
two -- given Harman's theory.
Consider the first alternative (the third
is clearly dependent on the first -- again, assuming the truth of the theory under consideration
--
while the second won't be open to question here): let us
suppose that a gene or set of genes, G1, is
responsible (a) in whole, or (b) in part for this innovation. Given either
of these sub-options, the presence of G1
would be wholly or partly
responsible for the transformation that took the group(s) to which NN and
MM belong from the use of W1 to the use of
W2
(or which enabled that development to take place).
Again, consider the first of these two
sub-options:
[1a] Let G1 be wholly responsible for this
development. In that case, G1 would clearly enable whoever
possessed it (i.e., in this case, MM) to use the wider repertoire W2.
Unfortunately, however, unless several speakers in their group
simultaneously possessed G1, MM would have no one to
talk to who was capable of understanding her, andno one from
whom she could have learnt this new language, or these new 'words'. In such circumstances, MM
would be either self-taught, or a spontaneous master of this wider
vocabulary. Neither option is credible, and both undermine the social nature
of language.
Even supposing MMwere
self-taught (or was a spontaneous master of this proto-language), no one else
would be able to understand her
unless the same were true of the rest of the group. In which case, there would be
no survival value
either for MM, or her group, resulting from this development. As
Christina Behme notes:
"Furthermore, the result of
the one-time
Merge mutation would have to have been exceptionably stable
(allegedly Merge did not change from the moment it appeared in one lucky hominid
(called by Chomsky (2005b) Prometheus), some 50,000-100,000years ago), and would have had to result
immediately in massive selectional advantages for Prometheus. According to
Chomsky (2009b,
2014b), it is foolish to speculate about the role of communication when
attempting to account for language evolution. This seems to imply that, whatever
advantage was conferred on Prometheus, he could not have communicated his novel
cognitive powers to other members of the breeding group." [Behme (2014b), p.677.
Referencing conventions altered to conform with those adopted at this site. It is
worth recalling that
Mary Shelley's
novel, Frankenstein, was subtitled The Modern Prometheus. Links
added.]
Of
course, these problems will only be multiplied
if, as Chomsky imagines, language (in its entirety) appeared
all at once.114
If we now weaken an earlier restriction
placed on
this option and suppose that several members of MM's group
possess G1 (and hence,
W2) at the same time.
Little in fact changes. Whenever it was that G1
first appeared in the population it would have had to have taken place all at once in more than one
speaker at the same time, otherwise the linguistic innovations of one member of the group would
be useless, and would be incapable of being learned or comprehended by others. Hence, unless we
suppose there to have been a simultaneous group mutation, this isn't a credible
scenario, either. Quite apart from the fact that widespread simultaneous
group mutations (of exactly the same sort) aren't plausible (and, as far as I'm aware, have never been
observed in nature), this option also
undermines the idea that language is a social phenomenon.114a0
This would mean that language is neither socially acquired nor learnt by group
interaction based on cooperative labour. In this particular case, language
couldn't have been taught toMM and
all the
rest by her/their parents, relatives or members of her/their local tribe or group. That is because this
proto-language was assumed to have been
genetically initiated; hence, ex hypothesi, MM, or her generation, would
have been the
very first to innovate.114a
Even if we now suppose that there had
been a group mutation of some sort resulting in members MM's linguistic
community allinnovating together, they would all have to demonstrate roughly
comparable expertise with the same novel vocabulary all at once or they
would fail to communicate, once more. [Why that is an important consideration
here is connected with the alleged survival value of mutations like this,
which, if there is no pay-off in this respect, would be zero.] But, how might this remarkable coincidence have come about? Unless we suppose
that there is a close connection between words and genes -- so that, for example,
where there was a gene (or set of genes), g1, there would be a word
(or set of words), w1
--, but, the presence in each individual of the sameunlearnt vocabulary is as difficult to account for as it is implausible.
On the other hand, if this
group mutation was also responsible for simultaneous grammatical (and not
just lexical) innovation, the difficulties become even more formidable. Not only
would the language of such innovators be incomprehensible to non-innovating
parents, relatives and members of the same community (who don't share these genes), we would be faced with a
scenario where these novice innovators would (as it were) all wake up one
morning as
expert
speakers in this new tongue.115 They would all demonstrate a
semi-miraculous facility with these new verbs
(never having heard or used them before), knowing (implicitly) that they weren't nouns -- this
remarkable capacity revealed
by their use of them as non-nouns, non-naming words. [I am not
suggesting they were intuitive grammarians, here! I am merely adverting to the
novel way they would have to use such 'words'.] Such a scenario might feature in
Hollywood B-movies, but only the most credulous of film buffs would buy it.
It
could be objected to the above that it badly misconceives the origin of language
since it supposes it to have been developed suddenly, which is absurd. The
genetic changes that occurred in our ancestors needn't have been expressed linguistically at first, but could have
developed as a result of other selection pressures, or random changes.
In that case, let us suppose that G1gradually formed in the said tribe, and further that this gene (or set of
genes) wasn't immediately responsible for these hypothetical linguistic
innovations. Instead, let us surmise that G1 initially assisted in the survival of members
of the group in some other (as yet unknown) way, and was preserved in the gene pool for
that reason.116
At some point, of course, the
exaptative
-- or perhaps
"pre-adaptive"
-- proto-linguistic properties of G1 would have
to have kicked-in, in order for G1
to assume its linguistic role in the target group.
Hence, G1
would have had to become operative in this novel capacity at a future date -- in response, say, to
new selection pressures -- allowing it to take up its innovative role, assuming functions not necessarily
related to those that had given rise to its original preservation in the
gene pool. In that
case, although G1 might appear in a population as a gene (or
set of genes) selected for other reasons, subsequently, it could, under certain circumstances,
facilitate the development of a proto-language. There are many
examples of this sort of pre-adaptation (or exaption) in nature, or so we are told.117
However, as Aitchison points out:
"But this by-product view is
highly unlikely, as language is too complex. Exaptation -- a re-use of an
existing structure -- is undoubtedly a powerful force in evolution. But in all
documented cases, complex structures are used for simple purposes, and not vice
versa. A type of wading bird uses its wings as a sun shade: there is no evidence
of any bird using what was originally a sunshade as wings. You can use a
television as a paperweight, but you can't use a paperweight as a television.
The complexity of language, and the interwoven adaptations of the mouth, larynx
and brain make it unlikely that language developed as an accidental by-product."
[Aitchison (1996), pp.74-75. Bold emphases added.]
Nevertheless, if we ignore for now this unhelpful fact, even given the
revised scenario, we
would have to suppose that several members of the group who possessed G1
would have to innovate linguistically at the same time, and in exactly
the same way, as the
pre-adaptive (or exaptative) properties of G1 kicked-in. If
this weren't the case, no one would be able to understand anyone
else in that group who began to speak in this novel way. A loneinnovator -- whatever the genetic pre-dispositions of the group happened to
be, and howsoever quickly or slowly they became manifest -- would suddenly find
themselves uttering strange unrecognisable sounds -- somewhat like the
aforementioned
ecstatic Pentecostals -- to a
group of bemused onlookers. This still fails to be a
credible picture.117a
Even so, as noted above, the idea that all or
most of those possessing G1
would innovate together is no less implausible. But, even if they had
innovated together, an entire throng
of 'talkers' would face each
other, not as lone a Pentecostal ecstatic, but as a dissonant gaggle of babblers
all suddenly barking strange noises at one another.
Of course, the possession of a new set of words (or noises, as these will
have initially presented themselves) doesn't imply that the individual (or individuals) concerned knew
how to use them. Compare that with those who hear foreign words now. Do they
know how to use a strange vocabulary upon hearing it? But, they alreadyhave
a language; these proto-linguists haven't.
Even worse, there would be no one in this
proto-linguistic community who could assist or guide them in this respect --
unlike today, when novel or even foreign words are first encountered. Moreover,
the possession of rules for the use of such terms, which failed to specify what
was or what was not their correct application, would, naturally, imply that they weren't
rules to begin with.117a0
Furthermore, members of this
group of innovators would either have been (a) Born with this gene (or set of genes)
already in
place, or they would have (b) Acquired it as a result of a mutation later in life.
Either way, plainly, they wouldn't have been taught, or socialised
into, this new proto-language by
their parents, siblings or carers,118 even though we are now to suppose
that they could
all use this new tongue in the same way, with the same grammar and
vocabulary, and would all comprehend one another in the same way, at the same time,
right from the get-go.
We have to suppose
that all this, or something like it, did in fact take place otherwise an appeal to
biological, genetic or physiological principles to account for linguistic
innovation would be an empty gesture.118a
That being so, none of these pre-historic,
novice innovators would need to be schooled by anyone in the proper use of their new
vocabulary with its novel grammar; the employment of these ground-breaking linguistic
abilities would spring forth in
each individual spontaneously, and all in the same way, at the same time. And, we would have to suppose this
were so no matter how basic or rudimentary this extension to the old 'dialect' (of
'nouns') proved to
be, or how gradually it was introduced. Every party to these new social norms
should have to innovate together at roughly the same rate (or the proposed links
with the biological origins of linguistic novelty would begin to weaken, once
more) --, even
though, as they proceeded to do this, they would fail to comprehend each other's new sounds,
words, or grammar. [Why that is so will be explained presently.]
Hence, these
innovations would still be useless,
and wouldn't be preserved.
Moreover, if we examine the original
postulate, these innovations were said to be of a specific sort.
Earlier it was
hypothesised that there once was:
...a proto-linguist, NN,
at time, t1
(i.e., X thousand years ago), who has in her linguistic repertoire a limited set of 'words', W1. Suppose
this set contains only 'proto-nouns' used to 'name' things.
Suppose also that at time, t2
(Y thousand years later) another
proto-linguist, MM (who perhaps belongs to the descendants of the group
to which NN belonged), has in her repertoire a wider set of 'words', W2.
Further, let W2 contain
'nouns'/'proto-nouns' and 'verbs'/'proto-verbs', which are used to say things. Finally, let MM be
the first to innovate in her group.
The question is: How would a member of
MM's group recognise a verb asa verb if they had never heard, used or
encountered one before, let alone know from then on how to use one? Since we are dealing with
human beings (or proto-humans) here, not
gene-driven automata, this consideration is no meredetail. Naturally, we
should only want to say that such proto-linguists were using verbs if they
employed them in the same way that we now use them to form, for example,
indicative sentences (among other things). And how might they have done that if they
don't yet know
(in practice, not in theory!) the difference between these two grammatical forms -- that is, between nouns and
verbs? If they don't in fact appreciate this difference (once more, in practice, not in
theory), what justification is there for us now saying that they must
have understood what they were communicating, or even that they knew how to
go about doing it? What possible reason could
there be for concluding that a new noise uttered (or employed in a novel way) for
the very first time in history is indeed a verb if it
wasn't then used, even in a rudimentary way, as we use verbs today? In fact, if there is no good reason,
then this science fiction account of the origin of speech can't even get off the
ground.
Of course, as noted above, the innovators
involved in this scenario needn't
have consciously grasped the import of what they were doing, nor need
they have had any idea at all about grammar (and, in this case, plainly, they
couldn't possess such knowledge). But, in practice, if they
couldn't use these new forms in the way that we now do, even if only in a
highly simplified manner, there is no good
reason to characterise their innovative sounds as part of this novel grammatical
category, i.e., as verbs. And since we are now trying to theorise about the steps that might
have led up to
the invention of language by human beings (and, once more,
if we
avoid deliberating
over the introduction of a new set of noises produced by automata), we
have no other way of conceptualising this series of moves. Hence, given what
we currently mean by the use of verbs, we now have no good reason to
suppose that such innovators could use these new forms -- if the process of
innovation is characterised in this way.119
[Why this is so was explained earlier;
more details are given below.]
On the other hand, even if these
innovative moves were plausible, radical developments like this couldn't facilitate communication.
Members of this pre-historic band of novice innovators would confront each other
with sounds they wouldn't recognise; nor would they know what to do
with them. Recall that at some point in history they would be the very first
individuals to use language in this new way on the planet; in that case, they would comprehend one another
no better than ecstatic Pentecostals do now.120
Simply mumbling, or barking, sounds at other members of the group doesn't amount to
communication. [On that, see Note 117a0 and Note 118a.]
The temptation to suppose otherwise --
i.e., that one or more of the above scenarios presents a credible picture of
how communication could have arisen -- might be have been motivated by the fact that we
are trying to imagine what life was like withoutlanguage, when,
of course, we are forced to do this from our perspective as sophisticated language
users already. As should seem obvious, even
historical
linguists and evolutionary scientists are heavily biased
in this respect, using analogies drawn from theway we communicate and innovate at present.
That is, of course, why such theorists find they constantly have to use
metaphorical language, liberally employing 'scare quotes' to that end.120a
Clearly, these two situations aren't
at all comparable; how we innovate today bears no comparison with how
proto-linguists once did this. Hence, we are in no position to form a clear picture of what it was like for human beings
before language was invented (any more than we can now put ourselves in the
'mind' of an ape or an ape-like ancestor of ours). Even an attempt to do this would be forced to
employ the very thing we are now trying to imagine we were once without
(i.e., language), applying it in situationswhere there was no language. It
isn't as if language is an afterthought, a dispensable, insignificant detail that we just happen
to get along with right now, something we can ignore when it suits us; it is
constitutive of our ability to comprehend anything whatsoever. As noted earlier,
human beings haven't yet figured out a way of thinking about anything that
doesn't use language, directly or indirectly -- nor are they ever likely to. And we can say this with some
confidence because of what the words in the last handful of sentences now mean.121
[Option (1b)
above suffers from the same weaknesses.]
Finally,
Professor of Linguistics,
Vyvyan Evans, concurs (but no one should assume I agree with everything Evans
has to say):
"There is one last big problem for the idea of a Universal Grammar. That is its
strange implications for human evolution. If language is genetically hard-wired,
then it self-evidently had to emerge at some point in our evolutionary lineage.
As it happens, when Chomsky was developing the theory, language was generally
assumed to be absent in other species in our genus, such as Neanderthal man.
This would seem to narrow the window of opportunity during which it could have
emerged. Meanwhile, the relatively late appearance of sophisticated human
culture around 50,000 years ago (think complex tool-making, jewellery, cave-art
and so on) seemed to both require and confirm this late emergence. Chomsky
argued that it might have appeared on the scene as little as 100,000 years ago,
and that it must have arisen from a genetic mutation.
"Stop and think about this: it is a very weird idea. For one thing, Chomsky's
claim is that language came about through a macro-mutation: a discontinuous
jump. But this is at odds with the modern neo-Darwinian synthesis, widely
accepted as fact, which has no place for such large-scale and unprecedented
leaps. Adaptations just don't pop up fully formed. Moreover, a bizarre
consequence of Chomsky's position is that language couldn't have evolved for the
purpose of communication: after all, even if a grammar gene could have sprung up
out of the blue in one lucky individual (already vanishingly unlikely), the
chances of two individuals getting the same chance mutation, at exactly
the same time, is even less credible. And so, according to the theory of the
language instinct, the world's first language-equipped human presumably had no
one to talk to.
"Something seems to have gone wrong somewhere. And indeed, we now believe that
several of Chomsky's evolutionary assumptions were incorrect. Recent
reconstructions of the Neanderthal vocal tract show that Neanderthals probably
did, in fact, have some speech capacity, perhaps very modern in quality. It is
also becoming clear that, far from the dumb brutes of popular myth, they had a
sophisticated material culture -- including the ability to create cave
engravings and produce sophisticated stone tools -- not dissimilar to aspects of
the human cultural explosion of 50,000 years ago. It is hard to see how they
could have managed the complex learning and co-operation required for that if
they didn't have language. Moreover, recent genetic analysis reveals there was a
fair bit of interbreeding that took place -- most modern humans have a few bits
of distinctively Neanderthal DNA. Far from modern humans arriving on the scene
and wiping out the hapless ape men, it now looks like early Homo sapiens and Homo
neanderthalensis could have co-habited and interbred. It does not seem
farfetched to speculate that they might also have communicated with one
another." [Quoted from
here; accessed 11/09/2017.]
One thing is reasonably clear: contemporary
linguistic innovators don't invent new areas of discourse as a result of genetic
mutations or variations. Not even the most died-in-the-wool, gold plated,
diamond studded sociobiologist would
argue along those lines. That is why they have had to concoct the entirely
fanciful "memetic" theory of linguistic
or cultural transmission to plug
theGrand Canyon-sized gap in their theory.122
But, even if they were
desperate enough to argue
that contemporary language development had a genetic basis, their theory
would emerge still-born, for only those possessed of the correct genes
would be able to comprehend it!
As
seems obvious, we comprehend languages new to us at present because they can be translated into our own, or
they can be explained to
us in terms we already understand. But, ex hypothesi, that option wasn't
available to these bands of proto-linguists -- for to do that they would have to
have possessed the requisite linguistic skills before they possessed them!
Let us now suppose that the
third
alternative above is correct: that is, that a dialecticalcombination
of cultural and genetic factors facilitated linguistic innovation (which is the
option Harman seems to favour -- in his journal article, at least).
Unfortunately, however, the problems outlined above also plague this
alternative. So, in order to comprehend a radically new innovation (i.e., the
introduction of 'proto-verbs'), MM, for example, would already have
to be in possession of the necessary linguistic resources. But,
as noted earlier,
comprehension is both public and language-based. It isn't something we do privately in our
heads. Hence, on this view, MM (or her group) would have to innovate before
she/they could understand that innovation, or anything said to one another!123
Howsoever this idea is re-packaged, we have as yet no idea how linguistic or
social change could be incorporated into, or caused by, our genetic make-up. On
the contrary, current genetic orthodoxy seems to suggest that the former can't happen -- since
extremely wide human
cultural diversity is in fact
supervenient on what is a largely stable genetic base, so we are told. And,
given current knowledge, there
seems to be no way that social change can be imprinted on the genome.124
If
that is so, a dialectical account (which sees some sort of mediation between
genes and cooperative social labour as fundamental to the origin of language and
thought) can gain no grip. In the end, like it or not, a materialist account of
language must take an HM-route, viewing language as a social phenomenon,
and nothing more.124a
There are several possible DM-escape-routes out of this
impasse. The first involves the postulation of an 'emergent' property of matter
which operates simultaneously in every member of MM's group, permitting
them all to use W2 at the same moment. Alas, this
potential 'escape-route' also suffers
from exactly the same problems as those outlined above. Unless all of these
quasi-miraculous innovators used the same words at the same time (and in
the same way), none of them would be able to understand what anyone
'said'.124b
Now, the weight of probabilities suggests that it is highly
unlikely that this band of proto-linguists would all hit on the very same words
at exactly the same time. Even dialectically-motivated genes
can't establish a settled vocabulary which is familiar to all its users,
and which they all know how to employ in the same way, from day one!
The operative word here, of course, is "familiar"; it is important to remember this
incredible scenario will only work if these proto-linguists
are all familiar with the words that have just been invented!
Is
there a single contemporary human being on the planet who understands even a
highly simplified foreign upon hearing it for the very first time? -- And
they are already sophisticated language users. It looks like this
dialectical-gene has stopped working, otherwise all of us would be able to pull
this trick off all the time. Many of us struggle with words in our own
language we have never heard before, for goodness sake! That's one reason we use
dictionaries. As far as we know, there were no dictionaries in the
Pleistocene.
It would be interesting to see how DM-fans
go about trying to "grasp"
that contradiction -- that these innovators were all familiar with
words that were entirely new to them. Any who fail to appreciate the
irony here need only reflect on the patently ridiculous idea that it is possible
to be familiar with someone who, or something that, has been encountered for the very first time.
On the other hand, if they weren't
in fact
familiar with these newly invented words (as seems to be reasonably certain), then whatever these
proto-linguists mouthed at one anther would be as comprehensible as Pentecostal
ecstatic linguists or aphasics (see below) are to us (let alone to one another)
today.
Even MM, the original innovator, would
have to be familiar with grammatically novel words (which she had never
heard before), as well as know how to use them correctly from the get-go. This isn't credible, to say the least.
Another way out of this
Dialectical Hole might involve
the suggestion that MM, or other members of her group, invented her/their
own private language and subsequently shared it with the rest. But, even
if such a private language were possible (and there are good reasons (over and
above Marx's view) to suppose
it isn't; on this, see the references
listed in Note 86), it would be impossible for this lone innovator to
explain (or communicate) anything to anyone else without the requisite
linguistic resources already being publicly available,by means of which this
could be achieved. As should seem obvious, this 'escape route' would have to
presuppose the existence of the very thing this 'private language' had been invented to explain:
a publicly usable language.
Again, we hit another brick wall.
This series of dead ends
isn't the least bit surprising: it isn't possible to use language to try to
get behind its conventions. Any attempt to do so will always run into the same
obstacles. That is because
language is irreducibly social; any endeavour to account for it along
alternative lines
has to appeal to, or has to utilise, surreptitiously, the very thing it had sought to explain. Hence, in order to make the above scenarios
even seem plausible, we had to suppose that early linguists innovated in the same way we do
today. Failing that, we are forced to concede that they innovated non-socially, as we
don't do today.124c
On the other hand, if they innovated back then as social atoms, individualistically,
they would fail to communicate. Moreover, we can't assume that such
proto-linguists innovated as we do today, for, plainly(!) they didn't possess a language,
whereas we do. They and the
group to which they belonged certainly hadn't
gone through the same social development we have experienced, which helped
create
the sophisticated language we now possess.
The reason why this barrier is non-negotiable
isn't just because we have to use some language or
other to express theories that endeavour to account for the invention of language.125
It is also because, in order to account for the development of language we have to
assume the existence of the very thing we sought to explain -- i.e., the ability
to use language based on historically-conditioned social convention -- so that communication
could take
place. Edit that out, and no wonder communication can't be accounted for
--, except by appeal
to myth, metaphor and miracle. [On this see Note 124c.]
Marxists, who inadvertently undermine belief in the social
nature of language (even if they pay lip service to it elsewhere),
play directly into the hands of the reactionary forces they otherwise seek to oppose.
It is no surprise then to see DM (given its alien-class origins) once again compromise HM
with sinister efficiency.126
Of
course, we needn't speculate about what might, or might not, have happened
hundreds of thousands of years ago, nor need we appeal
to confused religious hysterics to make this point. There are numerous
well-documented cases of aphasics (or, dysphasics)
who utter all manner of odd 'words' and 'sentences' that no one understands
-- even when they are (presumably) 'grammatically correct'. Indeed, a recent episode of House
(i.e., Failure
to Communicate) illustrated this phenomenon -- concerning a patient
suffering from
Wernicke's Aphasia.126a
In this particular episode of House, the afflicted individual came out
with (presumably) grammatically perfect sentences, the words comprising which his listeners could
in fact
understand, but they couldn't figure out why he had uttered them. So, after
he fell and banged his
head on the corner of a desk at an office party, he got up and said to his wife and other concerned
onlookers: "I flung the investment. Why I can sign? It's proficient. Why
disqualify the rush? I'm tabled." Later, he
mumbled (on several different occasions): "I displaced my function", "I
grapple average". "Of golf", "Keep the stain. Knife can't force",
"What is the durable? I got a till in the jug", and "Couldn't tackle the bear. They
took my stain."
So, even though he was using ordinary words in what looked like grammatically
correct sentences, or phrases, no one could
understand what he was trying to say because of the radically odd combinations he
uttered each time, and which seemed not to be relevant to anything happening in
the immediate vicinity. They all knew what his words meant, but they couldn't
grasp his
speaker's meaning.
In the end, (Doctor) House made an educated guess about his speaker's meaning (using
what were 'sort of' rhyming words) about
the last two sentences -- "Couldn't tackle the bear. They took my stain"
-- , interpreting his words to mean he was trying to say he was
bi-polar (hence the reference to the bear -- i.e., "polar bear"), and that he
had undergone a brain operation in a dirty hospital in the 'third
world' (hence "stain"). From that, House concluded this man had caught
cerebral malaria!
So, here, with sophisticated language-users all around him, this individual
still couldn't make himself understood even when he was using familiar words.
What chance would these proto-linguists have with 'words' (in fact, noises)
none of them had ever heard before? And, even if they had heard those noises
before, they would struggle even more than the listeners in that episode of
House to make any sense of them, since those characters in House at
least knew they were dealing with words (and, indeed, with nouns as well as
verbs).
It isn't possible to comprehend such radically flawed 'speech'. In that case,
the above ancestral novice innovators would confront each other as a group of
aphasics face us, or one another, today.
Moreover, we don't need to appeal to fictionalised US TV programmes to make this
point. As noted above, there are plenty of well-documented cases. One
interesting example was outlined in Aitchison (1996). This concerns a monk called
Brother
John who, in the 1970s, experienced a series of epileptic seizures. Whenever these
attacks occurred, he found he couldn't understand anything that was said to
him, and no one could comprehend his speech, either. Longer seizures, though,
had a more dramatic effect:
"He lost his ability to use
language, and was aware of this... [as] he [later] noted: 'I know that certain words I
say are not correct but I do not know which ones...'. When his spoken speech was tested
during a spell, similar sounding nonsense words tended to recur, often variants
of the nonsense sequence tuwari. When shown a picture of a
telephone...[he] said: 'That's it, there. The furi twar. No. Glarity tuware tuwa
tuware ari tuware tuware tuwarere tu tuware tu'." [Aitchison (1996), p.39. Italic
emphasis in the original; quotation marks altered to conform with the conventions
adopted at this site.
Paragraphs merged.]
Why
aren't we justified in concluding that Brother John was innovating, here? The answer is quite
plain: because what he 'said' was incomprehensible -- even to the individual
himself after the seizures abated. And the same can be said of
those hypothetical ancestral innovators, babbling away in their caves many thousands of years
ago.
It
could be objected that aphasics (etc.) are brain damaged, which means aphasia
(etc.) isn't relevant to the case in hand. Of course, it is undeniable that
aphasics are damaged individuals, but the point is that whatever the
aetiology,
the 'language' of anyone who utters nonsense (or who produces totally strange
noises) would be, and is, incomprehensible to those around them -- unless someone
'de-codes' it. But, in order to do that they would need a sophisticated language into
which it could be de-coded. Plainly, that would require the existence of a language in which this could be
done -- and, given the
scenario outlined above, that language would have to contain 'proto-verbs',
or no such translation could be effected.
Once more we would have to assume the existence of the very thing we sought to
explain!
Moreover, those without brain damage struggle to comprehend individuals who use language in odd ways, despite their expert training and
advanced technology. Proto-linguists, therefore, stand no chance of comprehending
innovators who use new words in new ways.
Anyway, what is a massive brain mutation other than damage to that organ?
Harman's resistance to the idea that human
development is "exceptional", and was sudden and took place relatively recently, was clearly motivated
by a fear that to think otherwise would be to make dangerous concessions to
Idealism. Indeed, his discussion appeared under the heading: "The new idealist
challenge", where he summarised what he took to be the main arguments central to
this challenge -- one of which was based on Gould and Eldredge's theory of
Punctuated Equilibrium. His conclusion was:
"The overall impact of these
different arguments has been to encourage a fashion in recent years which sees
'a distinctively human way of life' as arising very late in history, as a result
of a 'human revolution' which first produced culture and language." [Harman
(1994),
p.96.]
He then summed up his counter-argument
to the above in the
following way:
"So there has to be a
recognition of how quantity turns into quality, of how through successive
changes animal life gave birth to that new form of life we call 'human', which
had a dynamic of its own, shaped by its labour and its culture not by its genes.
But this should not lead to a collapse into a new
idealism which sees culture and
language
as emerging from nowhere in the fairly recent past." [Ibid.,
p.102.]
This partly explains why he rejected
Gould and Eldredge's theory:
"Finally, the
argument that punctuated evolution can take place does not, in itself, prove
that it did take place in such a way as to produce culture and language
suddenly. And there is one powerful argument against this -- that of brain size.
If the evolution of humanity was the result of very rapid changes towards the
end of a period of millions of years, then that is when you would expect the
most characteristic feature of
homo sapiens
-- the massive size of our brain compared to our bodies -- to arise. The
original formulation of the punctuated evolution hypothesis by Gould and
Eldridge (sic) in fact held to this view, contending that the brain hardly increased
in size for the million years
homo erectus existed. But, as Stringer
points out, there is 'little evidence' to back up this view.
"That leaves a
problem for any theory which sees the 'human revolution' as occurring all at
once half a million years ago with the replacement of
homo erectus
by homo sapien (sic), let alone 35,000 years ago after the evolution of
anatomically modern humans: why did late homo erectus have a brain twice
the size of the
Australopithecines, and the
Neanderthals
a modern sized brain? It could not have been simply to undertake the mental
operations which could be done by their ancestors millions of years before.
"At the same
time, it is inconceivable that our forebears of a million years ago could have
survived unless they had already developed ways of co-operating together to cope
with their environment and of transmitting knowledge to each other on a
qualitatively greater scale than is to be found among our ape cousins. For by
that time they were already moving out of the African valleys where their
species originated to colonise much of Eurasia, showing they were capable not
just of living in a certain restricted ecological niches, but of adapting a
variety of environments to their needs -- learning to discriminate between those
newly encountered varieties of plants that were edible and those that were
poisonous, learning to hunt new sorts of animals, learning to protect themselves
against new predators, learning to cope with new climates." [Ibid.,
p.99.]
Now, Harman's evidence and argument won't be disputed here, but this form of exceptionalism clearly suggested to
Harman that language and recognisably human psychological traits might not have
material roots in collective labour (etc.), implying some form of Idealism.127
But, we have already seen
that DM itself collapses into Idealism (since it is not only predicated
on
it, it was born out of it -- despite the materialist spin that is said to have
been inflicted on it -- on that, see
here). Moreover, Harman himself
had to use several inappropriate metaphors to support his case. He supposed, for
example, that specific developments in evolution would have "encouraged" others
to take place, and he attributed to natural selection a mystical sort of
agency,
saying things like the following:
"Natural selection would bring about…evolution
in the direction of ever larger, denser and more complex neural networks,
capable of directing and learning from intricate motor functions of the hand and
of using minute changes in gesture or voice to communicate…." [Harman (1994),
p.100. Bold emphasis added.]
This isn't just
Harman's shorthand for "Natural
causes eliminated those organisms that were less successful in reproducing,
because of factors X, Y or Z". That can be seen from the metaphor Harman found he
had to use. In fact, this mirrors
Darwin's own employment of the metaphorical (and thus the teleological) term
"natural selection". Indeed, Darwinism is shot through with teleological
concepts; and so is Harman's article.128
Of course, scientific metaphors are
entirely unexceptionable -- but only if they aren't taken literally! The suspicion
that Harman
has in fact done just this isn't helped by the way he attributes
direction to causes that he elsewhere says aren't "contingent".129
In that case, it isn't easy to reconcile the alleged capacity of natural forces
to control and direct events with theirsupposedly non-necessitating
nature.129a
Nor is it easy to explain how 'non-intelligent matter' is able to 'direct'
anything at all; nor, indeed, how objects and processes in nature can be 'directed',
'controlled', and 'determined' by anything else. Still less is it easy to comprehend how objects
and process succeed in 'obeying' these 'injunctions' everywhere and everywhen,
nor yet how these non-necessitating 'directives' command such unquestioning
compliance throughout the entire 'non-intelligent', material universe, for all of time.130
As intimated above, in order to
construct such a causal account of nature, theorists found that they had to animate
matter at every turn; they had to attribute to it a capacity to 'follow orders'
(as it were), or to 'obey' certain natural -- or, maybe, 'dialectical' -- 'laws'.
Clearly, that would make objects and processes in nature either intelligent
agents (capable of 'obeying' 'laws' without the benefit of an education and in
the absence of any sort of socialisation), or
they are rather dim agents, bullied and pushed about the place by a Universal Will.
Indeed, this is precisely what
is implied by a literal interpretation of such metaphors.130a
Now, far from it being the case that only human exceptionalism
implies Idealism,
Harman's approach itself ends up putting
'mind' before matter, for it depends on the (implicit) metaphysical doctrine
that natural events are directed by just such a Cosmic Will (a notion which is buried
behind all those metaphors).131
In fact, as pointed out earlier,
Harman found that he had to
anthropomorphise nature at the very start so that intelligence and purpose could
'emerge' from it at a later stage.132 To that end, causes
were
pictured as determining ('deciding', 'directing' or 'controlling') events
by natural law.132a These are "laws" which must be "obeyed" by 'lifeless,
non-intelligent' matter, this fable made 'acceptable' by the indiscriminate use
of metaphor -- i.e., in this case, the mystical correlate of those
aforementioned
"just-so-stories". [Since I have developed these points in more detail
elsewhere in this Essay, I will not rehearse them again here.]133
To be sure, Harman is concerned to defend
Engels's belief that labour contributed significantly to the development of
human language and cognition, since he sees in recent trends a resurgence of
Idealism. However, in order to defend Engels in this area it isn't necessary to deny human exceptionalism; in fact, as this Essay has
shown, the best way to defend the 'Labour Theory of Culture' is to emphasise our
uniqueness.
To sum up: there is nothing in current
Darwinian theory that allows for social change to be imprinted in the genome,
and there is nothing in Marxism that permits cultural phenomena to be
genetically encoded. Indeed, human cultural development has been
free-floating on a largely stable genetic base for tens of thousands of years. This explains why human beings
across the world can inter-communicate, and, with relative ease, share and enjoy cultural
artefacts with one another, even where their respective languages appear
to be
totally dissimilar, and their history could be more unlike.133a
Since it isn't possible for Neo-Darwinism to account for this, it is puzzling why this theory has been given any credence
(in this regard) by Marxists trying to account for language and thought.134
As we have seen, all that Harman himself could offer his readers in order to counter the reactionary ideas implied by
ultra-Neo-Darwinism was a tired old reference to Engels's
first 'Law', 'supported' by the hope that others will "see things" this
way.
Indeed, one might just as well try to stop an avalanche with
a "Keep Off" sign.
The minor genetic changes that have
taken place over the last ten or twenty thousand years are testimony enough to
the truth of the above comments. Over that period, cultural innovation has been
truly staggering
while genetic change has been relatively insignificant. Harman himself alludes to
this fact when he speaks of a social process that displays "a dynamic of its
own". As he will no doubt agree (but, alas,
he
is no longer with us!), the theory that accounts for this is called
--
HM. But, and once
again, it is DM that obscures the clarity that the latter theory brings to an
understanding of our own history and social development.134b
It is surely time we consciously selected out
this mutant, alien-class theory (DM); it is one aspect of our radical
inheritance we shouldn't thank our theoretical ancestors for passing down to us
without modification.
Another comrade who has also drifted
off into deep waters is
Andrew Collier. Collier recently attempted to controvert the
ideas of certain so-called "Wittgensteinians" with arguments drawn from the work
of Noam Chomsky,
Donald Davidson,
Trevor Pateman and
Roy Bhaskar,
among others.
One of Collier's main worries appears to
be Wittgenstein's alleged insistence that the reasons for an
action can't also be the causes of that action. Apparently, refuting this
particular idea
is important because it allows theorists to develop a 'naturalistic'
explanation of human action.
[At this point,
alarm bells should be ringing
in the head of anyone who accepts
HM.]
Of course, this issue is
connected with Collier's attempt to provide his readers with his own brand of
a prioriSuperscience, dressed up this time as something that
its adherents call
"Critical Theory". Indeed, Collier's argument forms part of a wider attempt to
establish the case for a Bhaskarean-style "Critical Realism".135
Unfortunately, the reader
will search long and hard (and to no avail) in Collier's book for a direct
reference to any of Wittgenstein's actual arguments. [In fact, there is
no reference at all to the Philosophical Investigations (or any
other of Wittgenstein's works) in the
Bibliography of Collier (1994).] A strange omission one might feel for an
avowed realist;
precious little correspondence with reality
evident here!
What the reader will find in place of
evidence and argument, however, are several rather vague allusions to the opinions of
certain "Wittgensteinians" about this or that aspect of 'mental' phenomena --
alongside a few misguided comments lifted from a book by former
"Wittgensteinian", and one-time radical, Trevor Pateman.136
The substance of Collier's case
against Wittgenstein fares little better. Quoting Bhaskar, he argues as follows:
"What does it mean to say that reasons can be
causes? Bhaskar suggests that:
"'When something is cited as a cause it is, I
think, most typically being viewed as that factor which, in the circumstances
that actually prevailed, "so tipped the balance of events to produce the known
outcome"…'…." [Collier (1994), p.152, quoting Bhaskar (1979), p.106. I
am clearly using a different edition of the latter, since Collier gives p.83 as the location
of this passage. Quotation marks altered to conform with the conventions adopted
at this site.]
He then concludes that:
"Intentional actions involve beliefs and
desires…. It is hardly open to dispute that, given a desire for something,coming to have a belief about the way to get it may 'tip the balance', and so
be naturally described as 'the cause'." [Ibid., p.153. Bold emphasis added.]
Of course, a belief about "a way to get"
X
isn't a reason why X was wanted in the first place. If NN wants to
attend a
protest march, for example, and believes the Number 159 bus will take her there, that
surely isn't
the reason why she wanted to be on that march!
Nevertheless, it is worth asking whether the opening sentence of the first of the above passages,
namely:
"[W]hen something is cited as
a cause it is, I think, most typically being viewed as that factor which, in the
circumstances that actually prevailed, 'so tipped the balance of events to
produce the known outcome'...", [Ibid.]
is based on some the yet-to-be-published,
innovative research into the
sociolinguistics
of the terms in question -- or, whether it was just a guess.
If the former is the case, it is reasonable to ask where the data is to be found that
supports the view that a cause is "typically…viewed" as something that "tipped
the balance of events".
This
query isn't being raised out of mere cussedness, or even 'pedantry'; if someone
makes an empirical claim about the "typical" way the rest of us are supposed to
interpret (or use) a certain word that is in fact not even remotely like the way
they would normally employ it, some supporting evidence is the least one
should expect. [If there is any such evidence, perhaps Collier will include it in the second
edition of his book.] Even so, in abeyance of this data, we would be unwise to
revise our use of English just because of what Roy Bhaskar says
everyone else typically understands by the word "cause".137
In
fact, competent
users of language needn't wait for the results of such a study, or survey, before they
have to be informed
what the
words they already use mean. If they were stupid enough to wait for one
such, a study of the words "word" and
"study" (among others) would surely have to be carried out first before any conclusions could
safely be drawn. And how might that task be undertaken, for goodness sake?
Manifestly, if no one understands what these two words mean without a
study being commissioned to that end no one would know what was being
proposed.
In fact, as I am sure Collier
is also aware, when someone challenges another's understanding of a word (or
attempts to revise it), an appeal to everyday examples -- or in some cases a dictionary (but not a survey)
--, would,
in many cases, settle things. Collier,
unfortunately, omitted these simple steps. In order to put this right, several typical examples of the use of "cause" (and other related
terms) will be examined later, at which point the reader will be invited to decide for herself if they
are typical. In advance of that it is worth noting that no dictionary
worth its salt defines "cause" in the way Collier or Bhaskar suggest.
In fact The New Shorter Oxford English Dictionary (1993 edition) defines "cause"
in two ways:
"1
gen. That which produces an effect or consequence; an antecedent or
antecedents followed by a certain phenomenon…. A person or other agent who
occasions something, with or without intent…. A fact, circumstance or
consideration which moves a person to action; ground for action, reason, motive;
esp adequate motive or justification….
"1
v.t. Be the cause of, effect, bring about, occasion, produce, induce, make
(a person or thing to do, a thing
to be done, be done); (arch) bring it about that…. 2
v.t. Give reasons or excuses…."
Admittedly, this mentions "reasons" as part
of the definition, which connection is not in fact being denied here. In some
cases, clearly, causes are reasons (as, indeed, reasons are sometimes
causes) -- or, at least, they can be given as reasons. For example, if
asked for the reason why a certain tree fell over, one could point out that it
had become rotten with age and the wind finally caused it to topple in the wind. Although I won't quote it here, the same dictionary lists "cause" as
part of the meaning of "reason", too. What is in dispute, however, is
whether reasons are always causes, and whether causes are always 'balance-tippers'.
Readers of
the above dictionary entry will struggle long and hard to find any reference to balances
(tipped or otherwise) in connection with either term.
Someone might point to the following sentence
in the above dictionary definition:"A
fact, circumstance or consideration which moves a person to action...", and
argue that this is, to all intents and purposes, the same as "balance-tipping
cause". But, that isn't the implication of this part of the definition. Human
beings aren't permanently teetering on the verge of doing whatever they decide
to do, which is what the 'balance' metaphor implies. For something to be a
'balance-tipping' cause that would have to be the case -- the individuals
concerned would have to be in some sort of 'balance-tipping' -- or 'ready to be
tipped' -- state -- especially if this is
supposed to be "typical" in such causes. While something might indeed move us to
act, there is no suggestion that there was a 'balance' that needed tipping. For
example, suppose you find out that your house/flat is on fire and are "moved" to get out as
fast as possible. Does that imply you have been teetering on the edge of rushing
to safety for years and the fire finally nudged you in that direction?138
Of course, it isn't possible to settle
philosophical questions with an appeal to what dictionaries have to say; the above
considerations, and the dictionary quoted in Note 138, are only being
included to
counter Collier's claim that:
"It is hardly open to dispute that, given a desire for something,
coming to have a belief about the way to get it may 'tip the balance', and so
be naturally described as 'the cause'." [Collier (1994), p.153.]
Despite this, it is a little rich of Collier
-- who follows Bhaskar in this -- to quote what we might be inclined to say
as part of a general scientific/philosophical analysis of the way the
natural world is supposed to work, all the while criticising Wittgenstein and
various un-named Wittgensteinians for
allegedly doing the very same thing! [Cf., Collier (1994), pp.205, 214-15.]
If,
however, such a consideration of "what we say" had in fact been aimed at
reminding us how we actually use ordinary words (illustrated by a wide selection
of everyday examples, like those given in this
Essay) -- perhaps also aimed at
dispelling philosophical confusion, ensuring that we avoid employing
language inappropriately or misleadingly -- all well and good. We would then be
on the same page. Unfortunately, Collier
simply gestures at doing the former while falling headlong into
performing the opposite of the
latter, an outcome which the whole exercise had originally been aimed at sidestepping.
Hence, far from representing some sort of "Critical Realism", Collier's account more closely
resembles "Uncritical Idealism", wherein what we say determines
a causal law, in reality! The end result is that all the effort he and
others have put into providing an 'objective' account of nature and society has been
vitiated since it is now plain that the whole exercise is based on the
reification of
Bhaskarean jargon, and little else!
Indeed, this approach illustrates quite nicely a point
made in Button, et al (1995), quoted several times already:
"As to the widespread
disparagement of attempts to resolve philosophical problems by way of appeals to
'what we would ordinarily say', we would proffer the following comment. It often
appears that those who engage in such disparaging nonetheless themselves often
do what they programmatically disparage, for it seems to us at least arguable
that many of the central philosophical questions are in fact, and despite
protestations to the contrary, being argued about in terms of appeals (albeit
often inept) to 'what we would ordinarily say...'. That the main issues of
contemporary philosophy of mind are essentially about language (in the
sense that they arise from and struggle with confusions over the meanings of
ordinary words) is a position which, we insist, can still reasonably be proposed
and defended. We shall claim here that most, if not all, of the conundrums,
controversies and challenges of the philosophy of mind in the late twentieth
century consist in a collectively assertive, although bewildered, attitude
toward such ordinary linguistic terms as 'mind' itself, 'consciousness',
'thought', 'belief', 'intention' and so on, and that the problems which are
posed are ones which characteristically are of the form which ask what we
should say if confronted with certain facts, as described....
"We have absolutely nothing
against the coining of new, technical uses [of words], as we have said. Rather,
the issue is that many of those who insist upon speaking of machines' 'thinking'
and 'understanding' do not intend in the least to be coining new, restrictively
technical, uses for these terms. It is not, for example, that they have decided
to call a new kind of machine an 'understanding machine', where the word
'understanding' now means something different from what we ordinarily mean by
that word. On the contrary, the philosophical cachet derives entirely from their
insisting that they are using the words 'thinking' and 'understanding' in the
same sense that we ordinarily use them. The aim is quite
characteristically to provoke, challenge and confront the rest of us. Their
objective is to contradict something that the rest of us believe. What the 'rest
of us' believe is simply this: thinking and understanding is something
distinctive to human beings..., and that these capacities set us apart from the
merely mechanical.... The argument that a machine can think or understand,
therefore, is of interest precisely because it features a use of the words
'think' and 'understand' which is intendedly the same as the ordinary use.
Otherwise, the sense of challenge and, consequently, of interest would
evaporate.... If engineers were to make 'understand' and 'think' into technical
terms, ones with special, technical meanings different and distinct from
those we ordinarily take them to have, then, of course, their claims to have
built machines which think or understand would have no bearing whatsoever upon
our inclination ordinarily to say that, in the ordinary sense, machines do not
think or understand." [Button, et al (1995), pp.12, 20-21.]
Despite this initial worry, one might also be
forgiven for wondering what relevance the following emphasised words possess:
"Intentional actions involve beliefs and
desires…. It is hardly open to dispute that, given a desire for something,
coming to have a belief about the way to get it may 'tip the balance', and so
be naturally described as 'the cause'." [Ibid., p.153. Bold emphases added.]
Even supposing it were true that means-end
reasoning preceded some of our actions, how would that show reasons were
causes, let alone 'balance-tippers'? Of course, if something is to be
regarded as the cause of something else, and this is supposed to form
part of a novel scientific analysis of the processes involved, evidence would be required -- otherwise any old fable
(or, indeed, any old "just-so story") could be presented as 'hard science'.
Nevertheless, Collier's 'analysis' purports to be
a little more
prosaic, based on what we might ordinarily say and think. If so,
as noted above, a
wide selection of examples of everyday use would have been much more appropriate
and helpful. Unfortunately,
Collier confined his 'research' to an examination of what Roy Bhaskar
had to say, not what ordinary speakers actually say.
A consideration of a wider selection
of the sort of ordinary examples that Collier unwisely omitted reveals a different,
and far more complex, story (one that isn't easy to squeeze into Bhaskar's a priori,
dogmatic
and ultimately Idealist straightjacket).
Further developing an earlier example: if, say, comrade
NN believes that running for a bus
will help her catch it, and so acted upon that belief, would this count as the reason
why she caught the bus? Even if running for a bus caused her to catch it would
that be her reason for catching it -- running for it? It seems it should if Collier were
correct. However, if, when asked for the reason why she caught the bus,
NN replied: "Because I ran for it" (and not, for example, "Because
I wanted to get to
the march on time"), we would be somewhat bemused, and rightly so. But, if
reasons were causes, we wouldn't be at all puzzled by such a response. We would accept any
'balance tipping' cause as a surrogate reason, or as the reason, no
matter how odd it might seem. In that case, NN could have answered "Because of the 'Big
Bang'", and that would be the end of the matter -- since that event tipped every subsequent
'balance', by default.138a
And, what if it turned out that the
'balance-tipping cause of NN actually catching the bus was the fact that the driver
accelerated away rather lazily, and forgot to close the doors, since he was
day-dreaming about his new girlfriend? Was thatNN's reason? It
was part of the cause.138b
Naturally, it could be argued that these
responses are absurd because (i) Collier does not claim that all causes are
reasons,139 and (ii)
He connects
desires and intentions with the
reasons, or causes, of actions. Clearly, the "Big Bang" and the
aforementioned driver's
actions couldn't serve in that capacity, nor could they cause NN's actions in the required manner.
But, Collier's own account actually implies
several of the above crazy conclusions. He says:
"Intentional actions
involve beliefs and desires…. It is hardly open to dispute that, given a desire for something,coming to have a belief about the way to get it may 'tip the balance', and so
be naturally described as 'the cause'." [Ibid., p.153.]
In that case,
NN could reply that she
caught the bus because she wanted to, which, even on Collier's terms, would be an
empty explanation. NN caught the bus because she wanted to catch it, and her
wanting to catch it caused her to catch it, so this must be the reason why she did it.
But, her real reason might have been to get to the march on time, which she would
readily have volunteered, if asked.
Well, perhaps this again misses the point?
In fact, it doesn't; the
cause of NN's catching of the bus and the
reason she caught it are separate matters. Collier's account simply runs
them
together.
However, if we examine Collier's
words more carefully, another puzzle soon emerges:
"…coming to have a belief
about the way to get [what one desires] may 'tip the balance', and so be
naturally described as 'the cause'." [Collier (1994), p.153.]
From this, it looks like it is the presence
of a belief or set of beliefs about the means to attaining our ends (and not
so much our desires as
such) that is supposed to cause our actions.
At least this interpretation would rule out the
"Big Bang" response mentioned earlier --, but, alas, it would do so only at the
expense of introducing several
other absurdities. For instance, as pointed out above, no one would say that the reason why they did
something could be identified with any of the beliefs they had about how they
might go about attaining what they had, in this case, wanted. If they did, then
NN could reasonably reply that the reason she
caught the bus was that she ran for it (since, presumably, she must have
believed that running for the bus would help her catch it)! Indeed, Collier
himself could
declare that the reason he wrote his book was that he had a word-processor,
which he bought because he believed it would help him write it!
To be sure, it could be objected that Collier
didn't mean that any old beliefs were relevant, only those that are relevant to
attaining our ends. But, the above examples were specifically chosen because they
were of that sort.
Even so, it could be countered that the above reasons are
contributory causes in disguise, not the cause. In the first case, running for the bus
helped cause
NN to catch it, and buying a computer was part of the causal background that led
to the said book being written.
However,
even if that were
the case, could any of these be
described as "balance-tipping causes"? That is far from clear. In some
cases, perhaps so, in others, maybe not -- and that might be the case even if
the candidate reason in question was the cause. So, the cause of
NN
catching the bus could be the fact she ran for it, but the reason she caught it
was she wanted to go on a march.
Clearly, one of the problems here is that it
is all too easy to confuse two different meanings of "reason". In order to
illustrate the latter connotation, consider another example: If
comrade MM attends a march against the Nazi BNP and gave as his reason for being
there that he ran for the bus withNN (since that was the easiest way to get there), would we count as his reason for demonstrating against the BNP
that he ran alongside NN? In one sense of "reason" perhaps we might, but in
another sense it would be a joke. If MM had been asked why he was on the march,
and he replied "Because I ran for a bus with NN", few would accept that as
anything other than a supercilious remark, at best.
The superficial plausibility of Collier's
argument relies on conflating these two meanings of "reason". One of these
connotations arises in connection with an explanation others might give why something
happened, and this might indeed involve reasons being causes, but this isn't
always the case. Nor is it even typically the case with the reasons an agent
might volunteer. It would be perfectly acceptable to explain why MM caught the
bus that he ran alongside NN (and this might even be recruited as one of the
contributing causes for his successfully catching it), but it wouldn't count as part
of MM's reason for going on that march, or even for catching that bus.
Furthermore,
in connection with explanations of this sort, not every use of "because" is causal.
Consider, for example, the
following:
B1: NM broke union rules
because the constitution says workers must have a ballot before each work-to-rule.
B2: Two is a
prime number
because it has exactly two factors, itself and one.
B3: DM-Athletic's first goal didn't count because the striker was
offside.
B4: You can't be nominated for
that position because you aren't a paid-up member of the party.
None of these uses of "because" is at
all plausibly causal, even if all are explanatory. In B1, the union rules didn't cause the alleged infraction; in B2, a number isn't caused to be prime by the
said definition,
and so on.139a
It could be objected once more that
the above misrepresents Collier's argument, for it assumes that he thinks causes are
reasons, when he merely argues that reasons are causes. But, as we shall see below, Collier's
'theory' does indeed imply this since his account of causation suggests that
causes are in fact surrogate reasons, and that everything that has
happened (or will ever happen), is subject to the operation of the mysterious
'Cosmic Will' we met earlier.
Anyway, even if this distinction
were
disallowed (for some reason), would things really be as Collier depicts them? Are beliefs and
desires normally regarded -- i.e., in everyday language (for, as pointed out
earlier, what else could Collier/Bhaskar's own term "naturally described"
amount to?) -- as little more than 'balance-tippers'? Is this
metaphor -- which pictures the mind as a sort of see-saw or weighing scale --,
any better than other figures of speech we have learnt to distrust in the
Philosophy of Mind? The few everyday examples Collier presents in support of
his rather bold claims suggest that his case is far more hyperbole than
hypothesis. Wittgenstein himself
characterised
this particular theoretical malady (i.e., the urge to generalise from a few
'non-standard', or specially-concocted, applications of a word to all of its
instances) as an intellectual disease, a pernicious case of philosophical malnourishment brought on by a diet of too
few examples.140
Hence, it would seem that Collier's somewhat counter-intuitive thesis -- wherein
broad results have been built on narrow evidence -- is further confirmation of
that untoward diagnosis.
Consider, therefore, these
additional, perfectly ordinary examples:
C1:
NN believed that the bus
was going to be 20 minutes late, so she decided to walk.
C2: NM wanted to go to France for his
holidays, so he booked a flight to Paris.
C3: MM believed that
Kier Starmer was a
socialist, so he voted Labour.
C4: The angry workers wanted a
better deal, so they went on strike.
C5: The boss knew the workers
wouldn't back down, so she gave them what they demanded.
C6: The philosopher wanted to make a
point about causation, so he wrote a book.
[These are all boringly typical sentences,
which we all recognise as such as soon as we read them.]
Are we really supposed to believe that in any
of the above there is a "balance" that has been "tipped" in a particular direction?
Or even that it is natural -- or "typical", to use Collier's word, again
-- to re-describe any of them that way?
Are commuters in a constant state of
equilibrium, hovering between walking and riding, to such an extent that a bus that is
20 minutes late will tip them one way, while one that is only 19 minutes 59
seconds late will send them the other? Are bosses permanently teetering on the edge of settling with recalcitrant workers, all the
while perched in a state of equilibrium between calling in the Police and
capitulating to militants? Are Labour voters wobbling on a knife-edge,
half-Tory, half-Labour? Does it take a Kier Starmer to unbalance
them? Do philosophers sit on fences all day long, dithering about writing books
and articles until a sufficiently unbalancing thought or desire enters their
heads? Do people linger in a state of indecision over proposed holidays to
France?141
It
could be argued that what Collier really means is that NN's reasons for doing
X,
Y or Z aren't so much the cause of her X-, Y-, or
Z-ing, but of NN's intention
to bring about X, Y or Z. In the earlier example, for instance,
NN wanted to go on
the march; that is what caused her to initiate a series of voluntary and intentional
actions related to whatever she believed would bring it about.
If
asked why she ran for that bus, NN would perhaps reply "I wanted to go on the
march." So her reason (i.e., her aim to join the march) caused her to initiate a
series of actions (getting dressed, having breakfast, leaving her flat, running
to the bus stop, etc.). Now while there might be, and, indeed are, a multitude of
other causes of this series of events (clearly they are far too many to list,
but they range from her beliefs about the properties of the physical universe in
her vicinity, the
behaviour of others (such as bus drivers), her knowledge of the transport system, and
so on), the main cause of the series of actions depicted above was NN's reason
(i.e., to go on that march), which informed all of her other relevant
voluntary actions and intentions that morning. In that
sense, her reason was the most important cause of her actions because this
lay behind most, if not all, of the rest, and this is what "tipped the balance".
Without that reason, and hence those intentions and voluntary actions, she
wouldn't have done what she did.142
Or, so it could be argued...
There
are, however, several problems with this way of seeing things:
(1)
NN may in fact have two or more perfectly good reasons why she did what she did
(i.e., running for that bus), some of which could even be operative at the same
time as the first given reason. For instance, these could comprise two or more
of the following: (a) To go on the march, (b) To test out a painful leg muscle,
or new pair of trainers, (c) To beat her companion to the bus, (d) To wake
herself up, (e) To avoid getting wet, (f) To escape from a dog/annoying man, (g)
To burn some calories, (h) To increase/test her fitness..., and so on. Which
one of these is the 'balance-tipping' cause of her running for that bus?
Some
might want to argue that the 'balance-tipping' cause is always the main
reason why agents do what they do. But, in the above case, NN might have as her
main reason that she wanted to go on the march, but what actually made her run
for the bus was her desire to test out a painful leg muscle.
Now,
if we were to argue that in that case her desire to test her painful
leg is the 'balance-tipping' cause in this instance, then we would already have
resiled
from the above thesis, that such 'balance-tipping' causes constitute the main
reason or cause why people do things. Even so, the 'balance-tipper' here was the painful leg. But,
that leg must have been in that condition for some time. Does this mean that
'balance-tipping' causes can last for weeks? [More on this presently.]
It
could be countered that it was her desire to test her leg that was in fact the
'balance-tipping', main cause,
in this instance. Hence, if her leg hadn't been painful the day before, she wouldn't have
run for the bus today; she might have walked instead. Indeed, and in this case
we would seem to have a reason which is also a cause (a possibility that hasn't in fact been
ruled out in this Essay). But, is it really a 'balance-tipper'? [More on
that
presently.]
Anyway, this needn't always be the case. As noted above, there could be
several reasons why NN ran for that bus.
Supporters of the 'reasons-as-causes' view of things might now want to occupy a fall-back
position, and argue that the set of reasons why agents do what they do
constitute the cause of what they do. That response brings us to the second
reason (no pun intended) why the general equation of reasons and causes is misguided.
(2)
As noted above, it is important to distinguish two senses of "reason":
(a)
"Reason" as an explanation an agent would give for why they did
X, Y or Z, and,
(b)
"Reason" as something that caused the X-ing,
Y-ing or Z-ing.
Now,
it might not always be possible for an agent to be able to say what caused
something to happen, but it is impossible for an agent not to know his/her
reasons for doing something intentional.142a0
This shows that reasons and causes aren't always the same. This
distinction becomes all the more glaring when motives are introduced into the equation
(which are, in fact, more natural candidates for psychological causes
than reasons). So, if NN were to φ for reason, R, then NN must know what
R is -- that is, NN must be aware of R and be
able to volunteer R as their reason if asked. On the other hand, if NN's motives for φ-ing were
F, G, or
H, then it is not always the case that NN will be aware of F,
G, or H, or be able to say what they are.
[The Greek symbol used above is explained
here. R is a
dummy variable for propositions or clauses that could be advanced as a reason
(such as "I wanted to go on the march", or "I want to prove Collier and Bhaskar
wrong"); F,
G, and H are
dummy letters for noun or verb phrases that could express motives (such as "To
annoy the neighbours" or "Jealousy"). It is worth adding that we
can't be unaware of our intentions, but we don't always know our motives.
That shows that intentions and motives aren't the same.]
Collier in fact has an answer to this:
"That reasons can be causes
is also a necessary condition of the phenomenon known as rationalization
(in the Freudian sense). This occurs when the reasons sincerely given for an
action by an agent are not the real reasons. Thus we may suppose that
Henry VIII
sincerely believed that he had his marriage to
Catherine of Aragon
annulled because it was contrary to
canon law,
whereas the real reason was that she had not provided him with a male
heir, or perhaps that he fancied
Ann Boleyn.
What is the force of 'real reason' here? Surely causally efficacious
reason.... In questioning one's mental
states in this way, one is, among other things asking whether one's putative
reasons are one's real reasons, i.e., the reasons that are effective. For
instance, I may come to question whether my believing a scandalous story about
an odious political leader is really caused by the evidence for the story or my
desire to vilify that leader, or whether my depression is caused by the state of
the world or the state of my digestion." [Collier (1994), p.154. Links
added; paragraphs merged.]
However, from the above, it is clear
that Collier has confused motives with reasons
tout court. Motives can indeed be reasons in sense (2b)
above, but not in sense (2a). Sure, Henry VIII
might have given as his reason for annulling his first marriage that it wasn't a
legitimate union in canon law (since Catherine had been married to his deceased
brother,
Arthur),
while his motives might have been as Collier suggests they were. It won't do, however,
to call motives reasons in sense (2a), unless Henry would have given
one of them
as his reason, and not have them attributed to him by others after the event.
Even so, the example Collier chose isn't a
happy one, for it shouldn't be news to Marxists that members of the ruling-class
hide their real motives and their real reasons, and advance in public the
reasons they give for taking a certain course of action for ideological or
'public relations' purposes. Nor will it do to appeal to Freud's
pseudo-scientific theories in support, either. Human beings were surely aware of
ulterior motives long before Freud was born, and that concept isn't made any the clearer by having Freud's obscure jargon, and his even
more dubious a priori speculations, imposed upon it.
Anyway, it isn't as if Henry VIII was unaware
of his real reason for divorcing his first wife, so this example doesn't undermine the distinction made above.
Collier's second example (about the odious politician) also trades on confusing
reasons with motives. If an individual doesn't know the reason why they believe a story,
then there isno reason (in sense (2a)) why
they do, only possible motives for so doing (2b).
Having said that, Collier's response does nothing to erode the clear distinction
drawn above between intentions and motives. Unless we suppose Henry VIII wasn't
in control of anything he did, he can't fail to have known his intentions, what
he aimed to by what he did. That is quite different from him knowing what his
motives were, or what his real motives were. He might not have known
either of these, but then again he might have known one or both of them. But he
can't have been unaware of what he aimed to do, what his intentions were. With
respect to intentions, there is no "might have", or "might not have", here. If
someone does something intentionally, that is the same as saying they knew what
they were doing. Let us suppose that footballer (soccer player), MN,
intentionally kicks opponent, NP. The intention here was to kick NP,
but the motive might have been, say, jealousy. The referee, in awarding a red
card, couldn't care less about the motive, it is MN's intention that
prompts
him to order MN off the pitch.
Another serious difficulty with picturing
causes as 'balance-tippers' is that it is incapable of explaining other clear
instances of causation which:
(1) Don't involve any obvious change, let
alone account for an upset equilibrium/'balance';
(2) Are the result of "negative
causation";
(3) Involve 'disconnected' causes;
(4) Are part
of complex social changes; or,
(5) Involve multiple causes.
[(3), (4)
and (5) clearly overlap somewhat and sometimes.]
Taking each in turn:
(1) For
example, we might want to say that the reason (sense (2b))
that a bridge remains standing is because of its careful construction and
design, along with the properties of the materials used to build it (and, of
course, the forces operating upon that bridge and its surroundings), etc., etc. Clearly, this sense of causation requires no reference to any
balances that have been "tipped" anywhere. And yet, this use of "cause" covers
countless examples of causation at work every day throughout the universe --
those
that operate on mountains, continents, planets and galaxies. Indeed, the cause
or causes that maintain the steady motion of the planets, stars and galaxies around
their relevant centres of rotation (or, even along their individual
world lines,
etc.) can't easily be accommodated by such highly restrictive, Bhaskarean figures of speech.
(2) Examples of negative causation include
the following: (a) MM fell sick because he forgot to take his medicine;
(b) The forest caught fire because it hadn't rained in six months; (c)
Dialectical Wanderers lost their last soccer game because they had five players
sent off for persistently abusing the referee. It isn't easy to see any of these
as 'balance-tippers' since such causes are privations and hence do not in any real sense exist.
(3) Disconnected causes are those where there
doesn't seem to be any obvious fact of the matter connecting an alleged cause
with its supposed effect. There are many examples of this. [In what follows, I
borrow heavily from Hitchcock (2003). See also Schaffer (2004).]
(i) Assassins, A and B, plan to kill victim,
V.
B takes aim and A shouts "Fire!". V overhears this, ducks and so
B misses. Now, did A's shouting
"Fire!" cause V to survive? Had A not shouted, B
wouldn't have fired, and
V
would have lived. Then again, if V hadn't overheard the command, he wouldn't have ducked,
and so would have been killed. So, what is the "balance-tipping" cause here?
(ii) The
probability that a woman will develop thrombosis
is raised by her taking birth control pills, but it is also raised by pregnancy.
So, does the pill cause or prevent thrombosis? If the pill "tips the balance",
which way does it "tip" things?
(iii) It rains heavily throughout April,
thoroughly soaking forest, F. In early May,
lightning hits F, but because it is still damp, the lightning strike fails to
set off a conflagration. By June, F has dried out, so that when lightning strikes again, a
fire breaks out. But, if there had been a dry April, and thus a fire in May,
there wouldn't have been one in June. So, did the April rains cause the fire in June? Was
this a "balance-tipping" cause?142a
[There are many more examples of this sort of
cause given in Hitchcock
(2003).]
It could be objected that the above aren't
relevant since Collier's
theory relates to the causes underlying human action.
Hence, if we are to account for those, a successful theory must address the
material preconditions of an action and explain why a particular course of
events unfolded as it did, as opposed to some other alternative. This would involve, for
instance, a consideration of the factors that "tipped the balance"
leading to NNφ-ing
instead of ψ-ing.
[Again, the
use of Greek symbols like these is explained
here.]
But, example (i) above does in fact concern human action,
as does (ii), indirectly. Anyway, are we really meant to believe that 'balance
tipping' causes only apply to human beings? In that case, what sort of naturalistic explanation of human
action is this that divorces the alleged causes of human action from the course
of events in nature?
(4) The idea that causes are
'balance-tippers' doesn't seem to apply to complex social change, either.
[Which, of course, makes it a useless idea when it comes to
HM.]
Consider, for
example, the question: "What caused the
Confederate States to lose the
American Civil War?" Or, more colloquially: "Why did the Confederacy
lose the Civil War?" [Where this "why" is asking for a reason as a
cause.]
Was it the massive three-to-one superiority in manpower
(not counting African-Americans, i.e., the slave and ex-slave populations), and two-to-one superiority in military manpower
in favour of the North over the South?
Or, was it the vastly superior productive and logistic capacity enjoyed by the Northern
economy? Could it have been the divisive nature of Southern politics, where many States
put their "rights" ahead of the collective war effort? Or, could it have
been the growing
dissatisfaction of Southern non-slave-owners with what came to be seen as a "rich man's war,
poor man's fight"? Maybe
it was the growing opposition to conscription in the South as the war progressed? [This isn't to suggest that some
of these
factors
didn't apply to the Northern war effort; for example, the
widening opposition to the draft (which was given the 'Hollywood Treatment' in Gangs
of New York) as the war progressed into 1863 and 1864.] Then again,
maybe it was the
many slaves who defected to the North? Alternatively, could it have been the fact that the
North developed a
sophisticated telegraph communications network during the war, which allowed
strategic and tactical decisions to be communicated far more
efficiently and decisively? In contrast, the telegraph network in the South
was patchy, at best. The same could be said about the
railway system in the North. Indeed, the rail network in the South was
nowhere near as efficient -- as Wikipedia notes:
"The system was fragile
and was designed for short hauls of cotton to the nearest river of ocean post.
During the war new parts were very hard to obtain, and the system deteriorated
from overuse, lack of maintenance, and systematic destruction by union
raiders.... In addition, the
Confederacy suffered from two key railroad deficiencies. The first was the lack
of a true rail network; instead, rail lines usually connected ports and river
terminals to points inland. This lack of inter-railway connections caused many
railroads to become useless once the
Union
blockade
was in place. A second concern was a
break of gauge; much of
the Confederate rail network was in the
broad gauge format, but
much of North Carolina and Virginia had standard gauge lines. Southern railroads
west of the Mississippi were isolated, disconnected, and differed widely in
gauge....
As troop movement
began in earnest in May and June 1861, a crippling problem was discovered; many
rail lines terminated in towns without connecting to continuing lines.
Instead, cargo would have to be unloaded, driven across town, and then reloaded.
Soldiers, and other passengers, would often have to stay overnight to catch a
continuing train the next day. When the Confederate
government attempted to rectify this problem, they ran into local opposition.
Towns preferred the lack of connection, since it required the hiring of
teamsters and populated hotels with guests. Railroad operators, while not
opposed to connecting lines, were opposed to the possibility of sharing rolling
stock with rival companies."
[Quoted from
here; accessed 16/07/2014. Only one link added;
paragraphs merged.]
Were any of
these the 'balance-tipping' cause?
Or, if we switch to purely military
matters, could it have been the death of
Stonewall Jackson (arguably the Confederacy's most able and effective general) in May 1863 after the
Battle of Chancellorsville? Perhaps it was the North fortuitously finding
Robert E
Lee's
battle
plans before the
Battle of Antietam? Maybe it was
General Ulysses S. Grant's
brilliant Vicksburg
Campaign in 1862-63, which resulted in the fall of that city on July 4th
1863 -- a prize
Jefferson Davis, President of the Confederacy, called "the nail head that
holds the South's two halves together" --, allowing the North to control the
Mississippi
River, and thus the heart of the Confederacy? Maybe it was the failure of
Lt.
General Richard S. Ewell to take
Culp's Hill
on the evening of the first day of the
Battle of Gettysburg, which conceded the higher ground south of the town to the
Army
of the Potomac. Had Ewell's corps taken
that hill, the entire Union Army would have been routed, leaving the capital, Washington, wide open
to Confederate attack. Could it have been Union
General Howard's success in holding Cemetery Hill on the first day of that
battle? Or,
was
it the stoic defence of
Little
Round Top by the 20th Maine led by
Colonel Joshua
Chamberlain on the second day, which prevented the entire
left flank of the Union Army from being turned? Maybe it was
General George Armstrong
Custer's
cavalry charge on the last day of the battle, which prevented a Confederate
flanking move on the Union right? Alternatively, could it have been the
uncharacteristically disastrous tactics Lee employed throughout the
battle (these perhaps being motivated by his overconfidence as a result of the major Confederate victories
at
Fredericksburg and Chancellorsville over the previous six months), the worst of which
being his ordering
of
Pickett's Charge on the last day of the battle, against the advice of his senior commanders?
On the other hand, maybe it was General
J. E. B.
Stuart's
dereliction of duty, which prevented Lee from obtaining
crucial intelligence about the disposition of
General Meade's forces
in the run-up to Gettysburg? Or, could it have been the strategy and
tactics employed by the three most able Northern Generals --
Ulysses
S. Grant,
William Tecumseh Sherman and
Philip
Sheridan -- in the closing two years of the war? Maybe it was a result of the
policy of total war waged by these generals throughout 1864/65 -- best
exemplified, perhaps, by
Sherman's march from Atlanta to the sea in the autumn of 1864, and then his
utter
devastation of South Carolina in early 1865?
Returning to more political
or logistic reasons, could it have been the fact that the
North already possessed a sophisticated bureaucracy capable of running the war,
enlisting men and collecting taxes, while the South had none to speak of, and were forced to
improvise from the start? Then again, maybe it was the political weakness of the South,
which had no identifiable political parties as such, and thus no coherent or overall strategy that
commanded the support of the majority of the Southern population? Alternatively,
perhaps it was
Abraham
Lincoln's masterful leadership of the North, and his administration's
Emancipation Proclamation, freeing the slaves?
Which one of these (and there are
many more) was the
'balance-tipper'?142b
(5) If five men move a piano up
some stairs, which
one is the 'balance-tipping' cause? Do we also include in
this the
stairs, the pulley and the ropes they used? As should seem obvious, this is a
highly simplified example of cooperative labour. Good luck, therefore, to
anyone brave enough to try and find the 'balance-tipping' cause of any of
the following: (a) The construction of the
Pyramids; (b) The building of the
Great
Wall of China; (c) The digging of the
Suez Canal;
(d) The
Apollo Moonlanding (in 1969); (e) The construction of The World Wide Web --
or, indeed, the countless examples of collective labour (big or
small) I am sure the reader can suggest for herself.
However, with respect to the vast majority of
the things people do, this sort of analysis would be misleading, if notincomprehensible.
In order to
make that clear, consider the following example: imagine a certain comrade, NS, say, who has been
in Party, YYYY, for 30 years, who originally joined in order to help build a
revolutionary movement with the aim of changing the world. That was her reason 30
years ago, and let us suppose that this reason has remained unchanged to this day. And yet, if
this reason had been a
"balance-tipping" cause all those years ago, and it remained her reason
during the intervening years, and it is still the reason she would give if asked why she is a revolutionary
today,
what "balanced" items are there left in her mind that still
need
"tipping" thirty years later? If this is still a valid reason for her -- and it is manifestly 'non-tipping' at present
-- how were things
any different 20 or 30 years ago, or at any point in the intervening years? Are we to suppose that this comrade had been teetering on the
verge of giving up revolutionary politics at every single moment during those
thirty years?
Hence, in this clear but familiar case (involving a long term application of
rationality, and the extended direction our lives can take) reasons are
manifestly not causes -- or, at least, not "balance-tipping"
causes.
Again, it could be argued that the desires
and beliefs that induced the change all those years ago were
clearly the cause of comrade NN becoming a revolutionary, which is all
that Collier requires for his modest theory to work. Hence, those factors explain why
NN's life
altered course in the past in the way it did.
Naturally, if that were so, we would no
longer have a general theory connecting reasons to causes -- or, indeed,
any
at all that links natural events to "balance-tipping" situations -- just a
particular fact about a certain individual. Such a retreat would also contain a
damaging admission: if reasons are causes only when they are applied to dramatic
changes in the direction a particular life might take, they couldn't form the
basis of a general theory of human action. So, what are they doing in Collier's
book?
It could be objected once more that the above
"retreat" to a minimalist position (with respect to the identification of
reasons and causes) doesn't imply that reasons are causes only when they
involve major life changes. [Apart from this pro-Collier response merely
amounting to a flat denial, it has little to recommend it. I shall return to
it below.]
On the other hand, it could be argued that
the reasons that operate behind even minor decisions could be, and are in
fact, governed by "balance tipping" causal laws, which, in combination with other laws, might
lend support to the
view that reasons are indeed causes.
However, this is where this picture
begins to lose whatever plausibility it might once even seemed to possess. Consider this example:
NN decides to
turn to page seven of
The Socialist (her copy being the one dated 13/12/2017. She does this at precisely 16:01:36
GMT on Saturday
16/12/17 in
a Café in
Camden after a long and
tedious paper sale. Are we
really supposed to believe that
there is a causal law governing, or constituting, her 'reason' for doing this? If so, it
can't be a general law -- since it applies only to her. But, unless we
now suppose that specific laws
similarly determine other examples of unique (but minor) events and
decisions -- perhaps based on a strict, one-one correlation, so that a new
law of nature is instantiated every time anyone makes the most trivial of
decisions -- no law at all could be operating in this case. Clearly, there couldbe no
causal law that stated that just NN would turn to page seven when sat in just
this café in Camden after a paper sale on just that Saturday, in
just this particular year at precisely that moment in time. Nor
could such a 'law' be generalised to state that anyone else just like her, at the same moment and in
the same place, would do
the same thing -- as any half-way decent law should --, for both her and
the moment are unique. Even if this 'law' predicted (or could have been used to
predict) what NN would do and when she would do it -- ex hypothesi --, it
couldn't apply to anyone else who wasn't identical with her. That is, it
couldn't apply to MM sat in that very same café in Camden after a paper sale on just that Saturday, in
just this particular year at precisely that moment in time unless
MM was identical to NN. But, no one can do exactly what NN did, and be subject
to this 'law', unless two or more individuals can occupy exactly the same
regions of space and time (and perhaps also have identical brain chemistry/world-lines).
In which
case, this 'law' would be a 'law' that applied uniquely to NN at that
precise moment in time, and to no other. But, why should we want to call that a
"law" as opposed to a unique event? And if it isn't a law, how then might we
justify the use of the word "cause" here? How would it be different from
mere
coincidence, or from an
adventitious event?143
It could be argued that the above is a
spurious objection because the said comrade did what she did for no specific
reason, even though it was a voluntary action. However, all we need to do is
alter the details slightly to neutralise this worry. Suppose NN turned to page 7
because she wanted to read a letter she had sent to the paper. The rest then applies as
before.
It could further be
objected that this sort of analysis would in fact mean that no event in
the history of the entire universe would be subject to causal law.144
That is because, on this view, they must all be regarded as unique in some respect or other,
since all events in fact take place at specific times and places and are subject
to one-off individuating circumstances. So, if every event is an
irreducible
particular, then none of them could be governed by a general causal law.
Of
course, events may be grouped in many ways; for example, they can be arranged into
types, and it is this that allows us to say that as a
token
of a set of type events they are governed by this or that law. But, there could
be no type law that applied to just these specific circumstances, and for
the above reasons. Once more, in the example concerning NN, only those
occupying the same regions of space and time, and who were identical to NN,
would be subject to the above 'law' -- even if it were a type law.144a
Even if that were so, no
DM-enthusiast could either appeal to it or accept it, for not only would such a view necessitate
an appeal to the LOI (overtly or covertly: by the explicit or implicit use of
the phrase "same law", "same set of events", or "identically
the same
person"), it would undercut their
commitment to the belief that everything is 'mediated' by everything else in the
entire universe. If everything were 'mediated' in this way, then everything would
be the cause of everything else, and we would have the exact obverse
of the view rehearsed above. And, if that
were the case, we ought to argue that comrade NN turned to the said page at just that
moment (etc., etc.) because of the "Big Bang", and be done with it.
Naturally, such an admission (that events can't be categorised by the use of the word "law") wouldn't be unwelcome
news for anyone who agrees with the views expressed at this site.
If philosophical theories are indeed both non-sensical and incoherent (as is argued in Essay
Twelve Part One), then it makes
no sense to suppose that there are 'laws of nature' (in the
metaphysical sense of that phrase), and thus, a fortiori,
that there are any such that could feature in a metaphysical theory of
the overall course of nature.145
Indeed, this observation would underline the claim made in Essay Thirteen Part
Two that 'laws of
nature' aren't something we discover as a result of a scientific (or
even a philosophical) analysis
of the universe (nor are they something that brute matter 'obeys' of its
own will -- anthropomorphically,
as it were). They belong to various "forms of representation", which enable us to make sense of the universe for
our own purposes. [On "forms of representation", see
Note 29.]
Indeed, far from it being the case
that reasons are causes, it now turns out that causes form part of the
reasons we give for why things happen, just as they form part of a
(hypothetically) valid account of the course of events in nature. Hence,
the former can't be used to explicate the latter without undermining our ability to
explain anything at all.
In that case, it is little wonder that Collier's metaphysical view of causation,
and every other such theory, so readily
collapses into incoherence.
[More on this below.]
Admittedly, it
might seem to some that the above is unfair to Collier, if not itself being dogmatic and
aprioristic. In which case, it could be argued that Collier's account is far
more sophisticated than has been suggested in previous sections.145a
Collier's theory not only covers
other relevant, but much more fundamental, topics, it does so from an overtly scientific
angle.
Or so an objector might argue...
In
that case, it might prove worthwhile to examine Collier's 'see-saw' theory of
the mind in more detail to find out if it is workable at any level.
"What does it mean to say that reasons can be
causes? Bhaskar suggests that:
'When something is cited as a cause it is, I
think, most typically being viewed as that factor which, in the circumstances
that actually prevailed, "so tipped the balance of events to produce the known
outcome"…'…." [Collier (1994), p.152. Quotation marks altered to
conform with
the conventions adopted at this site.]
To state the obvious, a balance
situated in relevant a force field moves when the
moment of one force on one
side of the fulcrum, or centre of rotation (real or theoretical), exceeds that of
any other forces opposing
it on the opposite side -- or on the same side if orientated in an opposite
sense. However, if this picture isn't just another inappropriate metaphor
for the way the mind functions, we need to ask whether there is anything
analogous to a 'fulcrum' situated in the brain or mind of the decisive and the
indecisive alike. So: Does the mind possess anything resembling a 'fulcrum'? Can it
even be pictured as if it did? Do 'mental causes' have moments about this
metaphorical point? Do certain beliefs work on the 'same side' of the supposed
fulcrum as others -- or on 'opposite sides', only in the 'reverse direction'
('clockwise' or 'anti-clockwise', as the case may be)? Can we multiply a cause
by its distance from this 'fulcrum' to obtain the 'moment of a reason'? Does the
mind even have a 'moment of inertia', such that, when set in motion, we can express it
as a sort of metaphorical
vector? If not, what sense can be given to the word
"balance" in this context? Indeed, what content does the verb "to tip" have when
applied to rationalisations?
[At this point, it is worth reminding
ourselves that this way of speaking about the 'mind' harks back to ancient,
mystical Greek ideas
about "rational" and "irrational" 'souls'.]
Of course, we already use the word "balance"
(and other related terms) to depict mental health in general. We say things like
the following:
B1: NP committed suicide while her mind
was unbalanced.
B2: NM has become totally unbalanced;
he now works for
the Tories.
B3: On balance, I think I'll take the
train.
B4: Having weighed things up, MM
decided to join the strike.
And so on. Nevertheless, the question
remains: Is there the least suggestion that there are causal
influences at work here?
Let us suppose that there is just such
a "least suggestion"; indeed, let us further suppose that there is a set of
neural events (or even 'emergent mental processes') that could cause any or all of the
events listed above (if we but knew what they were). In that case, how are we to characterise them? Perhaps
in the
following way?
B5: The set of neural
events, E, caused NN to φ.
B6: The set of mental or psychological
events, M, caused NN to ψ.
[Where
"φ" and "ψ" are dummy letters standing
for verb phrases expressing actions of the relevant sort, such as "run for
a bus", or "reject dialectics as incoherent".]
But, what kind
of analysis of causation is this? If, for example, one or both of
E and
M is/are to be regarded as
Humean-type causes (i.e., either E or
M are shorthand descriptions of regularities, or perhaps are
psychological associations),145b they wouldn't be the
kind of causal
agents Collier had in mind. That is because E and M might be
present and NN could fail to φ
or to ψ.
What Collier needs if his theory
is to work is some sort of natural necessity linking causes to effects
(reminiscent of Kant and Hegel's response to Hume, considered in more detail in
Essay Seven Part
Three) -- and that is precisely what we find:
"It will be clear that
transcendental realism involves a notion of natural necessity that is not
reducible to regular succession. It agrees with commonsense that to say that A
makes B happen is to say more than that A-type events are generally followed by
B-type events…. Here I want to spell out what the 'extra' is, and in doing so
meet the objection that any such extra must be unwarranted by the evidence…. Bhaskar's first marker for this element is
the word "power", which is itself empirical enough -- it merely indicates what a
given kind of thing can do, given the right conditions: dogs can bark,
aeroplanes can fly, cricket balls can smash green houses, and so on. So far, the
objector is going to say, nothing has really been added: how does 'Rectory Ale
sends you to sleep because of its dormative power' differ from 'Rectory Ale
sends you to sleep'?… [T]he stratified nature of explanation, and
the dynamic nature of scientific enquiry, making each result the next matter for
investigation, give empirically justifiable content to the non-empirical part of
causal claims. Effects are ascribed to causal powers, causal powers to the inner
structure…of the causal agent. The 'extra' in the causal power is just this
structure…." [Collier (1994), pp.59-60; paragraphs merged.]
But,
no Humean of any intelligence is going to be impressed with this -- indeed, Hume himself criticised the use of the term
"causal powers" as a way of explicating causation, which Collier has just
re-discovered (as I am sure he is well aware).145c
Nevertheless, the novel feature Collier
introduced here is an appeal to 'micro-states' (i.e., "inner structure(s)") to
explain "powers". Even so, it might well be wondered what it is about such
states that could possibly turn a regular conjunction of events into a 'necessitating' cause. How
does causal necessity emerge from any state of matter -- howsoever large,
small, simple or complex it is? How does any such state bind or direct the future? Collier
doesn't say, and it is doubtful whether he or anyone else can.146
Just in case the above conclusion
should strike the reader as a little too glib, it might be wise to consider the
details a little more carefully, perhaps by means of the following:
B7: A causes B because of
microstate, M1.
But, what is it about
M1
that would necessitateB? Perhaps it is the following?
B8: M1
necessitates B because of a causal law, L1.
But this won't do, since the whole point of
introducing M1 in the first place was to explain causation
and the necessitating power of causal laws. It couldn't do this if it
depends on a causal law itself. We will have just argued in a circle!
Suppose then that we try the following:
B9: L1 is a causal
law because it depicts or involves the operation of force, F1.
But, "force" is just as obscure a
notion as "power" ever was. And we can't now appeal to further micro-states to
account for forces on pain of generating an infinite regress.147
At this point, advocates of
'natural necessity' have only two options left open to them.
(1) They
could appeal to some sort of 'rational' (i.e., in Bhaskar's case,
"transcendental") principle to bridge this gap in their theory -- abandoning the
'empirical' constraint they originally sought and claimed existed. [We saw Kant and Hegel (and then
Lenin) do this in Essay Seven
Part Three.]
Or,
(2) They
could try to
re-define
'natural necessity' as a sort of mathematical, logical, or quasi-logical
'necessity' -- perhaps motivated by the modern (but in fact,
ancient) practice of reifying the mathematical structures scientists
use in order to construct models of the world -- re-christening it, perhaps, "metaphysical necessity"
in order to bamboozle the easily bamboozled.
Unfortunately, the second of these options
isn't viable. Mathematical necessity has no connection to "objective reality" in
this respect. If it had, then plainly we should have another pseudo-explanation.
That is because, if natural necessity were predicated on mathematical
necessity, and it turned out that the latter had been derived
from the natural world itself -- perhaps, by means of the process of 'abstraction'
--
then mathematical necessity would be based on 'natural necessity', not the other way round.
On the other hand, if
mathematical necessity is based on an ideal relationship that is
alleged to exist between ethereal objects in an 'abstract' or 'Platonic Realm', then
the nature of necessity would be even more obscure. That is
because, on this account, the necessary connection that was supposed to exist between material
objects would now be re-configured in terms of a relation between
abstract objects located in this 'abstract' or 'Platonic Realm', which is itself
supposed to mirror the 'natural necessity' operating between material objects in
this world! That would lock this 'explanation' in another circle -- and a
thoroughly mysterious one, to boot. Indeed, if there is already a problem about the
nature of 'natural necessity' in this world, an appeal to even more obscure
forms of metaphysical or mathematical necessity in an invisible world
wouldn't constitute much of
an advance. How exactly one
abstract object can necessarily constrain or command the properties or behaviour
of another still remains to be explained
(unless, of course, we suppose once more that these abstract ideas or objects
are themselves 'mind-like'). Failing that, these Platonic 'necessities' would
have to be seen as merely brute -- i.e., non-rational, and thus contingent -- facts about ethereal 'objects'
like these, or
even about how we conceive them. But, the whole point of this
retreat into the Ideal was to locate the various 'essences' that explained the
brute contingency we supposedly see in material reality, as a disguised form of
'natural' necessity, in order to short-circuit an appeal to Humean contingency!
If so, it would imply that such 'abstract' or Platonic 'objects' are just
as contingent and brute as their this-worldly correlates ever were --, which
disconcerting fact
undercuts any reason we might have had to go beyond the material world
in order to account for
contingency in the first place!
In short, an appeal to mathematical necessity would be of little
help in explaining 'natural necessity' if it turns out that it relies on a mystical
correlate of it.
In connection with this, the reader is
re-directed to a passage that was quoted earlier:
"Empirical,
contingent
truths have always struck
philosophers as being, in some sense, ultimately unintelligible. It is not that
none can be known with certainty…; nor is it that some cannot be explained….
Rather is it that all explanation of empirical truths rests ultimately on brute
contingency -- that is how the world is! Where science comes to rest in
explaining empirical facts varies from epoch to epoch, but it is in the nature
of empirical explanation that it will hit the bedrock of contingency somewhere,
e.g., in atomic theory in the nineteenth century or in
quantum mechanics
today. One feature that
explains philosophers' fascination with
truths of Reason
is that they seem, in a
deep sense, to be fully intelligible. To understand a necessary proposition is
to see why things must be so, it is to gain an insight into the nature of
things and to apprehend not only how things are, but also why they cannot be
otherwise. It is striking how pervasive visual metaphors are in philosophical
discussions of these issues. We see the universal in the particular (by
Aristotelian intuitive induction); by the Light of Reason we see the essential
relations of
Simple Natures; mathematical truths are
apprehended by Intellectual Intuition, or by
a priori insight. Yet instead of examining the use of these arresting
pictures or metaphors to determine their aptness as pictures, we build
upon them mythological structures.
"We think of necessary
propositions as being
true or false, as objective and independent of our minds or will. We
conceive of them as being about various entities, about numbers even
about extraordinary numbers that the mind seems barely able to grasp…, or about
universals, such as colours, shapes, tones; or about logical entities, such as
the
truth-functions or (in Frege's
case) the
truth-values. We naturally think of necessary propositions as
describing the features of these entities, their essential characteristics.
So we take mathematical propositions to describe mathematical objects…. Hence
investigation into the domain of necessary propositions is conceived as a
process of discovery. Empirical scientists make discoveries about the
empirical domain, uncovering contingent truths; metaphysicians, logicians and
mathematicians appear to make discoveries of necessary truths about a
supra-empirical domain (a 'third
realm'). Mathematics seems to be the 'natural history of
mathematical objects' [Wittgenstein
(1978), p.137], 'the physics of numbers' [Wittgenstein (1976), p.138; however
these authors record this erroneously as p.139 -- RL] or the 'mineralogy of
numbers' [Wittgenstein (1978), p.229]. The mathematician, e.g.,
Pascal,
admires the beauty of a theorem as though it were a kind of crystal.
Numbers seem to him to have wonderful properties; it is as if he were
confronting a beautiful natural phenomenon [Wittgenstein (1998), p.47; again,
these authors have recorded this erroneously as p.41 -- RL]. Logic seems to
investigate the laws governing logical objects…. Metaphysics looks as if it is a
description of the essential structure of the world. Hence we think that a
reality corresponds to our (true) necessary propositions. Our logic is
correct because it corresponds to the laws of logic….
"In our eagerness to ensure
the objectivity of truths of reason, their sempiternality
and mind-independence, we slowly but surely transform them into truths that are
no less 'brutish' than empirical, contingent truths. Why must red exclude
being green? To be told that this is the essential nature of red and green
merely reiterates the brutish necessity. A proof in arithmetic or geometry seems
to provide an explanation, but ultimately the structure of proofs rests on
axioms. Their truth is held to be self-evident, something we apprehend by
means of our faculty of intuition; we must simply see that they are
necessarily true…. We may analyse such ultimate truths into their constituent
'indefinables'. Yet if 'the discussion of indefinables…is the endeavour to see
clearly, and to make others see clearly, the entities concerned, in order that
the mind may have that kind of acquaintance with them which it has with redness
or the taste of a pineapple' [Russell
(1937), p.xv (this links to a PDF); again these authors record this erroneously as p.v;
although in the edition to which I have linked, it is p.xliii -- RL], then the
mere intellectual vision does not penetrate the logical or metaphysical
that to the why or wherefore…. For if we construe necessary
propositions as truths about logical, mathematical or metaphysical entities
which describe their essential properties, then, of course, the final products
of our analyses will be as impenetrable to reason as the final products of
physical theorising, such as
Planck's constant."
[Baker and Hacker (1988), pp.273-75. Referencing conventions in the original
have been altered to conform with those adopted at this site. Italic
emphases in the original; links added.]
Hence, it
seems that if we scratched a devotee of 'natural necessity' hard enough, we would invariably
uncover a Pythagorean/Platonist, a Platonic Realist (i.e., a non-slippery,
but more honest,
Hegelian), or even a
Kantian Transcendental Idealist
lurking below the surface. Indeed, that is precisely what we find in the
writings of Roy Bhaskar, Andrew Collier and other 'Critical Realists'. But, as the
above confirms, even these rationalists have to fall back on brute facts
(contingencies) at some point to account for their assumed 'rational' principles, vitiating
the whole exercise. It is just a brute fact about us, so they allege, that we
think along these lines, with these concepts and categories. Of course, they
might be tempted to present us with 'rational' or 'necessary' principles that
inform us that we can think in no other way, but even that will be a
brute fact about how we (or, rather, they) use certain words -- and
they would be
distorted words, at that!
Of
course, Collier seems to be aware of this criticism, for he says:
"…[N]atural necessity must surely, exist, if
anywhere, in things independent of us, yet it is being presented as supplied
entirely by the human mind…. If the real structure of nature, and its consequent
necessities, do not make…structured theories essential to their explanation, the
structured theories must be more or less gratuitous. We must either retreat to
the flatlands of Humean succession of impressions, or advance to a theory of
real structures generating real necessities." [Collier (1994), p.61.]
However, apart from
a bluff rejection
of Hume's account of causation -- along with a veiled reference to the alleged
'gratuitousness' of
Conventionalism
-- Collier offers little (or no) justification for his introduction of the
phrase "natural necessity", and neither does he succeed in dispelling the
suspicion that the latter is the product of yet another reification of language.
As noted earlier, such a
fetishisation of
discourse only seems to work because it
misconstrues the products of the social relations among
human beings as the real relations between things, or as those things
themselves. Integral to this archaic and 'time-honoured error', linguistic categories are projected back onto
reality with specialist terminology (drawn from human interaction) thrown in for good
measure -- such as: "necessity", "law", "obey", "rational", and, in the case of dialectics,
"mediation", "contradiction" and "negation", etc.,
etc.
Because of this, it is no surprise to see human traits staring back at us from
the Ideal world constructed by means of this fetishisation! It is even less of a surprise to see this in the work of
modern-day Hegelians (be they from the upside down fraternity or the 'right way up'
tendency), since Hegel turned word-magic like this into an art
form. [More on this in Essay Three
Part One.]
To be sure, Collier motivated his argument by
dismissing the relevant theories found in Empiricism and
Idealism (pp.70-106), but there is nothing else in
his book (as far as can be ascertained) that presents even a weak case
for
the acceptance of "natural necessity".
Far worse, even if he were correct, Collier
offers little to help the bemused reader understand what the phrase "natural
necessity" could possibly mean. Is he serious (is anyone who
has ever
used this term serious?) that objects and processes force
other objects and processes to do their bidding? Is every particle in nature
at once both a
ruthless tyrant and docile slave, bully and victim rolled into one? Do they
impose
themselves on one another, but then meekly fall into line whenever this is done in
return to them,
as the various causal chains in nature and society unfold?
Does each and every object and process involved 'understand' its role in this intergalactic,
sadomasochistic dance? If not, why use confusing, anthropomorphic concepts (such
as "power", "force", "law", "rational" and "obey") in such contexts,
and pretend to mean anything literal
by them?
Hence, Collier's attempt to re-package
Bhaskarean 'transcendental realism' fails miserably. We are still
owed an account of 'natural necessity' that goes beyond mere phrase-mongering --
that is, we are owed an explanation that waves "goodbye" to the constant conjunction
of the words "natural" and "necessity" by those who have developed a distinctly Humean Habit
of talking this way.
At this point, the
unconvinced among us might be forgiven for wondering what all this has to do with an
analysis of mental causation. If E or M (from
earlier) caused NN to φ,
or to ψ,
then how would that explainNN's actions? Alas, given the account under
review here, it couldn't. That is because we still don't possess a
'philosophical' account of causation that isn't itself based on
the projection of
intelligent aims and intentions onto nature -- that is, one that isn't
predicated on yet another fetishisation
of language -- and which doesn't thereby treat the products of social interaction as if they
represented the real relations between things.148
Even if
Collier's account were acceptable, his theory of causation couldn't be explained without an appeal to "structural"
principles (at whichever micro-level his
Scientific Realism finally gives out
--
i.e., at the level where 'realists' finally have to throw their hands in the air
and appeal to the "brute facts" of nature,
mentioned in the long quotation above). But, "structural" principles can't themselves necessitate
anything. As noted earlier, to suppose otherwise would commit us to the idea
that matter is alive and sentient (and thus is capable of volitional action, being able
to
'understand' and 'obey' the laws that govern it -- which, naturally, would
picture matter as an agent in its own right). Either that, or it would require
an acceptance of the belief that
natural laws
themselves can somehow 'coerce matter' into obeying their dictates (thus turning
'laws' into agents). On both accounts, this would suggest that events
are governed by some sort of Intelligence or Will. From this it is but a short step to
the circular argument that human action has been explained by projecting
human capacities back onto reality -- as we saw in relation to the
Homunculus Fallacy, above.
In the end, this theory implies that
reasons are causes because causes are reasons -- that is, it is based on
the idea that every particle in nature recognizes and understands the
causal law that governs it and acts in accord with its dictates, and with good reason,when it
complies with, and 'obeys', that causal law! Either that, or it implies that reason itself has been
hypostatised
in nature so that 'it' can induce causation by the 'sheer power of its will'.
But, how is it able to do this? By the strength of its personality? Or, perhaps
by its ability to change things and move them about the place by the use of magical words? Can 'reason'
simply materialise and push things
around in a
seemingly intelligent manner?148a
The irony is that although Collier's theory
seems to be committed to a view of 'consciousness' that depends on the existence
of an intelligent agent in our heads to whom things are 'represented', in reality it is based on an even more
ancient idea: that
intelligence and rationality is found throughout nature, and not just inside our skulls!
The universe is thus OneBig
Mind, and causation represents the operation of itsReasononourinner
'reason'.
Again,
as above, so below.
The microcosm isn't just a reflection of the macrocosm, it is controlled by
it!
This, of course, accounts for the
ever-present
Bhaskarean appeal to 'transcendental principles' to knit his artificial system together;
since matter can't act with reason, reason must lend to matter its active
"powers". Nature is
re-enchanted thereby. The metaphysicalrequirement
for there to be some sort of 'natural necessity' running the show thus becomes the mother of its
own invention -- as, indeed, Collier argued:
"…[N]atural necessity must surely, exist, if
anywhere, in things independent of us, yet it is being presented as supplied
entirely by the human mind…. If the real structure of nature, and its consequent
necessities, do not make…structured theories essential to their explanation, the
structured theories must be more or less gratuitous. We must either retreat to
the flatlands of Humean succession of impressions, or advance to a theory of
real structures generating real necessities." [Collier (1994), p.61.]
None
of this is the least bit surprising
since Collier has appropriated far too many ideas from Bhaskar, who in turn sold his radical soul long ago to
that arch-mystic, Hegel. Hence, it is hardly surprising, either, to see Bhaskar return to the
Mystical Swamp from whence his
mentor's ideas originally slithered.149
Once more, we can see that the alleged
materialist flip Marxist dialecticians say they have performed on Hegel's system
was all
smoke and
mirrors.
Despite what they say, it would seem that Critical
Realists like Collier have yet to come to terms with Hume's analysis of
causation. Without an appeal to the existence of infinite clusters of hidden
intelligences programmed to run like mini-computers (somewhat analogous to the
Monads
of
Leibniz's
system), or perhaps without reference to 'The Big Idea' that runs the show (akin to
the 'development' of Hegel's
'Absolute' -- or, in the form it now assumes as the NON in DM) --,
Critical Realists finds they have no account of causation capable of generating
the required necessity short of using the (by-now-familiar and powerful) magictrick of putting a few 'scare quotes' around a handful of
ordinary words, inventing yet more jargon, drawing several intricate
diagrams and flow charts, and just toughing
things out.
[NON = Negation of the
Negation.]
To be
sure, Collier's exposition of Hume (whose
theory was itself based on an implausible
version of
associationist psychology
and an un-workable atomistic theory of language) only succeeds in undermining the
epistemological foundations of Hume's theory of causation; it doesn't
affect the latter's overall case against the explanatory capacity of "powers",
"forces" and "necessitating causes". In this regard, however, it is rather odd that Collier failed to
mention (let alone discuss) contemporary versions of Hume's theory, or any
of the 'ordinary language' accounts of causation he indirectly derides. Both
alternatives show that if necessity originates anywhere, it arises from
the language we use to formulate explanations, models and theories of the world,
not from nature itself. Collier pointedly failed to address this
awkward, but highly salient, fact.150
What then are we to make of Collier's claim that reasons are causes (as part of
his response to
Wittgenstein's alleged insistence to the contrary)? It now turns out that
there is little reason to accept either his or Bhaskar's analysis of causation,
and every reason to reject them both -- as confused at best, incoherent and anthropomorphic, at worst.
For it
now seems clear that if, given Bhaskar and Collier's views, reasons are causes, and causes are themselves
underpinned by 'natural necessity', then causes are in fact reasons in disguise.
Of course, this means that until we are given
good reason not to do so, Bhaskar's unbalanced theory needs
tipping into a dumpster.151
I don't propose to respond to Trevor Pateman's criticisms of Wittgenstein in the
present version of this Essay. Much of what
he has to say depends on accepting Noam Chomsky's implausible, dogmatic,
fanciful and
quasi-metaphysical theory that there exists a "language acquisition device"
[LAD], which works in tandem with "transformational grammar"
(now replaced by "unbounded
Merge")
underlying discourse -- alongside a host of
other "likely stories".
Geoffrey Sampson has recently written a
clear and accessible demolition of some of the central (empirical and conceptual)
foundations of Chomsky's theory (as well as those of other 'nativist' theorists, like
Bickerton,
Fodor and
Pinker).151aFiona Cowie
has done likewise with the alleged philosophical and scientific support for
Nativism.152
This includes the specific areas that seem to impress Collier the most -- i.e., those
connected with the
creolisation of
pidgin dialects and the seemingly miraculous
linguistic capabilities of children born to deaf parents, to which Pateman also
refers in order
to 'refute' the so-called Wittgensteinian
private language argument [PLA].153
Nevertheless, it isn't with
empirical issues such as these that I wish to take issue...
[I will add more details
here in a later re-write of this Essay.]
This process was analysed in more detail in
Essay Twelve Part One and Part
Two (not yet published -- a summary of which can be found
here).
It is
possible to trace this notion (i.e., that eternal verities can be derived from
thought alone) at least as far back as
Plato who,
because he was an arch traditionalist, would surely have been retailing ideas that were already in
circulation (see below). In the
Theaetetus, for example, he argued as follows:
"Soc. But must not the mind,
or thinking power, which misplaces them, have a conception either of both
objects or of one of them?
"Theaet. Certainly.
"Soc. Either together or in succession?
"Theaet. Very good.
"Soc. And do you mean by conceiving, the same which I mean?
"Theaet. What is that?
"Soc. I mean the conversation which the soul holds with herself in considering
of anything. I speak of what I scarcely understand; but the soul when thinking
appears to me to be just talking -- asking questions of herself and answering
them, affirming and denying. And when she has arrived at a decision, either
gradually or by a sudden impulse, and has at last agreed, and does not doubt,
this is called her opinion. I say, then, that to form an opinion is to speak,
and opinion is a word spoken, -- I mean, to oneself and in silence, not aloud or
to another: What think you?
"Theaet. I agree."
[Plato
(1997b), p.210. However, I have used Benjamin Jowett's on-line version, not
the one found in Plato (1997b). Bold emphases added.]
Similar ideas were expressed in another of his
dialogues,
the
Sophist (in this case, the exchange is between the
Eleatic Stranger and
Theaetetus himself):
"Str. And therefore thought, opinion, and imagination are
now proved to exist in our minds both as true and false.
"Theaet. How so?
"Str. You will know better if you first gain a knowledge of what they are, and
in what they severally differ from one another.
"Theaet. Give me the knowledge which you would wish me to gain.
"Str. Are not thought and speech the same, with this exception, that what is
called thought is the unuttered conversation of the soul with herself?
"Theaet. Quite true."
[Plato
(1997c), pp.287-88.
Once again, I have quoted Benjamin Jowett's on-line version, not Plato (1997c). Bold emphases added.]
[Details concerning other Ancient Greek
thinkers (including Aristotle) who took a similar line can be found in Sorabji
(2004), pp.205-26. For Aristotle's theory of language and meaning, see Modrak
(2009). This isn't to suggest, of course, that
Aristotle's theory of 'the
mind' is the same a Plato's; far from it. See also the opening pages of Hacker
(2012)
(quoted here),
a passage I have quoted
from Hacker (2013a), and a video of Hacker on this topic,
posted
here.
This
subject is
also connected with the idea
that there is something called "non-propositional thought", a
notion beloved of mystics the world over -- the belief that the 'soul' can
directly apprehend
esoteric, or
'ineffable,
truths' about 'God' and 'Ultimate Reality' without the mediation of
language, which meant that this could be done 'free' of social constraint,
especially that exercised by the 'lower orders'.
[The Eleatic Stranger above doesn't appear to believe this, though. On this, see also Sorabji (2005), pp.90-93, Sorabji (1982), and
Alfino (1988) (which is a response to Sorabji (1982)). On 'Divine Illumination', see
Pasnau
(2015).]
For such mystics, language
is only
able to "intimate" these 'unfathomable truths' indirectly:
"One of the best-known facts
about mystics is that they feel that language is inadequate, or even wholly
useless, as a means of communicating their experiences or their insights to
others. They say that what they experience is unutterable or ineffable. They use
language but then declare that the words they have used do not say what they
want to say, and that all words as such are inherently incapable of doing so.
"According to the
Mandukya Upanishad
the unitary consciousness is 'beyond all expression'.
According to
Plotinus,
'the vision baffles telling.' In a passage which I shall quote more at length
later,
Eckhart
says that 'the prophets walking in the light...sometimes were moved
to...speak of things they know...thinking to teach us to know God. Whereupon
they would fall dumb, becoming tongue-tied.... The mystery they found there was
ineffable.'
"And modern Europeans and Americans who report having had mystical experiences
feel the difficulty just as much as do the ancient or classical mystics.
R.
M. Bucke
says that his experience was 'impossible to describe'.
Tennyson says that his was 'utterly beyond words'.
J.
A. Symonds
states that he 'was not able to describe his experience to
himself' and that he 'could not find words to render it intelligible'.
Arthur
Koestler
says of his experience that 'it was meaningful though not in verbal
terms' and of his own [p.278] attempts to describe it that 'to communicate what
is incommunicable by its nature one must somehow put it into words, and so one
moves in a vicious circle.' Probably hundreds of similar statements could be
collected from all over the world." [W T Stace, Mysticism and Philosophy,
quoted from
here. Quotation marks altered to
conform with the conventions adopted at this
site. Links added.]
Indeed, Hegel's system itself was ultimately based on esoteric flights-of-fancy
like these, masquerading as some sort of 'dialectical logic', the contemplation of
which supposedly led the human 'soul' into mystical union with 'Being'/'God'.
[This
ancient notion later resurfaced in Engels's work in what appeared to be an outwardly secular form (on
that, see
here and Note 1a, below).]
Indeed, the
esoteric language Hegel himself used
suggests, to those who dote on jargon like this, that there are 'ineffable
truths' that lie beyond language, toward which Hegel and those who follow him
can only hope to grope, and which can be grasped (in howsoever an attenuated
form) by those who 'understand' dialectics -- i.e., exclusively those
capable of contemplating these 'ineffables' directly via the 'concepts' to which
Hegel alluded, as opposed to the actual words he used, but by no one else.
Of course, that is only
to be expected of
genuine
mystics, or, indeed, expected of the aspiring DM-Mystics who litter
Dialectical Marxism.
This helps
explain why so many dialecticians express a liking for
Daoism and/or
Buddhism, and why some even revert to open and honest mysticism (Michael
Kosok, Roy
Bhaskar,
Andrew Collier, Graham Priest, and
Terry Eagleton perhaps being the latest examples). Indeed, the tendency
for radicals to become mystics is what motivated Lenin to write
MEC. This also
explains why Counterfire, home of leading Dialectical Marxists and
ex-members of the UK-SWP, can publish an article extolling the virtues of
Buddhism, even roping Marx himself in this! On that, see
Ledwith (2023). It
also explains why dialectician extraordinaire,
Graham Priest, accepts
some form of Buddhism. On that, see here
and Priest (2014).
Then there is this unfortunate
video about Buddhism
and Science, by Priest. In addition, he has written about
the use of "contradictions" in Buddhism; cf., this
on-line article by Yasuo Deguchi, Jay
Garfield and Graham Priest (this links to a PDF). No wonder he also has a
penchant for Hegel and the latter's dalliance with "contradictions".
[The is more on this
below, as well as in Essays Nine
Part Two and Fourteen Part One
(a summary of the latter can be found
here).
I have said more about the connection between Daoism and 'dialectics' in Essay
Three Part Two. A detailed
survey of the close link that exists between Daoism and Maoist 'Dialectics' can
now be accessed here.]
This
is something that Voloshinov himself acknowledged, without perhaps fully
realising the full significance of that admission:
"There was good reason why
thinkers in ancient times should have conceived of inner speech as inner
dialogue...." [Voloshinov (1973, p.38. Italic emphasis in the
original.]
There is a useful summary of 'western',
mystical ideas about the 'soul', and the links this time-worn tradition enjoys with
contemporary 'science of
mind' in Leahy (2005). In fact, concerning the claims advanced above,
Leahy had this to
say:
"Western conceptions of
mind began in religion before moving first to philosophy, and then to science.
However, for two reasons psychologists have underestimated the influence of
religious ideas of the soul -- the ψυχή
(psychē) of our science -- on conceptions of mind and self. First, psychology is an
aggressively secular enterprise and psychologists like to think that they put
religion behind them when they assume their role as scientists. A more subtle
reason concerns the dominance of historical scholarship by Christian belief.
When we as psychologists read about past thinkers such as Plato and
Descartes,
not only do we look at them as protopsychologists, we see them through the eyes
of historians and classicists who until recently worked within a quietly but
unequivocally held Christian framework. That framework rarely intrudes
explicitly, but it filters out the rough splinters, odd conceptions, and obscure
but vital disputes concerning mind and soul held from Greek times through to at
least Descartes. Thus we psychologists inherit a conception of the mind subtly
shaped by forces of which we know little, drain it of its specifically
supernatural content (e.g., survival of bodily death), and fancy that what
remains is somehow natural and therefore a proper object of science....
"Although there are differences in
detail, religions around the world have a remarkably concordant picture of the
mind, positing the existence of two immaterial souls for two distinct
reasons.... The first, universal reason is to explain the difference between
living and nonliving things. The second, less universal reason is to
explain human personality....
"Greek religion and the concept of
ψυχή
underwent a profound change in the later fifth century BCE.... Traditional Greek
religious thought had insisted on a great gulf between the human and divine
worlds, downplaying the idea of personal immortality. However, in the wake of
the
Peloponnesian War, continuity between the human and divine worlds was the
theme of various new cults, often imported from the non-Greek east. In their
practices these new religions induced in worshippers ecstatic states through
which they might for a time join the gods, perhaps even briefly becoming the god
of their veneration. The ψυχή
became a personal, immortal soul, taking after death its rightful place in the
divine world of the gods. Plato was influenced by these new teachings, but
steered them in a less ecstatic, more philosophical and cognitive direction....
For Plato, the proper object of the soul's attention was indeed something
divine, but he taught that instead of seeking salvation through ecstatic
communion with the gods, the soul should seek salvation through philosophical
pursuit of eternal, transcendental Truth. In Plato's hands, the mind became
identified with reason, the ability to formulate and know the universal Truths
underwritten by the heavenly Forms." [Leahy (2005), pp.37-39. Bold
emphases added.]
"There are few names to which more diverse persons and disciplines lay claim
than the term 'Hermetic.' Alchemists ancient and contemporary apply the
adjective 'Hermetic' to their art, while magicians attach the name to their
ceremonies of evocation and invocation. Followers of
Meister
Eckhart,
Raymond Lull,
Paracelsus,
Jacob Boehme,
and most recently
Valentin
Tomberg
are joined by academic scholars of esoterica, all of whom attach the
word 'Hermetic' to their activities.
"Who, then, was Hermes, and what may be said of the philosophy or religion that
is connected with him? The early twentieth-century scholar
Walter
Scott, in his classic edition of the Hermetic texts, writes of a legend
preserved by the Renaissance writer Vergicius:
'They say that this Hermes left his own country
and travelled all over the world…; and that he tried to teach men to revere and
worship one God alone, …the
demiurgus
and genetor [begetter] of all things; …and that he lived a very wise and pious
life, occupied in intellectual contemplation…, and giving no heed to the gross
things of the material world…; and that having returned to his own country, he
wrote at the time many books of mystical theology and philosophy.'
"Until relatively recently, no one had a clear picture of either the authorship
or the context of the mysterious writings ascribed to Hermes. Descriptions such
as the one above are really no more than a summary of the ideal laid down in the
'Hermetic' writings. The early Christian Fathers, in time, mostly held that
Hermes was a great sage who lived before Moses and that he was a pious and wise
man who received revelations from God that were later fully explained by
Christianity. None mentioned that he was a Greek god....
"The British scholar R. F. Willetts wrote that 'in many ways, Hermes is the most
sympathetic, the most baffling, the most confusing, the most complex, and
therefore the most Greek of all the Olympian gods.'
If
Hermes
is the god of
the mind, then these qualities appear in an even more meaningful light. For
is the mind not the most baffling, confusing, and at the same time the most
beguiling, of all the attributes of life?...
"Hermes became best known as the swift messenger of the gods. [Later known as
the
Roman 'god' Mercury
-- RL.]
"Hermes is thus of a double origin. His grandfather is
Atlas, the demigod who
holds up heaven, but
Maia, his mother, already has a goddess as her mother,
while Hermes' father,
Zeus, is of course the highest of the gods. It is tempting
to interpret this as saying that from worldly toil (Atlas), with a heavy
infusion of divine inspiration, comes forth consciousness, as symbolized by
Hermes.
"Versatility
and mutability are Hermes' most prominent characteristics. His specialties
are eloquence and invention.... The common quality in all of these is again
consciousness, the agile movement of mind that goes to and fro, joining humans
and gods, assisting the exchange of ideas and commercial goods....
"While
Hermes is regarded as one of the earliest and most primitive gods of the Greeks,
he enjoys so much subsequent prominence that he must be recognized as an
archetype devoted to mediating between, and unifying, the opposites. This
foreshadows his later role as master magician and alchemist, as he was regarded
both in Egypt and in Renaissance Europe....
"The
Greek Hermes found his analogue in Egypt as the ancient Wisdom God
Thoth
(sometimes spelled Thouth or Tahuti)....
"Thoth
played a part in many of the myths of Pharaonic Egypt: he played a role in the
creation myth, he was recorder of the gods, and he was the principal pleader for
the soul at the judgment of the dead. It was he who invented writing [and
philosophy
-- RL]....
"Most
importantly, perhaps, for our purposes, Thoth acted as an emissary between the
contending armies of
Horus
and
Seth
and eventually came to negotiate the peace treaty between these two gods. His
role as a mediator between the opposites is thus made evident, perhaps
prefiguring the role of the alchemical Mercury as the 'medium of the conjunction.'...
"It
was to this powerful god that the Egyptian Hermeticists of the second and third
centuries A.D. joined the image and especially the name of the Greek Hermes.
From this time onward the name 'Hermes' came to denote neither Thoth nor Hermes
proper, but a new archetypal figure,
Hermes Trismegistus, who combined the features of both....
"From
a contemporary view, the figure of Hermes, both in its Greek and its Egyptian
manifestations, stands as an archetype of transformation through reconciliation
of the opposites. (Certainly
Jung
and
other archetypally oriented psychologists viewed Hermes in this light.) If we
are inclined to this view, we should rejoice over the renewed interest in Hermes
and his timeless gnosis. If we conjure up the famed image of the swift god,
replete with winged helmet, sandals, and
caduceus, we might still be able to ask
him to reconcile the divisions and contradictions of this lower realm in the
embrace of enlightened consciousness." [Quoted from
here, minor typo
corrected; accessed 10/10/2013. Quotation marks altered to conform with the conventions
adopted at this site. Bold emphases and links added.]
This tradition therefore linked human
thought (or, rather, the theories and musings of boss-class ideologues and mystics) with the 'Mind'
of 'God', via 'consciousness', philosophy, and esoteric
'contemplation'.
In connection with these new doctrines,
Alfino (1988) exposes the link that exists between (i) the
Identity Theory of Predication [IDP] (see below),
(ii) "non-propositional thought" and (iii) the
mystical ideas that emanated forth from the works of the leading
Neoplatonist,
Plotinus,
second only to Hegel himself as this entire tradition's
most
influential ideologue.
As Sara Rappe
notes:
"Plotinus anticipates
Descartes in arguing both that the soul as subject of perception cannot be an
extended substance, as well as in arguing that the mind necessarily knows
itself. Like Descartes, Plotinus also invokes an introspective or subjective
stance within his dialectical procedure." [Rappe (1999), p.250. Bold
emphasis added.]
[The
rest of Rappe's article explores the close affinities that exist between
Descartes and Plotinus.]
Not only was Hegel profoundly influenced by
these
Neoplatonists, he also imported the IDP from Medieval theologians and then
used it to motivate his 'theory' that 'contradictions' and 'negation' power the development of
'Being'.
[This was discussed in detail in Essays Three
Part
One and Eight Part Three;
see also
here.]
In fact, two out of the three volumes of
Hegel's
Lectures on the History of Philosophy were devoted to Ancient Greek
Thought -- one
volume of which concentrates on Plato and the Neoplatonists. [Hegel (1995a, 1995b);
Hegel's comments on Plotinus appear in Hegel (1995b), pp.404-431.]
As noted above, the
speculative trope about the "mystical union" of the 'mind',
or 'soul', of the
philosophical Knower and the Platonic (or, later,
Hermetic
'Unknown'),
subsequently re-surfaced as the 'problem' of "Subject/Object
Identity", which became the main problematic of
German Idealism.
Rappe again:
"For
Plotinus, any representation of the self
or subject of consciousness can never be complete and can never succeed in
conveying the self that it purports to represent. The fallibilism of any such
conveyance is a consequence of Plotinus's more general theory of knowledge
according to which
truth cannot be ascertained by means of linguistic or
conceptual representations. It can be apprehended only when there is an identity
between the
knower and the
known." [Ibid., p.253. Bold emphasis added.]
Knowledgeable readers will
no doubt have noticed the similarity between this form of Western
Mysticism and Eastern, meditative religions and philosophies -- for instance,
this
and
this --, which isn't surprising in view of the fact that these ideas
originated in the East (as Leahy, for
example, pointed out).
As this Essay proceeds the
following themes will unfold:
(1)
Concepts connected with
Platonic-Christian-Cartesian theories of 'Mind', or 'Soul', have come to dominate western thought
(and, as it turns out, 'Materialist Dialectics', too), albeit expressed in diverse idioms at different times;
(2) The doctrine that 'thought'
precedes language -- which idea has resurfaced these days in
Cognitive Science -- coupled with the notion that there can be such a thing
as 'languageless thought', is an archaic dogma that still mesmerises modern
theorists -- and that includes DM-theorists, too.
[These claims might
at first sight
seem somewhat controversial, but they will be substantiated throughout this Essay.]
So, once again, we see another set of
"ruling ideas", which have dominated, and still dominate, thought 'East' and
'West', confirm Marx's assessment:
"The ideas of the ruling
class are in every epoch the ruling ideas, i.e. the class which is the ruling
material force of society, is at the same time its ruling intellectual force.
The class which has the means of material production at its disposal, has
control at the same time over the means of mental production, so that thereby,
generally speaking, the ideas of those who lack the means of mental production
are subject to it. The ruling ideas are nothing more than the ideal expression
of the dominant material relationships, the dominant material relationships
grasped as ideas; hence of the relationships which make the one class the ruling
one, therefore, the ideas of its dominance. The individuals composing the ruling
class possess among other things consciousness, and therefore think. Insofar,
therefore, as they rule as a class and determine the extent and compass of an
epoch, it is self-evident that they do this in its whole range, hence among
other thingsrule also as thinkers, as producers of ideas, and regulate the
production and distribution of the ideas of their age: thus their ideas are the
ruling ideas of the epoch." [Marx
and Engels, The German Ideology, pp.64-65. Bold emphases added.]
[Exactly how
such "ruling ideas" are
connected with the 'justification' and rationalisation of class power will be
explored in Essays Twelve and Fourteen
(summaries here and
here).]
[For Hegel's ideas (which are related to the mystical
union of the 'soul' with 'God'/'Being'), see
Magee (2008).
For the Greek end of this sorry tale, see
Burkert (1972, 1985), Guthrie (1950), Onians (2000), Otto (2001), Owen (1966/86) --
particularly, pp.207-11 (i.e., of the 1986 version) --, Robinson (2008), pp.1-16, and Snell (1982).
On Plotinus, see Armstrong (1970b), Clark (1999), and Gerson (1994). On Augustine, see Copleston (2003), pp.40-90,
Markus (1970) and Teske (2001).
The best general on-line articles on this are
Lorenz
(2009), and
Shields (2016).]
[On the more recent origin of
'psychological jargon' see Danziger (1997) -- although, Danziger
doesn't
put things this way! On
this theme (which runs through the entire course of German Mysticism (oddly enough, including Wittgenstein's early
work -- however, on Wittgenstein's alleged 'mysticism', see
here)), cf., Weeks (1993). For the social origin and function of these myths,
see Harrison (1989) and Vernant (1990). The verbal tricks performed by Ancient Greek
Philosophers in order to concoct their fanciful theories are exposed in Barnes (2009),
Havelock (1983), Kahn (1994, 2003), Lloyd (1971), and Seligman (1962) --
although, these authors don't put things this way, either! This strand of 'western' ruling-class thought is
traced from Ancient Greek times up to the
Enlightenment in Wright and Potter (2000).]
Of course, it is important to
emphasise that there certainly were other ways of conceiving of the 'soul' in
Ancient Greece, theories which
were incompatible with the Platonic tradition. The theory developed by
Aristotle and subsequent Aristotelian
philosophers' (also known as "hylomorphism"
-- where the 'soul' was held to be the 'form' of the body, etc.) were quite distinct.
But this theory created almost insurmountable difficulties for Christian Aristotelians
(like
Thomas
Aquinas) when they attempted to reconcile it with Church Dogma, and who
had to face frequent accusations
of "heresy" as a result.
[On Aristotle's views see Everson (1995), and Guthrie (1990),
pp.277-330.
On the controversies in the Roman Catholic Church, see Copleston (2003); on
Aquinas, see Kenny (1994), and Kretzmann (1993). (See also
here.) It is clear, however, that the Platonic view of the 'soul' in the end
'won the day', Pasnau (2002).]
It is also important to point
out that the easy equation of words for 'Mind' and 'Soul' in fact distorts the
complex connection that exists between these two 'concepts' in ancient texts (for
example, in Aristotle's writings). We are on much firmer ground when
we try to understand these terms as they feature in doctrines circulating in the post-Augustinian
period, and more particularly in the post-Cartesian era.
[On this, see for
instance, Matson (1966). Also see, Hacker (2012,
2013). I have quoted the opening two pages of the latter,
here.]
Representationalism was a
dominant ideology in the Middle Ages, too. On this, see Hallett (2008),
pp.25-32. The modern period will be covered in Essay Twelve Part Seven (not yet
published). However, an excellent survey of this theme, as it runs
like a thread through the history of 'western' thought, can be found in Sokolov (1975),
pp.11-33.
Even so, Scott Soames sums up
what might be called the currently dominant view of the nature of language, in
the following terms:
"[T]he central semantic fact
about language [is]...that it is used to represent the world." [Soames (1989),
quoted in Chomsky (2000a), p.132.]
Since the seventeenth
century, however, and under the influence of early bourgeois theorists (like
Hobbes,
Locke and
Descartes), the "Cartesian
Paradigm"
has come to dominate philosophical and scientific thought -- and, what is more,
its main features have remained largely unchallenged to this day.
On that, see Coulter
(1993, 1997), Coulter and Sharrock (2007), Hacker (2007a), pp.233-56, Hacker
(2012, 2013a),
Read (2008),
Robinson (2008) -- Robinson's book is in fact entirely devoted to the analysis
and criticism of this Paradigm, both in its older Platonic form (see
above) and in its more
modern Cognitive Science/Cartesian re-incarnation --,
Ryle (1949a
(this links to a PDF), Shanker (1987d,
1996b), Savage-Rumbaugh, Shanker and Taylor (1998),
pp.77-138 (although the latter draws conclusions about animal 'thought' that are
diametrically opposed to the approach adopted in this Essay). See also the detailed study in
Chomsky (2009a) -- however, on that, cf., Note
1d, below. The Platonic and Cartesian origin of Chomsky's theory of
language are explored in detail in Knight (2016a). On this, see, for example,
Chomsky (1988), pp.3-6, 24-27, 138-52. On Descartes himself, see Cottingham (1992)
and Williams (1990). I have to say that even though I profoundly disagree with
Williams's defence of Descartes's project, it is an admirably clear study from
which I have learnt much. For a
Wittgensteinian criticism of Descartes's approach to 'the mind' (and upon which I
have based many of my own ideas), see Kenny (1992, 1993), although readers
shouldn't assume I agree with everything Kenny has to say. There is also an extended
discussion of this Paradigm (and Wittgenstein's arguments against it) in Wilson
(1998) -- although Wilson concentrates largely on Locke's version of it.
[See also Note 2.]
Medieval views of 'the mind' were largely dominated by Aristotle's theory. A
paradigm shift occurred when Descartes reoriented philosophical
psychology back toward the even more archaic
Augustinian-Platonic tradition.
[On this, see
Hacker (2012) (this links to a PDF), and Hacker (2013a). Again, I have
reproduced the opening two pages of the latter,
here.]
Anthony Kenny neatly summarises this
Dominant Paradigm
and its widespread influence in the following terms:
"Descartes's view of the nature
of mind endured much longer than his view of matter. Indeed among educated
people in the West who were not professional philosophers it is still the most
widespread view of the mind. Most contemporary philosophers would disown
Cartesian dualism but even those who explicitly renounce it are often profoundly
influenced by it. Many people, for instance,
go along with Descartes in identifying the mental realm as the realm of
consciousness. They think of consciousness as an object of introspection; as
something we see when we look within ourselves. They think of it as an
inessential, contingent matter that consciousness has an expression in speech
and behaviour. Consciousness, as they conceive it, is something to which each of
us has direct access in our own case. Others, by contrast, can only infer to our
conscious states by accepting our testimony or making causal inferences from our
physical behaviour." [Kenny (1992), p.2. Bold emphases added; paragraphs
merged.]
"In philosophy, ever since
Plato, the mainstream opinion has been that the mind is the organ of thought;
thinking is what the mind is for, and we act as we do because we think what we
do." [Fodor
(2011), p.24. Bold emphasis added.]
As Bennett
and Hacker note:
"Philosophical
reflection on human nature, on the body and soul, goes back to the dawn of
philosophy. The polarities between which it fluctuates were set out by Plato and
Aristotle. According to Plato, and the Platonic-Christian tradition of
Augustine, the human
being is not a unified substance, but a combination of two distinct substances,
a mortal body and an immortal soul. According to Aristotle, a human being
is a unified substance, the soul (psuchē) being the form of the body.
To describe that form is to describe the characteristic powers of human beings:
in particular, the distinctive powers of intellect and will that characterize
the rational psuchē. Modern debate on this theme commences with the
heir to the Platonic-Augustinian tradition: namely, the Cartesian conception of
human beings as two one-sided things, a mind and a body. Their two-way causal
interaction was invoked to explain human experience and behaviour.
"The greatest
figures of the first two generations of twentieth-century neuroscientists, e.g.,
Sherrington,
Eccles and
Penfield, were avowed
Cartesian dualists. The third generation retained the basic Cartesian structure,
but transformed it into brain–body dualism: substance dualism was abandoned, but
structural dualism retained. For neuroscientists now ascribe much the same array
of mental predicates to the brain as Descartes ascribed to the mind, and
conceive of the relationship between thought and action, and experience and its
objects, in much the same way as Descartes -- essentially merely replacing the
mind by the brain. The central theme of our book was to demonstrate the
incoherence of brain–body dualism, and to disclose its misguided
crypto-Cartesian character. Our constructive aim was to show that an
Aristotelian account, with due emphasis on first- and second-order active and
passive abilities and their modes of behavioural manifestation, supplemented by
Wittgensteinian insights that complement Aristotle's, is necessary to do justice
to the structure of our conceptual scheme and to provide coherent descriptions
of the great discoveries of post-Sherringtonian cognitive neuroscience."
[Bennett and Hacker (2008), pp.240-41. Bold emphases and links added. Italics
in the original.]
So, the
Platonic-Christian-Cartesian Paradigm still dominates Cognitive Science, and as we will see throughout
this Essay (especially
here) -- despite their erstwhile rejection of dualism -- DM-theorists en
masse have bought into this paradigm in one form or
another (largely thanks to Hegel).
Representationalism is
part-and-parcel of this view of 'the mind' and remains another dominant
paradigm in Cognitive Science, as the following author points out:
"The
notions of computation
and representation are not just common currency in cognitive science modelling.
To put it mildly, they are the building blocks of the discipline.
Alternative voices from a number of subdisciplines that call into question these
notions have periodically been raised. Unfortunately, after an initial, and
usually short, excitement they remain quiet. Silence is due mainly to two
reasons. On the one hand, the dominant
paradigmoverwhelms competitors
(sometimes due to 'pragmatic' considerations) with data already accounted for
and results to be accounted for, and on the other hand, alternative framings are
repeatedly absorbed and made innocuous. Both reasons are interrelated.
Alternatives raised, by default, carry the burden of proof in such a way thatthe dominant paradigm is the one that chooses what phenomena are in need
of explanation.... Problems start when the what limits the range
of options available when it comes to answering the how." [Garzón (2008),
pp.259-60. Quotation marks altered to conform with the conventions adopted at this site. Bold emphases alone added.]
In relation to this,
Hilary
Putnam notes the following about contemporary Philosophy of Mind:
"...[V]irtually no
philosopher doubted, from the time of
Locke
until roughly 1914, that, whatever concepts and ideas were, they were clearly
mental objects of some kind. And no large-scale and comprehensive demolition job
was done against this particularly wide-spread and influential philosophical
misconception until
Wittgenstein produced his
Philosophical Investigations...." [Putnam (1975b), p.7.]
As Richard Rorty also points out in relation to
the views of those who agree with Wittgenstein (which group he calls "Pragmatic
Wittgensteinians"):
"Pragmatic Wittgensteinians
tend to be historicist in their metaphilosophical views. They think that the
problems of pre-Kantian metaphysics, the problems that the naturalists [this is
Rorty's term for those who reject Wittgenstein's approach to philosophy -- RL]
have revived, are hangovers from a particular moment in Western intellectual
history. These problems originate not in a clash between common sense and
science, but rather between the immaterialist notions that Christian theology
had inherited from Plato and Aristotle and the mechanistic and materialist
world-picture sketched by Galileo and Newton. That clash was between
metaphysical outlooks, not between metaphysics and a pre-metaphysical
understanding of things.
"This clash produced the Cartesian
notion of ideas as appearances on the stage of an inner theatre, as well as the
Lockean account of words as signs of such ideas. More generally, it produced a
picture of knowledge as the attempt to acquire accurate mental representations
of non-mental reality. Representationalist accounts of the relation between
language and non-language emerged from the attempt to divide language into
assertions that represent real things and those that do not. On this historicist
view, Wittgenstein's importance lies in his having helped wrench us out of our
Cartesian-Lockean mind-set." [Rorty (2010), p.132. Bold emphasis added.
Readers should note the caveats concerning Rorty's work posted
here.]
In that case, according to the traditional
Platonic-Christian-Cartesian approach, 'mental objects' and 'concepts' are antecedent to discourse, a doctrine
that is integral to
Representational Theories of Mind and Language. This paradigm sees language as
primarily an "instrument of thought", not of communication, as
Chomsky
himself acknowledged:
"If so, then it
appears that language evolved, and is designed, primarily as an instrument of
thought. Emergence of unbounded
Merge in human
evolutionary history provides what has been called a 'language of thought,' an
internal generative system that constructs thoughts of arbitrary richness and
complexity, exploiting conceptual resources that are already available or may
develop with the availability of structured expressions. If the relation to the
interfaces is asymmetric, as seems to be the case, then unbounded Merge provides
only a language of thought, and the basis for ancillary processes of
externalization." [Chomsky
(2007), p.22. This links to a PDF, access to which is open, but requires free
registration.]
"It seems that
we must either deprive the notion 'communication' of all significance, or else we must
reject the view that the purpose of language is communication.... It is
difficult to say what 'the purpose' of language is, except, perhaps, the
expression of thought, a rather empty formulation." [Chomsky (2005a), p.230. I owe this reference to
Millikan (2005b), p.24. However, Millikan must have been using a different edition,
since my copy reads as I have reproduced it above, but not as Millikan has
represented it. Quotation marks altered to conform with the conventions
adopted at this site. Bold emphases added.]
"Now let's take
language. What is its characteristic use? Well, probably 99.9 percent of its use
is internal to the mind. You can't go a minute without talking to yourself. It
takes an incredible act of will not to talk to yourself. We don't often talk to
ourselves in sentences. There's obviously language going on in our heads, but in
patches, in parallel, in fragmentary pieces, and so on. So if you look at
language in the way biologists look at other organs of the body and their
subsystems -- so you take into account all its functions in talking to yourself
-- what do you get? What are you doing when you talk to yourself? Most of the
time you're torturing yourself.... So you might think you're being conned, or
asking why does this person treat me that way? Or whatever. So you could say
that the function of language is to torture yourself. Now, obviously, that's not
serious....
"In fact, a
very tiny part of language is externalized -- what comes out of your mouth, or
from your hands if you're using sign. But even that part is often not used for
communication in any independently meaningful sense of the term
'communication'...the overwhelming mass of language is internal; what's external
is a tiny fraction of that [and what's used in communication in some serious
sense is a smaller fraction still]. As functions are usually informally defined,
then, it doesn't make much sense to say that the function of language is
communication." [Chomsky (2012a), pp.11-12, quoted in
Behme (2014a)
pp.6-7. I haven't yet been able to check this particular book of Chomsky's.
The Behme link no longer appears to be working!
However, a corrected version of this paper can be accessed here.
(This links to a PDF).]
These
comments place Chomsky firmly in the Idealist,
Platonic-Christian-Cartesian
camp. [On this, see Note 1d, below.]
Notice how Chomsky has to appeal to "inner speech", and what
looks suspiciously like an anecdotal form of it, too. However, Chomsky failed to
say where he obtained the figure of 99.9%; but, even supposing he were only half
right, that wouldn't show that the primary function of language doesn't lie in
communication, as Christina Behme points out:
"The blinkers of my car are in off
mode most of the time, and they can be used for several purposes (signalling
slow speed when driving on the highway, signalling change in direction, greeting
a friend, entertaining my kids, etc.). At times they are misleading (when I
forget to cancel them after a turn or change my mind about turning), and there
are other means to signal a change in direction.... Yet, uncontroversially, the
function of the blinkers is to signal my intention to change direction." [Behme
(2014b), p.682.]
Indeed, my car is parked outside my flat for
long periods (I estimate that it is there well over 90% of the time), but that
doesn't imply I was wrong to enter its main use as "social and leisure purposes"
on my insurance application form. I certainly didn't even think to describe its
use as "the occupation of a parking space", even though that is what it does most of the
time.
Pieter Seuren also makes the following
not insignificant points:
"Sober reflection on the
well-known fact that sentences, and therefore also their token realizations as
utterances, never come without a speech act component immediately makes the
communicative function of the language faculty stand out among any other
functions it may have. Likewise for the obvious fact that the external
motor-perceptual component has been considerably adapted in the course of recent
evolution to the production and perception of speech sounds and hence public
use...." [Seuren (2004), p.135.]
This helps account for the emphasis placed on
Wittgenstein's work at this site, especially in relation to the
topics discussed in this Essay; for, as Putnam noted, Wittgenstein was the first
leading philosopher to challenge these "ruling ideas" (even though
he, too, wouldn't have put things this way!). Of course, this also helps explain
Chomsky's (and
Fodor's) unwavering hostility toward Wittgenstein, as the following commentator noted:
"The
basis for this
dislike by Chomsky
is surely to be found in a difference between a Cartesian and a social view of
language. Fortunately one does not have to speculate here. As it has been most
clearly accepted and claimed by Fodor (1975; 1980) among the Chomskyans,
according to them the whole mind has to be understood in a
solipsistic way.
Mental states have to be interpreted in themselves, and in the same way, rules
have to be interpreted just as properties of the isolated, self contained system
of the individual mind." [Pleh
(no date). This link connects to a downloadable .DOC file. Bold emphases
added.]
However, as noted in Essay Three
Part Two,
just like the proverbial
Humpty
Dumpty, fragmented discourse -- predicated on an atomised linguistic base
by theorists like Descartes, Locke, and now Chomsky and Fodor -- can't easily be put back together again, no
matter how many incantations theorists like Hegel (and Voloshinov) utter over its dismembered
corpse.
[The serious implications of this ideologically-motivated wrong turn will emerge as this Essay unfolds.]
In
that case, this tradition isn't a natural hunting ground for Marxists to go in
search of kindred spirits, even though
far
too many have been seduced by these Platonic-Cartesian-Chomskyan "ruling ideas".
[Some might wonder how "ruling ideas" can
rightly be attributed to a noted left-wing theorist like Chomsky. However, as we
have seen, even card-carrying Marxist classicists like Engels, Plekhanov, Lenin
and Trotsky found themselves importing "ruling ideas" into the workers'
movement. (Why they all do this is explained in Essay Nine
Part Two.) On Chomsky, see Knight
(2003,
2005,
2006,
2007a,
2007b,
2007c,
2010a,
2010b,
2011, 2016a,
2016b). On representational theories of perception,
see Travis (2004).]
1a.
This finds echo in Engels's thoughts on the matter:
"The great basic question of
all philosophy, especially of more recent philosophy, is that concerning the
relation of thinking and being." [Engels
(1888), p.593.]
To be sure, Engels might genuinely have
believed he was discussing the alleged relation between 'thought' and 'reality',
but the philosophical dice had already been heavily loaded against him.
As a result his ideas are also firmly located within the Platonic-Christian-Cartesian Paradigm,
noted above. [More on that in
Essay Twelve (summary,
here).]
1b.
Augustine
asserted the following: "Return to yourself. Truth dwells within" (quoted in Pasnau (2002),
p.208).
Thomas
Aquinas concurred: "Truth is only in the mind" (quoted in Hallett (2008),
p.26). Compare that with what Aquinas's teacher,
Albertus
Magnus, had to say: "Truth and falsehood in speech are
signs of truth and falsehood in the soul" (quoted in Hallett (2008), p.190, note
8). Hence, words were merely the means by which we communicate these
already-formed, private
thoughts to others.
1c.
In this Essay, I will
be concentrating solely on these two competing views of the role of discourse:
representation versus
communication. This isn't to suggest
that there aren't other important uses of language (e.g.,
interrogative,
imperative,
optative,
expressive, and creative -- or, indeed, its socially cohesive role, etc.,
etc.), or that there aren't political dimensions to language (explored,
for example, in Holborow (1999). Far from it. However, the above two historically dominant approaches to language seem
to me to be the
most pertinent to the aims of this Essay, and this site.
1d.
It is no surprise, therefore, to learn that perhaps the most influential
theorist in this area, Chomsky, chose to call one of his major works "Cartesian
Linguistics", subtitled "A Chapter in the history of rationalist
thought"; i.e., Chomsky (2009a).
Even so, Chomsky's version of the history of
linguistics hasn't met with universal acceptance; quite the reverse in fact. On
this, see Aarsleff (1970). [Chomsky's rather irascible reaction to such
criticism is recorded in
Behme (2014b), pp.686-92,
and Knight (2018), i.e., the
Preface
to the second edition of his 2016).]
2.
On modules, see Note 52a. Of
course, this ignores 'neural
nets'; I will say more about this side-issue in a future re-write of
this Essay.
Plainly, this amounts to what is in effect a bourgeois division of 'mental labour'
-- perhaps best represented in the work of
Daniel
Dennett (with his emphasis on
dim algorithms) -- illustrating, maybe, the free market degradation of the human psyche
in its most acute form. So, according to Dennett, de-skilling has hit a brain near you!
[On that, see for example, Dennett (1995), pp.53-60.] When
Harry Braverman wrote his classic study of the deskilling of work in late
capitalism, I'm not sure he envisioned a process of 'mental de-skilling'
underway at the same time! [Braverman (1999).]
This way of looking at the 'mind' and the
algorithms
it is supposed to run wasn't, of course, invented by Dennett, but goes back at least as
far as Turing
(however, see below); it now occupies an important role in modern versions of
the dominant
Cartesian Cognitive Paradigm; in place of
Descartes's "thought" and "mind", read "algorithm" and "module". [What was that
about a rose by any other name?]
[This Paradigm is criticised in Shanker (1998), pp.1-62. See also Shanker (1986c,
1987a, 1987b, 1987c, 1988, 1995, 1996b). Cf., also Bennett and Hacker (2021),
pp.34-38, 470-91, and Hutto (1997). On Dennett, see
Note 54, Note 97, as well
as
here. However, the best Marxist
discussion of this is still Robinson (2003b).]
In
fact, this notion
is undoubtedly an echo of Ancient Greek
ideas about the rational 'soul, the cosmic order, psychical "balance", and the
connection these had with ratio -- hence the word "rational" --, numbers and geometry
-- especially
in theories concocted by Plato, the
Pythagoreans
and the
Stoics. On that, see for example,
Furley (1973), Kenny (1973b), and Holland (1980b). Later, we will see this
ancient doctrine re-surface in theories of 'mind' concocted by the so-called "Critical
Realists".
Nevertheless, the aim of this computational approach to the 'mind' appears to be to
reduce 'mental processes' to
mechanically executed, 'simple algorithms' that require 'no intelligence' to run,
and which don't therefore involve anyone, or anything, attending to meanings (hence, once more, they operate independently of, and prior
to, language,
even in an evolutionary sense), nor
do they require the application of any kind of reasoning. Of course, this change of emphasis was an essential move in
the development of computers, but it has nothing to do with the way human beings
think, as John Haugeland points out:
"Reasoning (on the
computational model) is the manipulation of meaningful symbols according to
rational rules (in an integrated system). Hence, there must be some sort of
manipulator to carry out those manipulations. There seem to be two basic
possibilities: either the manipulator pays attention to what the symbols and
rules mean or it doesn't. If it does pay attention to the meanings, then
it can't be entirely mechanical -- because meanings (whatever exactly they are)
don't exert physical forces. On the other hand, if the manipulator does not pay
attention to the meanings, then the manipulations can't be instances of
reasoning -- because what's reasonable or not depends crucially on what the
symbols mean. In a word, if a process or
system is mechanical, it can't reason; if it reasons, it can't be mechanical." [Haugeland (1985), p.39. Italic emphasis in the original.
I owe this reference to Shanker (1988), p.252. Paragraphs merged.]
This is somewhat reminiscent of Gilbert
Harman's argument against representationalism, summarised
below.
2a0.
Or, if we are enjoined by
EP-ers like Richard Dawkins to resist the
machinations of our
genome, we must be more than the product of our
genes. But, if
that is so, what could
this 'something more' possibly be, given this deterministic view human
nature?
[On this, see Stove (2005), pp.172-97.
Readers should, however, read the warning I have posted
about Stove, here. I will say more about Dawkins and his 'selfish genes' in Essay
Three Part Five -- in the meantime, see
here.]
"By the end of World
War II, a new paradigm for the scientific understanding of human development had
emerged. For a variety of reasons...the idea that human nature is the preeminent
force at work in shaping the character of a person and the contours of her life
had given way to a resolute
empiricism. For the first time since the
Enlightenment..., experience, and not our innate endowment, was universally
accorded the primary role in the making of ourselves and our society. By the
time of the
war in Vietnam, however, the newly resurrected empiricist orthodoxy
was being challenged by a vigorous resurgence of
nativism
[a set of
doctrines involving
Innatism and a belief in
Universal Grammar, etc.
-- RL]. Far from playing a minor role in the making of a person, human nature
was again taken to be the chief factor at work in determining the nature of
people and the milieux in which they live.
"Today, 'Nativism Rules, OK'.
In both the popular and academic presses. books and articles claiming to have
found a 'substantial genetic basis for,' or 'instinct for,' or even 'the gene
for,' a variety of traits and behaviours have proliferated. Features as diverse
as scholastic performance, sexual orientation, violence, 'altruism, compassion,
empathy, love, conscience, the sense of justice,' poverty, alcoholism and other
substance abuse, susceptibility to diseases, sexual mores, the desire to rape
women, the attainment of concepts, language use, even attitudes toward divorce
and religion -- features that were formerly held to be substantially under
environmental control -- are now routinely claimed to be largely, if not wholly,
innate....
"Nativism's potential
for affecting profound changes in our views about ourselves and our society is
already becoming apparent. The nativist's shift of explanatory emphasis from the
environment to the genes -- or from the knowable and the manipulable to the
imponderable and immutable -- seemingly supports the more general shift in
sociopolitical attitudes that is occurring as the twentieth-century lurches to a
close. The growing obsession here [i.e., in the USA -- RL] and abroad with
racial, national, and sexual differences; the widely perceived failure of social
welfare programs to cure society's ills; the newly fashionable backlash against
attempts to enforce ideals of equality and civil rights; the calls for a
reinstatement of 'traditional morals' and 'family values'; the nostalgia for a
time when everyone had a place and knew that place and stayed in it -- these
kinds of views find a fertile seeding ground in the New Nativism. Conservative
politicians, moralists, and jurists apparently find overwhelming the inference
from 'innate' to 'right' and 'inevitable.' If the poverty and violence of our
inner cities are coded in the
genotypes of their inhabitants, then the
government programs aimed at ameliorating these conditions are pointless. If
poor scholastic performance among minority children is a consequence of their
substandard genes, then forget about
Head
Start
and other educational reforms. If the kind of serial monogamy
practices in our society destroys some men's biological 'right' to reproduce,
then we should tighten the divorce laws. If women are by nature less aggressive
than men, and if aggression is a factor in achieving social status and economic
success, then sexual inequality and the 'glass
ceiling' are here to stay. And so on." [Cowie (2002), pp.vii-xi.
Quotation marks altered to conform with the conventions adopted at this site; spelling
changed to UK English; minor typo corrected. Links added. (By the way, Cowie
added references to nativist literature in support of the claims she made about
Nativism.)]
"Male supremacy misrepresents all women as genetically inferior, manipulative
and stupid and reduces them to their reproductive or sexual function -- with sex
being something that they owe men and that can or even should be coerced out of
them.
Driven by a biological analysis of women as fundamentally inferior to men,
male supremacists malign women specifically for their gender.
Their thinly veiled desire for the domination of women and their conviction that
the current system oppresses men in favour of women are the unifying tenets of
the male supremacist worldview." [Quoted from
here; accessed 08/03/2018. Bold emphasis added; spelling modified to agree
with UK English.]
Update October 2024: Readers might like to check out this entertaining
2024 video that shows how EP has taken over the political (anti-'woke') right,
and the misogynistic 'Manosphere'
Video Four: EP Debunked
In
connection with this, it isn't surprising to see
Dawkins now promote and
praise as "utterly superb" a book by far right Tory ideologue,
Douglas Murray, which
argues in favour of the Nazi-inspired,
Great Replacement Theory.
But, Dawkins's future political trajectory was abundantly clear to anyone who
read his first book, The Selfish Gene, with open eyes [i.e., Dawkins
(2006)].
2b.
Some might consider this claim somewhat dubious, but they would be wise to shelve those doubts until Essay Twelve Part
Two has been published. Until then they should consult the background
details in Essay Twelve Part One,
and here.
3.
This will form the main topic of Essay Twelve Parts Two, Three and Seven (summary
here).
Of course, that is to express this topic
in a contemporary idiom! Nevertheless, as the above Essays will show, philosophers,
priests and assorted ruling-class ideologues have concocted a range of
esoteric
vocabularies
and specialised
lexicons, invented in order to give voice to their dogmatic theories -- but, and more
importantly, in order to rationalise the interests of their patrons -- the elite, the rich and
the powerful.
To be
sure, these theories have changed their content many times -- sometimes
dramatically -- in accord with the rise and fall of each successive Mode of
Production, and as the exigencies of the class struggle required. But, the form
these theories have taken has remained largely
the same for over two millennia: that behind 'appearances' there exists an invisible world,
which is accessible to thought alone
and which impervious to empirical investigation, a hidden world that is more real than the physical
universe we see around us.
Some might object that philosophical ideas can't
have remained the
same for thousands of years, across different Modes of Production, since that
would run
counter to core ideas in HM. But, we don't argue the same
for religious belief. Marx put no time stamp on the following, for
example:
"The foundation of irreligious criticism is:
Man makes religion, religion does not make man. Religion is, indeed, the
self-consciousness and self-esteem of man who has either not yet won through to
himself, or has already lost himself again. But man is no abstract
being squatting outside the world. Man is the world of man -- state,
society. This state and this society produce religion, which is an inverted
consciousness of the world, because they are an inverted world.
Religion is the general theory of this world, its encyclopaedic compendium, its
logic in popular form, its spiritual point d'honneur, its enthusiasm, its
moral sanction, its solemn complement, and its universal basis of consolation
and justification. It is the fantastic realization of the human essence
since the human essence has not acquired any true reality. The struggle
against religion is, therefore, indirectly the struggle against that world
whose spiritual aroma is religion.
"Religious
suffering is, at one and the same time, the expression of real suffering
and a protest against real suffering. Religion is the sigh of the
oppressed creature, the heart of a heartless world, and the soul of soulless
conditions. It is the opium of the people."
[Marx
(1975c), p.244.
Italic emphases in the original.]
The above remarks applied back in Babylon and
the Egypt of the Pharaohs,
just as they did in Ancient China and the rest of Asia, The Americas, Greece,
Rome and throughout Europe, Africa, Australasia, and as they have done right across the planet ever since.
The same is true of the core thought-forms that run through Traditional
Philosophy: that there is an invisible world, accessible to thought
alone --, especially since Marx also argued that:
"[P]hilosophy is nothing else but
religion rendered into thought and expounded by thought, i.e., another form
and manner of existence of the estrangement of the essence of man; hence equally
to be condemned...." [Marx
(1975b), p.381. Bold emphasis added.]
And:
"[O]ne fact is
common to all past ages, viz., the exploitation of one part of society by the
other. No wonder, then, that the social consciousness of past ages, despite
all the multiplicity and variety it displays, moves within certain common forms,
or general ideas, which cannot completely vanish except with the total
disappearance of class antagonisms. The Communist revolution is the most
radical rupture with traditional property relations; no wonder that its
development involved the most radical rupture with traditional ideas." [Marx
and Engels (1968b), p.52. Bold emphases added.]
This, of course, helps explain why Marx thought this
entire discipline (Philosophy) was based on
distorted language, and contained little other
than empty abstractions and alienated thought-forms -- and, indeed, why he turned his back on it from the
late 1840s onward. [On that, see
here.]
For such theorists,
philosophical
knowledge of this hidden world was the onlygenuine knowledge. Empirical knowledge of this (material) world
was limited, inadequate, contingent, and thus couldn't count as genuine, on a
par with the 'superior knowledge' available to philosophers. This
invisible world soon found itself populated with "Universals", "Forms", "Ideas", "Essences",
and other assorted 'abstract'/'Ideal' objects, but through the centuries it has
always remained an
occult, a priori,
'Ur-universe',
which constitutes the 'Being' that Absolute Idealists and assorted mystics declare is the real aim of
the 'soul's' desire for mystical unity (mentioned in Note 1, above).
"Empirical,
contingent
truths have always struck
philosophers as being, in some sense, ultimately unintelligible. It is not that
none can be known with certainty…; nor is it that some cannot be explained….
Rather is it that all explanation of empirical truths rests ultimately on brute
contingency -- that is how the world is! Where science comes to rest in
explaining empirical facts varies from epoch to epoch, but it is in the nature
of empirical explanation that it will hit the bedrock of contingency somewhere,
e.g., in atomic theory in the nineteenth century or in
quantum mechanics
today. One feature that
explains philosophers' fascination with
truths of Reason
is that they seem, in a
deep sense, to be fully intelligible. To understand a necessary proposition is
to see why things must be so, it is to gain an insight into the nature of
things and to apprehend not only how things are, but also why they cannot be
otherwise. It is striking how pervasive visual metaphors are in philosophical
discussions of these issues. We see the universal in the particular (by
Aristotelian intuitive induction); by the Light of Reason we see the essential
relations of
Simple Natures; mathematical truths are
apprehended by Intellectual Intuition, or by
a priori insight. Yet instead of examining the use of these arresting
pictures or metaphors to determine their aptness as pictures, we build
upon them mythological structures.
"We think of necessary
propositions as being
true or false, as objective and independent of our minds or will. We
conceive of them as being about various entities, about numbers even
about extraordinary numbers that the mind seems barely able to grasp…, or about
universals, such as colours, shapes, tones; or about logical entities, such as
the
truth-functions or (in Frege's
case) the
truth-values. We naturally think of necessary propositions as
describing the features of these entities, their essential characteristics.
So we take mathematical propositions to describe mathematical objects…. Hence
investigation into the domain of necessary propositions is conceived as a
process of discovery. Empirical scientists make discoveries about the
empirical domain, uncovering contingent truths; metaphysicians, logicians and
mathematicians appear to make discoveries of necessary truths about a
supra-empirical domain (a 'third
realm'). Mathematics seems to be the 'natural history of
mathematical objects' [Wittgenstein
(1978), p.137], 'the physics of numbers' [Wittgenstein (1976), p.138; however
these authors record this erroneously as p.139 -- RL] or the 'mineralogy of
numbers' [Wittgenstein (1978), p.229]. The mathematician, e.g.,
Pascal,
admires the beauty of a theorem as though it were a kind of crystal.
Numbers seem to him to have wonderful properties; it is as if he were
confronting a beautiful natural phenomenon [Wittgenstein (1998), p.47; again,
these authors have recorded this erroneously as p.41 -- RL]. Logic seems to
investigate the laws governing logical objects…. Metaphysics looks as if it is a
description of the essential structure of the world. Hence we think that a
reality corresponds to our (true) necessary propositions. Our logic is
correct because it corresponds to the laws of logic….
"In our eagerness to ensure
the objectivity of truths of reason, their sempiternality
and mind-independence, we slowly but surely transform them into truths that are
no less 'brutish' than empirical, contingent truths. Why must red exclude
being green? To be told that this is the essential nature of red and green
merely reiterates the brutish necessity. A proof in arithmetic or geometry seems
to provide an explanation, but ultimately the structure of proofs rests on
axioms. Their truth is held to be self-evident, something we apprehend by
means of our faculty of intuition; we must simply see that they are
necessarily true…. We may analyse such ultimate truths into their constituent
'indefinables'. Yet if 'the discussion of indefinables…is the endeavour to see
clearly, and to make others see clearly, the entities concerned, in order that
the mind may have that kind of acquaintance with them which it has with redness
or the taste of a pineapple' [Russell
(1937), p.xv (this links to a PDF); again these authors record this erroneously as p.v;
although in the edition to which I have linked, it is p.xliii -- RL], then the
mere intellectual vision does not penetrate the logical or metaphysical
that to the why or wherefore…. For if we construe necessary
propositions as truths about logical, mathematical or metaphysical entities
which describe their essential properties, then, of course, the final products
of our analyses will be as impenetrable to reason as the final products of
physical theorising, such as
Planck's constant."
[Baker and Hacker (1988), pp.273-75. Referencing conventions in the original
have been altered to conform with those adopted at this site. Italic
emphases in the original; links added.]
In its modern disguise, this doctrine
has re-surfaced as part of the pseudo-scientific doctrine that held that,
au fond, the universe is 'rational', governed (if not constituted) by
Mathematics and/or 'logic'. These days, apparently, the brain, or the 'mind', is also controlled,
or run, by somewhat similar
'algorithms' (if Daniel Dennett is to be
believed) -- which theory might make sense
if reality were indeed 'Mind', otherwise not. [On this, see Essay Twelve
Part Four (to be published in 2018).]
4. However, several articles
published in a book that appeared after this section was first written [i.e., Kitching
and Pleasants (2002)] suggest that this observation is now slightly out-of-date.
5.
Incidentally, in the many references to Wittgenstein's work at this site, the
reader will be hard pressed to find a single allusion to "language games"
(saving the present one, of course -- or, where I quote Wittgenstein directly).
Wittgenstein introduced this metaphor to assist him compare and contrast the many uses
there are of language, as well as to help him draw an analogy between
language and rule-governed social behaviour. It wasn't meant to suggest that
the use of language is merely a game, or that it is simply there for amusement or
recreation, and is thus of little import. Nor yet that we play games when we use language, or
even that our 'view of reality' is 'relative' to a given game. [The last few
words have been put 'scare quotes' partly because Wittgenstein himself would
have questioned their employment in such contexts.] He would have been as
horrified as Christopher Norris clearly is (whom Rees quotes) at the blasé,
over-use of this metaphor, and even more appalled at the way it has been employed
to construct relativistic philosophical theses in the social sciences,
for example. [More on that in a later Essay.]
5a.
Indeed, Wittgenstein had this to say about the "objects" of
the Tractatus:
"What I once called 'objects',
'simples', were simply what I could refer to without running the risk of their
possible non-existence; i.e., that for which there is neither existence nor
non-existence, and that means: what we can speak about no matter what may be
the case." [Wittgenstein (1975), p.72; §36. Italics in the
original.]
Objects of sense experience, whose existence
can be verified (or otherwise), aren't, therefore, the sort of "objects" to which
Wittgenstein was referring. So, they can't be "sense data", nor are they
dependent upon 'experience' (otherwise it would be possible to suppose them not
to exist), and neither are they the 'logical atoms' of
Russell's theory, which do depend on sense experience.
There are no satisfactory on-line articles on the Tractatus, but
this source at least runs through some of the main interpretations of the
"objects" that Wittgenstein introduced -- although, it has to be
said, the author finally adopts
an erroneous view of them as point masses, similar to those found in
Hertz's work. As noted above, for Wittgenstein it was impossible to suppose
that the "objects"
of the Tractatus didn't exist, but it is easy to imagine this of Hertzian
point masses. In fact, if we were to imagine the non-existence of one of these
point masses, we would have to appeal to these Tractarian 'objects' in order to do this! That was
the point of him saying: "what we can speak about no matter what may be
the case".
This article also outlines the many different interpretations there are of
these "objects", except the author omits the one adopted here
-- that is, that they
are logical objects (not to be confused with "logical atoms"), as Wittgenstein
himself indicated. Easily the best account of the Tractatus
and its "objects" is to be found in White (2006).
6.
Cf., Cornforth (1965), pp.111-30. Verification as a criterion of sense was
something Wittgenstein toyed with briefly in the late 1920s and early 1930s,
nearly ten years after
the Tractatus was published. Even then, the slant he put on
verification was entirely different from that touted by members of
the "Vienna
Circle" (with whose work Cornforth largely
confuses Wittgenstein's) -- who themselves also tended to conflate
Wittgenstein's aims with those of Russell. [On the difference between sense
and meaning, see here.]
On this, cf., Baker (1988),
Hacker (1985, 1996, 2000a), Hanfling (1981), Medina (2001), and Misak (1995). See
also the important qualifications in Diamond (1999).
Even so, the comments of
Communist Party member Cornforth weren't always consistent with the views
expressed by
the
CPSU (although that should surprise few of
those cognisant of their constant theoretical volt-faces). For example,
the Introduction to A
Textbook of Marxist Philosophy had this to say about Wittgenstein's
anti-metaphysical stance:
"The 'logical-analytic'
method of Wittgenstein and his followers is by no means the only modern
philosophy that approximates in certain points to the new dialectic....
It would appear, in fact,
that not only are scientific discoveries confirming the standpoint of
dialectical materialism but that Western philosophers are increasingly
discarding metaphysical concepts...." [Shirokov (1937),
pp.18-19. Paragraphs merged.]
Now, it may be because this
was written around about the time of the
Popular Front that the authors took a more conciliatory tone toward
Wittgenstein (whose work they, too, clearly confused with that of the Vienna Circle), but that just underlines how vacillating Stalinist authors were
between, say, 1930 and 1956, Cornforth included. So, compare the above
comments with those that Kuusinen published some twenty-five years later:
"The basic tenets of
neo-positivism were formulated by Bertrand Russell and the Austrian philosophers
Wittgenstein and
Schlick.
Its most prominent exponents are
Carnap
in the United States and
Ayer
in
Britain. It owes its origin to a desire to refurbish the subjective-idealist
philosophy of
Machism and adapt it to the present state of physics, mathematics and
logic." [Kuusinen (1961), p.57.]
It is worth noting, once again, that
Kuusinen offers no evidence at all to substantiate the allegation that
Wittgenstein was a Positivist, or even that he was keen to promote "Machism".
But, that is
par for the course; DM-theorists are notoriously cavalier with such details.
We may speculate that this
cavalier attitude might be because a "one-sided" emphasis on evidence might result in accusations of "Empiricism!" being thrown at
those foolish enough to request some. I know, because I have been accused of
being one such simply because I keep asking
for the evidence that supports or substantiates the hyper-bold theses concocted by
DM-theorists -- or, to be more accurate, the evidence that supports the
theses they have lifted from Hegel and other assorted mystics. The easiest way
to stifle counter-accusations like this is to adopt just such a
cavalier and slipshod attitude toward evidence, or and the need to produce any. There seems
to be no other way to account for this attitude, which has been echoed right across
all areas of Dialectical Marxism. [Substantiation for that particular allegation
can be found in most of the Essays published at this site, but more specifically
in Essay Two (link above) -- and, in Lenin's case,
here.]
6a0. Indeed,
one rather
incautious 'comrade' even accused me of "superficiality" while having the
cheek to refer his
readers to Gellner's hyper-superficial analysis of OLP!
In a review of Kitching and
Pleasants (2002), David Stern has this to say about Gellner:
"While Gellner's critique,
largely composed of shoddy rhetoric, insinuation and personal abuse, created
considerable controversy, it was dismissed by the philosophical establishment at
the time. However, his caricature of Wittgenstein was enormously attractive to
those who needed a convenient rationale for dismissing him. It has since become
conventional wisdom in many quarters, and especially among social scientists,
and is certainly part of the reason why a relatively small number of social
scientists on the left have taken a serious interest in Wittgenstein." [Quoted
from
here.]
More-or-less the same can be said about the
influence of Marcuse's book, which has almost single-handedly been responsible
for prejudicing the minds of countless Marxists against Wittgenstein and OLP.
6a01. Readers
might be tempted to ask: Exactly what is it that allows Ms Lichtenstein to think
she can rise above such social forces?
The answer is simple: (i) An implacable determination not
to propound or promote a single philosophical theory -- this being an important first step in the right
direction -- coupled with an equally strong determination (ii) To cleave to the deliverances of common understanding and (iii)
To take seriously Marx's advice (see below) and heed the protocols expressed in, and
that are constitutive of, the vernacular. [Precisely how these steps manage to achieve that end will be
explained in detail in Essay Twelve Part Seven.]
What is more,
Marx plainly
held similar views:
"The philosophers have only
to dissolve their language into the ordinary language, from which it is
abstracted, in order to recognise it, as the distorted language of the actual
world, and to realise that neither thoughts nor language in themselves form a
realm of their own, that they are only manifestations of actual life." [Marx
and Engels (1970), p.118. Bold emphases added.]
Ordinary language is
therefore by implication a manifestation of "actual life"; that is, it is a
manifestation of life that hasn't been distorted by
class-motivated ideology. This also follows on from another piece of advice from
Mar-- advice that the vast majority of DM-theorists have yet to heed:
"One has to 'leave philosophy aside'...,
one has to leap out of it and devote oneself like an ordinary man to the study
of actuality...." [Marx
and Engels (1976), p.236.
Quotation marks altered to conform with the conventions
adopted at this site.]
Now, this might or might not help guarantee that the
thoughts expressed in these Essays rise above "such social forces", but one
thing that will most certainly guarantee that they don't is the adoption
of
boss-class forms-of-thought, beloved
of DM-fans.
These and other related topics (i.e.,
those concerning colour) are
discussed at length in Wittgenstein (1958, 1980c). Cf., Glock (1996), pp.81-84,
Hacker (1987), Hanfling (2000), Harrison (1972, 1973), and McGinn (1991).
Cf., also Hardin (1993), and Westphal
(1991). However, the approach adopted by the latter two books is incompatible with the
method employed at this site. They are offering a theory of colour, I am not --
nor will I, and for reasons outlined in Essays
One and
Twelve Part One.
However,
on this topic, the reader should consult a minor modern classic in this area:
Stroud (2000b). I can't recommend this book too highly.
Incidentally, it is a bad idea to confuse concepts with objects, words, mental
processes or structures (which view of concepts seems to have infected much of
post-Kantian 'Continental Philosophy'). That is because it reduces
sentences and propositions to lists, and lists say nothing. The approach adopted at
this site treats a concept as the expression of a linguistic rule. There are
good reasons to prefer that approach over the traditional analysis of
concepts and predicates, among which are the following -- as I noted in Essay Three
Part One (slightly
edited):
There are many advantages to this
neo-Fregean way of analysing language; several will be outlined later in this Essay. However, for
present purposes, the main advantage is that it isn't possible to interpret schemas like "ξ
is a warmonger" as a name of anything, least of all the name of an
Abstract Particular -- nor,
indeed, the name of
some sort of 'mental entity'. [The use of Greek letters like this is explained
here and here.]
This
modern approach also incorporates the word "is" into the
predicate expression (or, rather, into the linguistic functional expression),
short-circuiting questions about
whether
or not it is an "is" of identity or an "is" of predication.
No less important is the fact that it allows us to drop entirely from logic the words "predicate"
and "predication",
thereby casting into
oblivion two-and-a-half millennia of wasted effort, wrong turns, and aimless
metaphysics -- at the same time as
completely
undermining a
key argument in Hegel's 'Logic'.
The same can't be said about the results
of subject/predicate analysis
found
in Traditional Logic. For example, the alleged predicate -- "a warmonger"
-- looks like it designates or
names a class, concept, category, group, Idea, 'mental construct', or Abstract
Particular, all of which readings of the phrase would destroy the generality
expressed in the original propositions (for reasons explored in the main body of
this Essay). The post-Fregean approach using "ξ
is a warmonger" entirely side-steps this pitfall.
Subsequent Logicians and Philosophers were only too eager
to take the aforementioned 'wrong turn', adopting the 'Term
Logic' they inherited from Aristotle, which was largely based on the
traditional analysis of predicate expressions. The quotations taken from
DM-sources (many of which have been posted in Note 1e and
Appendix B),
expounding the 'process of abstraction', clearly illustrate how dialecticians
have wandered blindly into this bear trap. Novack provides us with a
particularly good example (but
there are many others):
"This law of identity of opposites, which so perplexes and
horrifies addicts of formal logic, can be easily understood, not only when it is
applied to actual processes of development and interrelations of events, but
also when it is contrasted with the formal law of identity. It is logically true
that A equals A, that John is John…. But it is far more profoundly true that A
is also non-A. John is not simply John: John is a man. This correct proposition
is not an affirmation of abstract identity, but an identification of opposites.
The logical category or material class, mankind, with which John is one and the
same is far more and other than John, the individual. Mankind is at the same
time identical with, yet different from John." [Novack (1971), p.92.
Bold emphasis added.]
By concentrating on "man" as the assumed predicate
expression (instead of "ξ
is a man"), Novack reduces this word to the name of an idea, category, set,
concept, or
class, thus destroying
the unity of the proposition.
[While
Novack doesn't explicitly say "man" is the name of an idea
(even though it is plain from what he does say that "man" is indeed such a name), other DM-theorists
do. [Again,
see Note 1e and
Appendix B
for examples.]
The
above, traditional analysis of predicates
not only conflates what is said about certain "subjects" with the means
by which this is done, it also predisposes theorists
to think of "man" as the name of an 'object of thought', since the word plainly
doesn't name anything in the 'outside world'. As soon as we look at this
'problem' through 'traditional spectacles', we are forced to search for
something for this word to name. The temptation then becomes irresistible to
look insideourselves to find its supposed referent. Hence, because of this
simple error of logic, a whole body of Traditional Metaphysics and Philosophy of
Mind has been conjured into existence. This simple error of logical syntax forced DM-theorists to
drop their theory into the solipsistic quagmire outlined in Note 4, above. Of course, the other
route taken by Traditional Theorists was to postulate a hidden world underlying
(or lying behind) appearances, to which such predicates expressions referred (or
the objects and processes in which predicates supposedly 'inhered'), an approach
to predicate expressions that underpins Platonism and
other forms of Classical Realism. [These topics will be explored in more detail in
Part Two.]
However, since "ξ
is a man" is an expression for a linguistic rule, the temptation to
confuse it with anything 'mental' -- or with the name of anything 'internal',
or even anything which resides in 'heaven', perhaps with 'God' (etc.) -- totally
vanishes. At a stoke, that removes this entire topic from the hidden, internal
world of mythical and uncheckable 'mental' processes -- or, indeed, from the ghostly world of
Forms, Universals, Categories and Concepts --, and places it squarely in
the public domain. On this account, the mastery of a concept is no more
nor no less mysterious than the mastery of a linguistic skill, publicly exercised and
capable of being inter-subjectively taught and hence checked. Paraphrasing Wittgenstein: a whole cloud of
metaphysics condensed in -- or, rather, distilled out of -- a drop of grammar. [Wittgenstein (2009), p.233,
§315;
(1958),
p.222. (This links to a PDF.) The online edition
of the latter has the page numbering all
wrong; it puts this sentence on p.22*6 (sic!).]
This is just one more advantage
this way of analysing language
possesses (further outlinedbelow). The use of Greek
letters -- as in "ξ
is a man" -- reminds us that this inscription is properly to be viewed as
the linguistic expression of a rule, which places this 'concept' in the
public domain. So, given this interpretation, concepts and predicates aren't 'mental
constructs' (as tradition would have us believe), they are expressions of linguistic rules we use
to communicate with one another, among other things.
7.Wittgenstein's deep respect for Frege can be gauged from the many
references he made to him in his published and un–published work -- they far
out-number those of any other philosopher --, and in the remarks
recorded by many of his pupils. On this, see Hallett (1977), pp.764-65, 791. On
Wittgenstein's overall debt to Frege, see Diamond (2010), Dummett (1991b), Reck (1997,
2002b), and most of the other essays in Reck (2002a). Indeed, and on a personal
note, I can recall
Peter Geach (one of Wittgenstein's closest
post-war friends) commenting -- when asked by me what attitude Wittgenstein had adopted
toward Frege when he (Geach) knew him -- that to his mind Wittgenstein always spoke of Frege as he
imagined Aristotle would have spoken of Plato.
8.Indeed, if Watson rises to the challenge he will find the
exact opposite of what he unwisely asserted about Wittgenstein and AFL. In
fact, those who bother to
check will discover several of Wittgenstein's comments have been used by
dialetheic
logicians in
support of their view that contradictions can be 'true'! [Cf., Priest (2002,
2004, 2006, 2007); and see below.]
A more balanced view of
this matter (and one that displays a rather more secure grasp of Wittgenstein's
overall method) can be found in Goldstein (1986, 1988, 1989, 1992, 1999). Graham
Priest's work will be
reviewed in a later Essay; in the meantime the reader is encouraged to check
this out, which is a
review of one of Priest's recent books, written by a leading logician,
Hartry Field. See also
Slater (2004) -- now Slater (2007b) -- , Slater (2007c), and Field (2008),
pp.361-92.
Anyway, it is clear that
Wittgenstein was happy to question
AFL in ways that alarmed less progressive contemporaneous classical logicians. [Confusingly,
the phrase "classical
Logic" now refers to Fregean and Russellian logic, etc., not AFL.]
In fact, one only has to look at the reception given to Wittgenstein's work on the
Philosophy of Mathematics, for instance, to see the extent to which his ideas horrified
classicists. On this see, for example, the first edition of Benacerraf and Putnam (unfortunately, the
second edition (i.e., Benacerraf and Putnam (1983)) has had all of this material removed)
-- Benacerraf and Putnam
(1964), i.e., Anderson (1958), Dummett (1959), and Bernays (1959).
On the controversy provoked
by this aspect of Wittgenstein's work, see Monk (2007), and Shanker (1987d).
The best on-line article on Wittgenstein's Philosophy of Mathematics is
Rodych (2018).
Here is what I have
said about this topic in another
Essay:
Here are just a few of the
many things [Wittgenstein] had to say
about contradictions in notebooks and lectures
during his 'middle period':
"I want to talk about the sense in which we
should say that the law of contradiction -- ~(p.~p)
[the 'dot' stands for 'and', and the tilde (i.e, '~') for 'not';
'p' is a propositional letter; '~(p.~p)' is counted as a tautology in the
Propositional Calculus(i.e., true under every
interpretation) -- RL] -- is a true proposition. Should we
say that if '~(p.~p)' is a true proposition, it is true in a
different sense of the word from the sense in which it is a true proposition
that the earth goes around the sun?
"In logic one deals with tautologies --
propositions like '~(p.~p)'. But one might just as well deal
with contradictions instead. So that the
Principia Mathematica
would not be a collection of tautologies but a collection of contradictions.
Should one say that the contradictions were true? Or would one then say that 'true' is being used in a
different sense?" [Wittgenstein (1976), p.187. Links added.]
"The laws of logic, e.g., excluded middle and
contradiction, are arbitrary. This statement is a bit repulsive, but
nevertheless true. In discussing the foundations of mathematics the fact that
these laws are arbitrary is important, for in mathematics contradiction is a
bugbear. A contradiction is a proposition of the form p and not-p. To
forbid its occurrence is to adopt one system of expression, which may recommend
itself highly. This does not mean we cannot use contradiction." [Wittgenstein
(1979a), p.71. Italic emphasis in the original.]
"I am prepared to predict that there will be
mathematical investigations of calculi containing contradictions, and people
will pride themselves on having emancipated themselves from consistency too."
[Waismann (1979), p.139.]
And we have
already
encountered this passage (which is directly reminiscent of Hegel and/or
Engels):
"But
you can't allow a contradiction to stand! -- Why not?...
"It
might for example be said of an object in motion that it existed and did not
exist in this place; change might be expressed by means of contradiction."
[Wittgenstein (1978), p.370.]
Hence, it is
now quite plain why
Turing, according Wittgenstein,
believed that he (Wittgenstein) was trying to introduce "Bolshevism" into
Mathematics: because of his criticisms of the irrational fear of contradictions
among mathematicians. [Cf., Monk (1990),
pp.419-20; see also Hodges (1983), pp.152-54.]
The above comments (and there were plenty more like it in the 1930s) plainly demonstrate
the influence both of DM and Hegel on Wittgenstein's thought. Which other major
(or even minor) non-Hegelian philosopher was arguing along these lines in the
1920s and 1930s? Which one was surrounded on all sides by active and leading
Marxists with whom he regularly discussed DM?X21
Indeed, and
more recently, Graham Priest and Richard Routley, both Marxists,
thought that they could recruit Wittgenstein to the
Paraconsistent and
Dialetheic 'cause', even if they had a few
criticisms to make of certain aspects of his work:
"Though dialetheism is not a new view, the
word itself is. It was coined by
Graham
Priest
and
Richard
Routley
(later Sylvan) in 1981 (see Priest, Routley and Norman, 1989, p.xx).
The inspiration for the name was a passage in Wittgenstein's Remarks on the
Foundations of Mathematics, where he describes the Liar sentence ('This
sentence is not true') as a Janus-headed
figure facing both truth and falsity (1978, IV.59). Hence a di-aletheia is a
two(-way) truth. Unfortunately, Priest and Routley forgot to agree how to spell
the 'ism', and versions with and without the 'e' appear in print." [Priest, Berto
and Weber (2018), quoted from
here. Links added.]
The passage
to which the above two authors refer reads as follows:
"Why should Russell's contradictions not be conceived as
something supra-propositional, something that towers above the propositions and
looks in both directions like a Janus head? N.B. the proposition F(F) -- in
which F(ξ) =
~ξ(ξ)
-- contains no variables and so might hold as something supra-logical, as
something unassailable, whose negation itself in turn only asserts it.
Might one not even begin logic with this contradiction? And as it were
descend from it to propositions.
"The
proposition that contradicts itself would stand like a monument (with a Janus
head) over the propositions of logic." [Wittgenstein (1978), p.256. Italic
emphases in the original.]
However, in
his later work, it is clear that Wittgenstein abandoned this way of seeing
things:
"There can be no debate about whether these
or other rules are the right ones for the word 'not'.... For without these
rules, the word has as yet no meaning; and if we change the rules, it now has
another meaning (or none), and in that case we may just as well change the word
too." [Wittgenstein (2009), §549, footnote, p.155e.
(This links to a PDF.)]
Hence, if
the negative particle typically maps a truth onto a falsehood, or vice versa,
then a contradiction can't be true, but must either be senseless or false....
Wittgenstein's unorthodox view of contradictions isn't the only area of his work
where it looks like he was echoing ideas we normally associate with Hegel and DM; he
also held non-standard views about the so-called 'Law
of Identity' [LOI]:
"Roughly speaking, to say of two things that they are identical is nonsense, and
to say of one thing that it is identical with itself is to say nothing at all.
"Thus I do not write 'f(a,b).a = b', but 'f(a,a)' (or
'f(b,b)'); and not 'f(a,b).~a =
b', but 'f(a,b)'....
[Wittgenstein explains what he is doing here:
"Identity of object I express by identity of sign, and not by using a sign for
identity. Difference of objects I express by difference of sign." (5.53,
p.105.) -- RL.]
"'A thing is identical with itself.' -- There
is no finer example of a useless sentence.... It is as if in our imagination we
put a thing into its own shape and saw that it fitted." [Wittgenstein (2009),
§216, p.91e. (This links to a PDF.)]
"'a = a' is a perfectly useless proposition."
[Wittgenstein (1976), p.283.]
"The law of identity, for example, seemed to
be of fundamental importance. But now the proposition that this 'law' is
nonsense has taken over this importance." [Wittgenstein (1993), p.169.]
I can think
of very few Analytic Philosophers (who have not already been influenced by
Hegel and/or Wittgenstein), if any, who would argue this way. But, this isn't surprising
given what we know of the opinions of his communist friends.
These are just a few of the passages I have
quoted in the above
Essay, which show that, in some respects, Wittgenstein had been heavily influenced by Hegel and
(possibly even) by Engels.
9. This
is a phrase Wittgenstein did not, of course, use. However, it is important to note
that even in his early work Wittgenstein was at pains to avoid the
reification of language. In later writings he acknowledged that in his
earlier work he hadn't gone far enough in that direction; indeed, it became
clear to him that he had overlooked several other important features of language,
among which were: its diversity, its origin in human practice, and
its role intercommunication. In short, he admitted he had ignored the "anthropological"
nature of discourse. [This is reported in Monk (1990), p.261. On this in
general, see here.
Also see
Hacker (2013c).]
11.
Cf., Parrington (1997), and Holborow (1999). Sean Doherty has also contributed
to the debate in a recent review of Holborow's book. [Cf., Doherty (2000).]
However, he adds little that is relevant to the aims of this Essay in addition to what Parrington and Holborow have to say. Even more recently, Dave McNally has also
weighed in;
cf., McNally (2001, 2004). On McNally, see
here.
[I will add
several more comments on McNally's work in a later re-write of this Essay. In
addition, I will examine Lecercle (2006) and Lecercle and Riley (2004). I will also be
examining the contribution of Chik Collins -- i.e., Collins (1999, 2000, 2003,
2004).]
It would require an entire Essay in itself to
review the many misconceptions, inconsistencies and confusions that litter Voloshinov's work. Hence, I have largely limited myself to
commenting only on those aspects of his theory that seem to be the most relevant
to issues raised at this site.
Nevertheless, and to
be fair to Voloshinov, there are a number insights in his writings that prefigure ideas
found in a more sophisticated and developed form in Wittgenstein's work: those connected
with the social nature of language and the meaning of words. On the first, see, for example,
Note 23, below; on the second, this:
"Meaning is a function of the
sign and is therefore inconceivable...outside the sign as some particular,
independently existing thing. It would be just as absurd to maintain such a
notion as to take the meaning of the word 'horse' to be this particular, live
animal I am pointing to. Why if that were so, then I could claim, for instance,
that having eaten an apple, I have consumed not an apple but the meaning of the
word 'apple'." [Voloshinov (1973), p.28.]
Compare that with Wittgenstein's comments,
aimed at what he alleged was
Augustine's view of language:
"These words, it seems to me,
give a particular picture of the essence of human language. It is this: the
individual words in a language name objects -- sentences are combinations of
such names. -- In this picture of language we find the roots of the following
idea: Every word has a meaning. This meaning is correlated with the word. It is
the object for which the word stands....
"It is important to note that
the word 'meaning' is being used illicitly if it is used to signify the thing
that 'corresponds' to the word. That is to confound the meaning of a name with
the bearer of the name. When Mr. N. N. dies one says that the bearer of
the name dies, not that the meaning dies." [Wittgenstein (1958), §1, p.2e, and §40,
p.20e. Italic emphasis in the original; Quotation marks altered to
conform with
the conventions adopted at this site.]
These parallels are quite striking. It would
be interesting, therefore, to find out whether the important intellectual stimulus
Wittgenstein received from his discussions with
Sraffa
actually came indirectly from Voloshinov or Vygotsky, or from some other common
source:
"Even more than
this…criticism I am indebted to that which a teacher of this university, Mr P.
Sraffa, for many years unceasingly practiced on my thoughts. I am indebted to this stimulus for themost consequential ideas of this book."
[Wittgenstein (1958), p.viii. Bold emphasis added.]
[On Sraffa's influence on Wittgenstein, see
Kitching and Pleasants (2002). For the 'Augustinian Picture' of language, see
Glock (1996), pp.41-45, Baker and Hacker (2005a), pp.1-28,
and Baker and Hacker (2005b), pp.48-72. See also
here. I have covered this issue
(concerning the likely influence on Wittgenstein of Sraffa, Bakhtin and
Voloshinov), here.
However, Wittgenstein's view of Augustine's theory of language wasn't entirely
accurate, but it did serve as a point of departure for him. On this, see Kirwan
(2001) and Burnyeat (1987).
(This links to a PDF.)]
Despite this, it is clear that Voloshinov's
ideas were seriously compromised by his adoption of traditional ways
of theorising about language, 'mind' and philosophy --, perhaps too much for him
to have appreciated the inconsistency and confusion this introduced into his
work. They will become apparent as this Essay unfolds. [On this, see, for
example, Note 23.]
12.The word "appears" has been put in 'scare' quotes because, if what Voloshinov
appears to have said were in fact correct, then not even he could
have begun to report on it, and neither could anyone else have determined what
other speakers might 'appear' to have said about his words
-- or, for that matter, about anything else on any subject whatsoever, including the sorts
of things
Parrington and Holborow have to say. Why this is so will be outlined presently in the
main body of this Essay.
12a.
We must be careful to distinguish between different forms of
contextualism. One type might emphasise the
circumstances surrounding an utterance. But, even these can vary. One sub-variety
here will speak about the social or historical circumstances, the interplay
between speakers, the physical setting, and the role an utterance might play in
situations like these. Another sub-variety might concentrate on the individual who utters a
given set of words, the psychological associations, 'values', beliefs, motives,
intentions and other idiosyncrasies they might bring to the occasion.
Voloshinov, Parrington and Holborow appear to have run both of these together
without giving much thought to the fact that they aren't (plainly) the same.
Indeed, this shows up in the obvious tension that exists in their work between
their desire to give a social account of language and meaning and what they
finally deliver, which is its exact opposite. [On this, again, see Note 23.]
The above meaning of "context" is to be contrasted with the sort
of sentential, grammatical and logical contextualism one finds in the Fregean/Wittgensteinian tradition
-- and at this site. On
that, see here.
13.
Open texture and "indexicality" will be tackled presently.
13a.
We may note in passing that here we have yet another dialectician
imposing
his theory on the facts. There is no way that Voloshinov could possibly have
known that the following is the case:
"Any word used in actual
speech possesses not only theme and meaning in the referential, or content,
sense of these words, but also value judgement: i.e., all referential contents
produced in living speech are said or written in conjunction with a specific
evaluative accent. There is no such thing as a word without evaluative
accent." [Voloshinov (1973), p.103. Italic emphasis in the original.
Bold emphasis added.]
In fact, there are countless words that don't possess an "evaluative accent" -- for instance, "the", "or", "and", "is",
"will be", "between", "skylight", "question", "comma",
"river", "orange peel", "Voloshinov", "etc.", etc.
Now, Voloshinov might have meant that there
is no use of words that doesn't also entail an evaluation of some sort, but,
clearly, that is a
separate matter -- and still not obviously true. Examples of several non-evaluative
sentences (even when used) will be given in the main body of this Essay,
presently.
13b.
It is worth adding here that this doesn't imply that I believe words
have an "intrinsic meaning"; I am simply alluding to ideas that Voloshinov is himself
criticising. On what I do think about meaning, see, for example,
here and
here.
"'When
I use a word,' Humpty Dumpty said, in a rather scornful tone, 'it means just
what I choose it to mean, neither more nor less.'
Figure Ten: Humpty
Dumpty Dialectics?
"'The question
is,' said Alice, 'whether you can make words mean so many different
things.'
"'The question
is,' said Humpty Dumpty, 'which is to be master -- that's all.'
"Alice was too
much puzzled to say anything; so after a minute Humpty Dumpty began again.
'They've a temper, some of them -- particularly verbs: they're the proudest --
adjectives you can do anything with, but not verbs -- however, I can
manage the whole lot of them! Impenetrability! That's what I say!'
"'Would you
tell me, please,' said Alice, 'what that means?'
"'Now you talk
like a reasonable child,' said Humpty Dumpty, looking very much pleased. 'I
meant by "impenetrability" that we've had enough of that subject, and it would
be just as well if you'd mention what you mean to do next, as I suppose you
don't mean to stop here all the rest of your life.'
"'That's a
great deal to make one word mean,' Alice said in a thoughtful tone.
"'When I make a
word do a lot of work like that,' said Humpty Dumpty, 'I always pay it extra.'
"'Oh!' said
Alice. She was too much puzzled to make any other remark.
"'Ah, you
should see 'em come round me of a Saturday night,' Humpty Dumpty went on,
wagging his head gravely from side to side, 'for to get their wages, you know.'
"(Alice didn't
venture to ask what he paid them with; so you see I can't tell you.)
"'You seem very
clever at explaining words, Sir' said Alice. 'Would you kindly tell me the
meaning of the poem called "Jabberwocky"?'
"'Let's hear
it,' said Humpty Dumpty. 'I can explain all the poems that ever were invented
just yet.'
"This sounded
very hopeful, so Alice repeated the first verse:
'Twas
brillig, and the slithy toves,
Did gyre and
gimble in the wabe:
All mimsy
were the borogroves,
And the mome
raths outgrabe.'
"'That's enough to begin with,' Humpty Dumpty interrupted: 'there are plenty of
hard words there. "Brillig" means four o'clock in the afternoon -- the
time when you begin broiling things for dinner.'
"'That'll do
very well,' said Alice: 'and "slithy"?'
"'Well, "slithy"
means "lithe and slimy." "Lithe" is the same as "active." You see it's like a
portmanteau
-- there are two meanings packed up into one word.'
"'I see it
now,' Alice remarked thoughtfully: 'and what are "toves"?'
"'Well, "toves"
are something like badgers -- they're something like lizards -- and they're
something like corkscrews.'
"'They must be
very curious-looking creatures.'
"'They are
that,' said Humpty Dumpty: 'also they make their nests under sundials -- also
they live on cheese.'
"'And what's to
"gyre" and to "gimble"?'
"'To "gyre"
is to go round and round like a gyroscope. To "gimble" is to make holes
like a gimlet.'
"'And "the
wabe" is the grass-plot round a sundial, I suppose?' said Alice, surprised
at her own ingenuity.
"'Of course it
is. It's called "wabe," you know, because it goes a long way before it,
and a long way behind it-----'
"'And a long
way beyond it on each side,' Alice added.
"'Exactly so.
Well then, "mimsy" is "flimsy and miserable" (there's another portmanteau
for you). And a "borogove" is a thin shabby-looking bird with its
feathers sticking out all around -- something like a live mop.'
"'And then "mome
raths"?' said Alice. 'I'm afraid I'm giving you a great deal of trouble.'
"'Well, a "rath"
is a sort of green pig: but "mome" I'm not certain about. I think it's
short for "from home" -- meaning that they'd lost their way, you know.'
"'And what does
"outgrabe" mean?'
"'Well, "outgrabing"
is something between bellowing and whistling, with a kind of sneeze in the
middle; however you'll hear it done, maybe -- down in the wood yonder -- and,
when you've once heard it, you'll be quite content. Who's been repeating
all that hard stuff to you?'
"'I read it in
a book,' said Alice."
If only dialecticians were as 'clear'
as Humpty Dumpty at
explaining the obscure terms in their 'theory'!
14a.
Of course, it could be pointed out that if enough speakers of a language
intend to use words in certain ways, then those intentions will, as a
matter of fact, change the meaning of those words for that sub-group, and maybe
later for the entire speech community. Look at how the word "sick",
for example, has changed recently for many young speakers of English.
Maybe so, but then "sick" has become a new
word, typographically identical with the old one. This new word has now come to
mean "crazy", "cool", or "insane" (with the words "crazy" and "insane", too,
having new meanings -- they no longer imply some sort of psychiatric disorder
when used this way), but no one supposes that when it is used in sentences like
"Is your sister sick? I saw her at the doctors today", and "My dog is sick; I
think I'll take it to the vets" the word "sick" means "crazy", "cool", or
"insane". Not even those who use "sick" to mean "crazy", "cool", or
"insane" when they are with their friends (who also speak this way) will fail to
understand someone who says they are going to be sick, or that they have to go
home to take care of their sick mother. And when the BBC inform us that "Glasgow
the UK's sickest city", they aren't trying to tell us how "crazy", "cool",
or "insane" it is.
To state the obvious: new words can certainly
be introduced into the language, and old words can assume new meanings. But, if
the meaning of any and all words were to change with each use, or their
meaning in the language were sensitive to occasions of use, and/or the
intentions and aims of speakers on each occasion of use, then we would surely
lose grip on the word "word" itself, not to mention "occasion", "intention" and
"aim"! [On this, see Note 15 and
later in the main body of this Essay.]
Nor is this to deny that words mean what they
do because human beings use them, but they use the words they chose to
utter or write because of the meaning they already have. Otherwise, any words would
do.
15.
One wonders what the supporters of "contextualism" have to say about a woman who
says "No!". Does this mean "No!", or do we have to concede ground to
rapists and male chauvinists who seem to think it means "Yes!" in certain
circumstances? And yet, if this word means what it says (in all circumstances),
then why
not other words?
16.
This seems to be a direct implication of Voloshinov's view! If a word is
never used, or if it has fallen out of use, then it must be meaningless.
That being so, in between its various episodic employments, that word must also cease to have a meaning. If so,
why would anyone want to use such a meaningless word? For example, does anyone
now actually want to use
"BuBuBu"? And yet, on this account, it would seem that "BuBuBu" is no different
from "cat", "door" or "coffee mug", since it would
seem that each of these is meaningless until it is used!
In fact, if this were so, no one would
bother to translate long dead languages. If an archaeologist unearthed an
ancient artefact that had, say, a strange inscription on it that was written in
an unknown language, she would surely be misguided in her attempt to translate it.
If the words on this inscription are no longer in use they must surely be
meaningless, given this 'theory'.
Alternatively, if these long dead words once had a
"theme" (when, say, they were originally written or inscribed), which was unique to that event, that
"theme" would be inaccessible since it has now become forever lost. In
either case, such an inscription would beuntranslatable.
Indeed, had
Jean-François Champollion (and others) taken a similar view of the
Rosetta
Stone, the meaning of Ancient Egyptian hieroglyphs would still be a mystery.
They would have no meaning if Voloshinov were correct -- so why would
anyone bother
trying to decipher them?
[There
are more details
on this,
here. Concerning another similar episode, the decoding of
Linear B,
see here.]
Update June 2013: We now read this
from the BBC about a new study of this manuscript:
"The message inside 'the world's most mysterious medieval manuscript' has eluded cryptographers, mathematicians and linguists for over a century. And for many, the so-called Voynich book is assumed to be a hoax. But a new study, published in the journal Plos One, suggests the manuscript may, after all, hold a genuine message. Scientists say they found linguistic patterns they believe to be meaningful words within the text. Whether or not it really does have any meaningful information, though, is much debated by amateurs and professionals alike. It was even investigated by a team of prominent code breakers during WWII who successfully cracked complex encrypted enemy messages, but they failed to find meaning in the text. The book has been dated to the early 1400s, but it largely disappeared from public record until 1912 when an antique book dealer called Wilfrid Voynich bought it amongst a number of second-hand publications in Italy.
"Marcelo Montemurro, a theoretical physicist from the University of Manchester, UK, has spent many years analysing its linguistic patterns and says he hopes to unravel the manuscript's mystery, which he believes his new research is one step closer to doing. 'The text is unique, there are no similar works and all attempts to decode any possible message in the text have failed. It's not easy to dismiss the manuscript as simple nonsensical gibberish, as it shows a significant [linguistic] structure,' he told BBC News.
"Dr
Montemurro
and
a
colleague
used
a
computerised
statistical
method
to
analyse
the
text,
an
approach
that
has
been
known
to
work
on
other
languages.
They
focused
on
patterns
of
how
the
words
were
arranged
in
order
to
extract
meaningful
content-bearing
words. 'There
is
substantial
evidence
that
content-bearing
words
tend
to
occur
in a
clustered
pattern,
where
they
are
required
as
part
of
the
specific
information
being
written,'
he
explains.
"'Over
long
spans
of
texts,
words
leave
a
statistical
signature
about
their
use.
When
the
topic
shifts,
other
words
are
needed. 'The
semantic
networks
we
obtained
clearly
show
that
related
words
tend
to
share
structure
similarities.
This
also
happens
to a
certain
degree
in
real
languages.' Dr
Montemurro
believes
it
unlikely
that
these
features
were
simply
'incorporated'
into
the
text
to
make
a
hoax
more
realistic,
as
most
of
the
required
academic
knowledge
of
these
structures
did
not
exist
at
the
time
the
Voynich
manuscript
was
created.
"Though
he
has
found
a
pattern,
what
the
words
mean
remains
a
mystery.
The
very
fact
that
a
century
of
brilliant
minds
have
analysed
the
work
with
little
progress
means
some
believe
a
hoax
is
the
only
likely
explanation."
[Quoted
from
here.
Accessed
23/06/2013.
Quotation
marks
altered
to
conform
with
the
conventions
adopted
at
this
site.
Bold
emphases
added;
link
in
the
original.
Some
paragraphs
merged.]
While another expert, a cryptographer, had
this to say:
"'There are about 25 examinations of the Voynich manuscript and most of the
results show the text has similarities with natural language. This new
examination is one more of this kind,' says Klaus Schmeh, a cryptographer. 'While we know a lot about the statistical properties of the text, we don't
know enough about how to interpret them, which is one of the problems with the
new research. We need to find out how different languages, encryption methods,
and text types influence the statistics.
There have been numerous encrypted texts since the Middle Ages and 99.9% have
been cracked. If you have a whole book, as here, it should be 'quite easy' as
there is so much material for analysts to work with. That it has never been
decrypted is a strong argument for the hoax theory.'" [Ibid.
Quotation marks altered to conform with the conventions adopted at this site.
Paragraphs merged.]
We also read the following:
"Gordon Rugg,
a mathematician from Keele University, UK,
is one such academic. He has even produced
his own complex code deliberately similar to
'Voynichese' to show how a text can appear
to have meaningful patterns, even though it
is 'gibberish hoax text'. He says the
new findings do not rule out the hoax
theory, which the researchers argue. 'The
findings aren't anything new. It's been
accepted for decades that the statistical
properties of Voynichese are similar, but
not identical, to those of real languages. I don't
think there's much chance that the Voynich
manuscript is simply an unidentified
language, because there are too many
features in its text that are very different
from anything found in any real language.'
"Gordon
Rugg
does not
believe
it
contains
an
unknown
code,
which is
another
theory
of what
the text
may be:
'Some of
the
features
of the
manuscript's
text,
such as
the way
that it
consists
of
separate
words,
are
inconsistent
with
most
methods
of
encoding
text.
Modern
codes
almost
invariably
avoid
having
separate
words,
as those
would be
an easy
way to
crack
most
coding
systems.' As to
its
enduring
appeal,
an
unsolved
cipher
could be
'hiding
almost
anything',
says
Craig
Bauer,
author
of
Secret
History:
The
Story of
Cryptology.
"'It
could
solve a
major
crime,
reveal
buried
treasure
worth
millions
or in
the case
of the
Voynich
manuscript,
rewrite
the
history
of
science,'
he adds.
Dr
Bauer's
opinion
of
whether
it is
meaningful
is often
swayed,
he
admits.
While he
recently
believed
it to be
a hoax,
the new
analysis
has now
shifted
his
opinion.
But
despite
this, he
still
believes
it is a
made up
language,
as
opposed
to a
real
naturally
evolving
one, or
'it
would
have
been
broken
years
ago. However,
I still
feel
that
it's
very
much an
open
question
and I
may
change
my mind
a few
times
before a
proof is
obtained
one way
or the
other.'
"But Dr
Montemurro
is firm
in his
belief,
and
argues
that the
hoax
hypothesis
cannot
possibly
explain
the
semantic
patterns
he has
discovered.
He is
aware
that his
analysis
leaves
many
questions
still
unanswered,
such as
whether
it is an
encoded
version
of a
known
language
or
whether
a
totally
invented
language. 'After
this
study,
any new
support
for the
hoax
hypothesis
should
address
the
emergence
of this
sophisticated
structure
explicitly.
So far,
this has
not been
done.
There
must be
a story
behind
it,
which we
may
never
know,'
Dr
Montemurro
adds." [Ibid.
Quotation
marks
altered
to
conform
with the
conventions
adopted
at this
site. Paragraphs
merged.]
And now, in February 2014, there is this:
"A
breakthrough has been made in attempts to
decipher a mysterious 600-year-old
manuscript written in an unknown language,
it has been claimed. The Voynich Manuscript,
carbon-dated to the 1400s, was rediscovered
in 1912, but has defied codebreakers since. Now,
Bedfordshire University's Stephen Bax says
he has deciphered two words, which could
lead to more discoveries. The manuscript,
which some think is a hoax, is full of
illustrations of plants and stars, as well
as text. It has been latched on to by
supporters of a whole range of strange
theories including some linking it to
Leonardo da Vinci or even aliens....
"In June last
year, Marcelo Montemurro, a theoretical
physicist from the University of Manchester,
UK, published a study which he believes
shows that the manuscript was unlikely to be
a hoax. Dr Montemurro and a colleague, using
a computerised statistical method to analyse
the text, found that it followed the
structure of 'real languages'.
"In February
this year, a paper published in the journal
of the American Botanical Council said one
of the plant drawings suggested a possible
Mexican origin for the manuscript. Prof Bax,
an expert in applied linguistics, said he
had been working on the Voynich Manuscript
for about two years. He said he had managed
to find the word for Taurus, alongside a
picture of seven stars (seen as part of the
zodiac constellation of Taurus) and the word
Kantairon alongside a picture of the herb
Centaury. Prof Bax
said he had been trying to crack the
manuscript using his knowledge of medieval
texts and his familiarity with Semitic
languages like Arabic.
"'I hit on
the idea of identifying proper names in the
text, following historic approaches which
successfully deciphered Egyptian hieroglyphs
and other mystery scripts, and I then used
those names to work out part of the script,'
he said. 'The manuscript has a lot of
illustrations of stars and plants. I was
able to identify some of these, with their
names, by looking at medieval herbal
manuscripts in Arabic and other languages,
and I then made a start on a decoding, with
some exciting results. My aim in reporting
on my findings at this stage is to encourage
other linguists to work with me to decode
the whole script using the same approach,
though it still won't be easy. But already
my research shows conclusively that the
manuscript is not a hoax, as some have
claimed, and is probably a treatise on
nature, perhaps in a Near Eastern or Asian
language.'
"Prof Bax
said he hopes a conference can be arranged
later this year, to bring together experts
on the manuscript." [Quoted from
here; accessed
19/02/2014. Quotation marks altered to
conform with
the conventions adopted at this
site. Several paragraphs merged.]
The incapacity of cryptographers to ascertain
the "theme" of manuscripts like this has clearly played no part in the success or
failure of researchers to decipher these texts (whether or not they have ever
heard of "theme", or care a fig about it).
But, let us suppose
The Voynich Manuscriptis
a hoax; in that case, and if meaning were
occasion-, speaker-, author-, or intention-sensitive, the meaning of its words
would be plain; they would mean something like this: "You lot are idiots for not spotting
this hoax earlier!", or "How clever must I be for fooling you smart Alecs for so long!"
-- or
even "The meaning of this manuscript is that it hasn't got one!" -- or some such.
Indeed, cryptographers would be able come to a similar conclusion within minutes
of discovering such manuscripts, thus saving themselves the bother of having to
decode them. All they need do in such circumstances is reflect on their
own subjective response to these texts -- and that would be what these strange words mean.
That they
don't do this, and show no sign even
of entertaining for one second such a crazy strategy, suggests
that they don't think meaning is occasion-sensitive, at
least not here -- and that they are right to reject such an approach.
[On this, see Kennedy and Churchill (2004); see also
here, and, more recently,
here.]
"Last week, a history researcher and television writer named Nicholas Gibbs
published
a long article in the Times Literary Supplement about how he'd
cracked the code on the mysterious Voynich Manuscript. Unfortunately, say
experts, his analysis was a mix of stuff we already knew and stuff he couldn't
possibly prove. As soon as Gibbs' article hit the Internet, news about it spread
rapidly through social media (we
covered it at Ars too), arousing the skepticism of
cipher geeks and scholars alike. As Harvard's Houghton Library curator of
early modern books John Overholt
put it on
Twitter, 'We're not buying this Voynich thing, right?' Medievalist Kate
Wiles, an editor at History Today,
replied,
'I've yet to see a medievalist who does. Personally I object to his
interpretation of abbreviations.'
"The
weirdly-illustrated 15th
century book has been the subject of speculation and
conspiracy theories since its discovery in 1912. In his
article, Gibbs claimed that he'd figured out the Voynich
Manuscript was a women's health manual whose odd script was
actually just a bunch of Latin abbreviations. He provided
two lines of translation from the text to 'prove' his point.
"However, this isn't
sitting well with people who actually read medieval Latin.
Medieval Academy of America director Lisa Fagin Davis
told The Atlantic's Sarah Zhang, 'They’re not
grammatically correct. It doesn’t result in Latin that makes
sense.' She added, 'Frankly I’m a little surprised the TLS
published it... If they had simply sent to it to the
Beinecke Library, they would have rebutted it in a
heartbeat.' The Beinecke Library at Yale is where the
Voynich Manuscript is currently kept. Davis noted that a big
part of Gibbs' claim rests on the idea that the Voynich
Manuscript once had an index that would provide a key to the
abbreviations. Unfortunately, he has no evidence for such an
index, other than the fact that the book does have a few
missing pages.
"The idea that the book
is a medical treatise on women's health, however, might turn
out to be correct. But that wasn't Gibbs' discovery. Many
scholars and amateur sleuths
had already
reached that conclusion, using the same evidence that
Gibbs did. Essentially, Gibbs rolled together a bunch of
already-existing scholarship and did a highly speculative
translation, without even consulting the librarians at the
institute where the book resides.
"Gibbs said in the
TLS article that he did his research for an unnamed
'television network.' Given that Gibbs' main claim to fame
before this article was a series of books about
how to write and sell television screenplays, it seems
that his goal in this research was probably to sell a
television screenplay of his own. In 2015,
Gibbs did an interview where he said that in five years,
'I would like to think I could have a returnable series up
and running.' Considering the
dubious accuracy of many History Channel
'documentaries,' he might just get his wish." [Annalee
Newitz at the Ars Technica website, accessed
03/01/2018. Quotation marks altered to conform with the
conventions adopted at this site. Links in the original.
Several paragraphs
merged.]
Update August 14 2019: Although,
this article in The
Washington Post (pay wall!) says every attempt so far to decipher this
manuscript has in fact failed.
However, not one single expert involved in this controversy has bothered
to point
out the obvious -- or, what is 'obvious' to DM-fans who look to
Voloshinov for guidance -- that since no one knows the "theme" of this manuscript,
it can't be decoded. Don't they 'understand' dialectics, for goodness sake!
It could be argued, once again, that written
and spoken languages are different. But, that isn't so; on that, see
here.
Finally, someone could object that Voloshinov has an answer to this in his
distinction between "theme" and meaning. That seemingly promising life-line will
be severed,
presently.
17.It would be interesting to see how defenders
of Voloshinov's 'theory' try to extricate him from this particular quagmire by
uttering or writing nothing but meaningless sentences!
[Once more, on Voloshinov's appeal to
"standard meanings", see Note 18a0,
below.]
Again, it could be objected that this
confuses "theme" and meaning, or it ignores the distinction between
them. This topic will be examined
presently, where we will see
that Voloshinov himself confuses these two words/'concepts'!
18.Of course, since these
"possible meanings" -- expressed in some language or other -- haven't actually
been uttered yet, the words comprising them must be
meaningless, too! In which case, there can be no "possible meanings" for an
uninterpreted utterance to mean, if Voloshinov were correct.
And, even if the above were based on a
misunderstanding of what Voloshinov himself meant by his theory (irony
intended), these hypothetical "possible meanings", if expressed in some language
or other, will face the same occasionalist protocols: that they, too, must
be occasion-sensitive and in need of interpretation, and hence susceptible to their own
"possible meanings"..., and so on.
On the other hand, if this (looming) infinite
regress is capable of being stopped at any point -- i.e., if we take some utterance, or
some "possible meaning", to be sui generis
comprehensible without reference to its own occasion of
utterance or interpretation --, then there seems to be no good reason why this can't
apply to the original utterance itself.
18a0.
It could be argued that an utterance would have a meaning (and known to have
that meaning) by the one uttering it. But, if so, that meaning must be known
before it was uttered, otherwise the one doing so wouldn't have uttered it,
but would have come out with a different utterance. But, if that were so, the
meaning of that utterance can't be occasion-sensitive since it hasn't been
used yet! Either that, or it would have no meaning since it hadn't been uttered
in a given situation.
It could be replied that whoever utters a
given sentence knows what she intends that sentence to mean (in those
circumstances) before the relevant words leave her mouth. But, if that were
so, then, and once more, any words would do, and we would be back where we
were a paragraph ago.
On the other hand, if a specific set of
words is chosen, then those words must have standard meanings, and known to
have them -- or, and once again, any old words would do. And yet, if any old
words won't do, then meaning can't be occasion-sensitive.
It is worth pointing out that a standard
meaning isn't the same as an intrinsicmeaning.
I
return to this issue in the
main body of this Essay, where I discuss what I have called "immediate languages".
18a. Indeed, if the meaning of words changed with each utterance, the
Oxford English Dictionary, for example, would need updating several hundred thousand
times per second, twenty-four hours a day, seven days a week!
[Again, on Voloshinov's appeal to
"standard meanings", see Note 18a0, above.]
18b.
It isn't easy to decide if every occasion of utterance is indeed unique
since the phrase "occasion of utterance" is itself hopelessly vague! Is it a
geographical location? If so, how wide should the net be cast? Are the relevant
surroundings of an utterance circumscribed by a circle with radius five metres, ten
metres, fifty metres? Or, is it an ellipse, a hexagon? If we go into three
dimensions, is it a sphere, an
ellipsoid,
an
hyperboloid, or some other
quadric?
Or, is the time/timing of the utterance
decisive, here? Once again, are we talking about the exact length of time the relevant
words took to be uttered, or a few seconds either side? Do we all therefore have
to carry around with us accurate stopwatches to make sure we don't become too
generous (or too restrictive) -- or too random -- in our timing of any given
utterance?
Alternatively, do interpersonal,
psychological considerations enter in? If so, which of these countless
"considerations" are relevant? And where can we access or download the manual that
informs us what is and what isn't allowed? Or, do we all just have to guess?
It could be objected that the above questions
are far too pedantic and nit-picking. But, these are the sort of questions that a
genuine science would be ready to ask and then answer. Not so with
Mickey Mouse
Dialectical 'Science'.
19.
As we will soon
see, there are
good reasons for rejecting this way of conceptualising language. Little sense can be made of such
a tightly constrained speaker's
idiolect --
i.e., some form of idiosyncratic "inner speech".
20.
This use of the word "know" isn't meant to relate to factual knowledge, since,
manifestly, few of us are aware of this facility. It is more akin to the
distinction drawn by
Gilbert Ryle
sixty or more years ago between "knowing that" and "knowing how". Here, it is as
much "knowing when" as it is "knowing how". So, just as we know when to put one
foot in front of another as we walk -- which knowledge can't be reduced to any
form of "knowing that" --, competent speakers know when certain
combinations of words can sensibly be
uttered, and thus the sorts of contexts in which they can plausibly appear or be
used.
Hence, we don't expect someone, for example, to comment on the remarkable size of Blue
Whales or the tensile strength of steel when they have been asked what sort of
day they have had at work, or what the time of the next train is. And that is why several of the sentences listed in
the main body of this Essay will stand out as extraordinarily odd, since the sort of
contexts in which they might occur, or be uttered, aren't those we meet every day, if
at all.
So, for
instance, the following sentences appear rather strange (situated where they are
on this page -- or your screen) since we can all imagine circumstances in which
it would be appropriate to utter them, none of which seem relevant:
V1: "Move, and you're dead!"
V2: "Your house has just
burnt down!"
V3: "Those pickets will stop
you strike-breaking!"
V4: "The Nazis know where you
live!"
V5: "Margaret Thatcher is
your biological mother!"
V6: "Tony Blair really
admires you!"
However, if Voloshinov's theory were correct,
and anything could mean anything, then each of the above will be appropriate in
any and all circumstances. Voloshinov has a reply of sorts to this which will be considered
below.
It is worth adding at this point that many
sentences that we either encounter, utter or write are completely novel, as indeed are one
or two of those listed above. For example, it is highly unlikely that anyone (other than
Thatcher's children, perhaps) has ever
heard or uttered, say, V5 before. But, as pointed out above, we can all imagine situations where
it would be appropriate to come out with V5. And that is because we already understand it.
Of course, this isn't to suggest that things are always this clear cut; however,
further consideration of this topic will take us too far away from the main aims of
this Essay. [Those who wish to explore this topic further should begin with the work
of Paul Grice
-- as well as the books and articles listed
here --
see, for example, Grice (1989).]
Ryle's distinction can be found in Ryle
(1946,
1949a)
-- this links to a PDF. Incidentally, I am well aware that this distinction has been
challenged recently -- for example, in Stanley and Williamson (2001). On this, see the
reply in Rosefeldt (2004) and Wiggins (2009, 2012). See also Brown (1970).
20a.
Of course, part of the problem here is that the word "meaning" is
itself highly
ambiguous. I tackle that particular 'difficulty'
here and
here -- and, indeed, throughout much of
the next two main sub-sections.
21.
Just because I have used "BBB XXX ZZZ QQQ TTT" to make the point that it is
meaningless that doesn't imply that "BBB XXX ZZZ QQQ TTT" means
"BBB XXX ZZZ QQQ TTT is meaningless". If it did then clearly "BBB XXX ZZZ QQQ
TTT" would mean "BBB XXX ZZZ QQQ TTT is meaningless", which in turn
would mean that "BBB XXX ZZZ QQQ TTT" isn't meaningless after
all! In which case, "BBB XXX ZZZ QQQ TTT" would imply that BBB XXX ZZZ QQQ TTT
is meaningless and BBB XXX ZZZ QQQ TTT isn't meaningless!
Nor does it mean that just because I
intended to show that "BBB XXX ZZZ QQQ TTT" is meaningless
that it is meaningless just because that was so intended by me. It was
meaningless before I used it, and after. If we exclude the possibility that this string of
letters is some sort of code, or is intended to be a code (on this, see below), intentions
can't turn babble into sense, nor
the other way round. But, that fact didn't prevent the present author from using "BBB XXX
ZZZ QQQ TTT" to point out that it was indeed meaningless. Neither does it
prevent anyone else understanding the present author's (speakers') meaning to
that end, even
though whatever was, or could be said by using "BBB XXX ZZZ QQQ TTT" is linguistically
meaningless, for all that.
Now, while the above
shenanigans might not have been without meaning (i.e., they weren't pointless), the
inscription used (i.e., "BBB XXX ZZZ QQQ TTT") was nonetheless
meaningless. But, this isn't a distinction a follower of Voloshinov can make. [On that, see
here and
here.]
This also shows that linguistic meaning
isn't based on the itinerant use to which sounds
or inscriptions are put.
It
could be argued that someone could intend "BBB XXX ZZZ QQQ TTT" to mean, say, "Memory card in office safe", as a sort
of code. In that case, it would seem that intentions can
create a meaning from
mere babble. But, even here, "BBB XXX ZZZ QQQ TTT" is code for "Memory
card in office safe", not that it literally means this in English. No one is going to
revise a dictionary and list as one of the synonyms of, say, the word "card" the letters
"XXX". The intention here is clearly that the recipient of the code should use
a code-decode manual, or key, to interpret the secret letters, not revise the language!
[On this, also see Note 14a.] In
this case, "BBB XXX ZZZ QQQ TTT" would acquire a (temporary) coded meaning from
the manual used to translate out of, and then back into, English (or some other natural
language as the case may be). Here, it is clear that "coded meaning" is a variant on "speaker's
meaning".
Of course, the problem is that it is all too
easy to conflate speaker's meaning with meaning in language. So, someone could
mean by "XXX" (speaker's meaning) "card" (for example, as part of the above
code), but, as we have seen, that wouldn't alter the meaning of "card" in
English. Clearly, in this case, "XXX" would gain its (temporary)
speaker's meaning from a word already in use and which already has a meaning in
English. It is only because of this that "XXX" could gain this temporary coded
meaning.
In addition, this
entire topic has been complicated by the fact
that "meaning" has many meanings already. [On that, see
here.] Furthermore, this is
further aggravated by the conflation of the meaning of words with the meaning,
and then the sense, of a sentence. As it stands, "BBB XXX ZZZ QQQ TTT" is
both a nonsensical and a non-sensical string of letters. [On the
distinction between "nonsense" and "non-sense" see
here.] It has no
meaning in language even though, when used, it has a speakers' meaning in this
Essay.
22.
This point doesn't undermine Wittgenstein's notorious claim that meaning and
use are the same -- since he never made such a claim!
What Wittgenstein actually said was this:
"For a large class of
cases -- though not for all -- in which we employ the word 'meaning' it can be
defined thus: the meaning of a word is its use in the language." [Wittgenstein
(1958), §43, p.20e. Italic emphasis in the original; Quotation
marks altered to conform with the conventions adopted at this site.]
Not everything that has a use has a meaning,
and not everything with a meaning will have that meaning connected with its use
(narrowly conceived). [For an example of a string of letters that has a use but
no meaning see Note 21, above.]
Several of the fine distinctions that can be drawn
here
alone are explored in great detail in Hallett (1967); cf., also Hallett (1984, 1988,
1991, 2008). See also Baker and Hacker (2005a), pp.129-58, Hanfling (1989),
pp.31-54, Hanfling (2000), Hacker (2010), Horwich (2010), Ryle (1971c, 1971d), Skorupski (1997),
and Whiting (2010b). However, these works (and, indeed, Wittgenstein's own comments)
should be read in the light of Kuusela (2005, 2006, 2008). Although, it should
be pointed out that many of the above also tend to run together the meaning of
words and the meaning, and then the sense, of a sentence.
Nevertheless,
even though it represented an important advance, Wittgenstein's discussion of
meaning is in the end unsatisfactory since he too failed to consider the many and
varied meanings of
the word "meaning" itself! [On that, see
here.] Wittgenstein wasn't a systematic thinker.
23. For example, Voloshinov's view that the
use of language (i.e., "signs") is a social not an individual psychological
phenomenon is underlined in the following passage:
"In point of fact, the speech
act, or more accurately, its product -- the utterance, cannot under any
circumstances be considered an individual phenomenon in the precise meaning of
the word and cannot be explained in terms of the individual psychological or
psychophysiological conditions of the speaker. The utterance is a social
phenomenon." [Voloshinov (1973), p.82. Boldemphasis alone
added.]
However, several other things Voloshinov
says run counter to this quasi-Wittgensteinian idea (that communication is
primary, and that 'acts of cognition' aren't private, 'inner' phenomenon). For example,
he tells his readers the following:
"The distinction between
theme and meaning acquires particular clarity in connection with the problem
of understanding….
"Any genuine kind of
understanding will be active and will constitute the germ of a response. Only
active understanding can grasp theme -- a generative process can be grasped only
with the aid of another generative process.
"To understand another
person's utterance means to orient oneself with respect to it, to find the
proper place for it in the corresponding context. For each word of the utterance
that we are in process of understanding, we, as it were, lay down a set of our
own answering words. The greater their number and weight, the deeper and more
substantial our understanding will be.
"Thus each of the
distinguishable significative elements of an utterance and the entire utterance
as a whole entity are translated in our minds into another, active and
responsive, context. Any true understanding is dialogic in nature….
Understanding strives to match the speaker's word with a counter word…."
[Ibid., p.102. Bold emphases alone added.]
"Idealism and psychologism
alike overlook the fact that understanding itself can come about only within
some kind of semiotic material (e.g., inner speech), that sign bears upon sign,
that consciousness itself can arise and become a viable fact only in the
material embodiment of signs...understanding is a response to a sign with
signs." [Ibid., p.11. Italic emphasis in the original. Bold emphasis
added.]
"In the first instance,
to
understand means to refer a particular inner sign to a unity consisting of
other inner signs, to perceive it in the context of a particular psyche....
"Self-observation
(introspection) is the understanding of one's own inner sign.... We do not see or
feel an experience we understand it. This means that in the process of
introspection we engage our experience into a context made up of other signs we
understand. A sign can only be illuminated with the help of another sign."
[Ibid., pp.35-36. Bold emphases alone added.]
So, while Voloshinov certainly gestures
toward a
sociological explanation of meaning, understanding and language, he actually opts for a
psychologistic and individualistic account of all three. Indeed, Meredith William's
judgement of Vygotsky could equally well be applied to Voloshinov:
"Vygotsky attempts to combine a social theory of
cognition development with an individualistic account of word-meaning.... [But]
the social theory of development can only succeed if it is combined with a
social theory of meaning." [Williams (1999), p.275.]
[The rest of that chapter in William's book
(i.e., Chapter Ten, 'Vygotsky's Social Theory of Mind', pp.260-81) is highly
relevant.]
It could be argued that this is unfair, since
Voloshinov repeatedly says that all signs, both inner and outer, are
sociological -- for example, on pp.12, 34, 37, and 82.
However, even though Voloshinov says
that all signs are social, it is also true that for him their processing is an entirely private,
'inner' affair. This is no more a social picture of meaning than would be one
concerning, say, the activities of coin hoarders, if they ignored the monetary
value of the coins
they collected and
attached to them their own sentimental or idiosyncratic value, and thereby
thought that the value of these coins depended on what they thought and felt
about them. Monetary value is
social, just as coins are social products; but idiosyncratic value isn't.
The important thing here isn't the response
of sign upon sign, as Voloshinov seems to think, but the social use of words.
Signs themselves are lifeless objects (as Voloshinov also seems to have acknowledged); it
takes human beings to give them life, and this takes place in a public, not a private, arena. As Peter Hacker
points out:
"It is indeed true that a
sign can be lifeless for one, as when one hears an alien tongue or sees an
unknown script. But it is an illusion to suppose that what animates a sign is
some immaterial thing, abstract object, mental image or hypothesised
psychic entity that can be attached to it by a process of thinking.
[Wittgenstein (1969), p.4: 'But if we had to name anything which is the life of
the sign, we should have to say that it was its use.'] One can try to rid
oneself of these nonsensical conceptions by simple manoeuvres. In the case of
the idealist conception, imagine that we replace the mental accompaniment of a
word, which allegedly gives the expression its 'life', by a physical correlate.
For example, instead of accompanying the word 'red' with a mental image of red,
one might carry around in one's pocket a small red card. So, on the idealist's
model, whenever one uses or hears the word 'red', one can look at the
card instead of conjuring up a visual image in thought. But will looking at a
red slip of paper endow the word 'red' with life? The word plus sample is no
more 'alive' than the word without the sample. For an object (a sample of red)
does not have the use of the word laid up in it, and neither does the
mental image. Neither the word and the sample nor the word and the mental
pseudo-sample dictate the use of a word or guarantee understanding.
"...It seemed to
Frege,
Wittgenstein claimed, that no adding of inorganic signs, as it were, can make
the proposition live, from which he concluded that [for Frege -- RL] 'What must
be added is something immaterial, with properties different from all mere
signs'. [Wittgenstein (1969), p.4.] He [Frege -- RL] did not see that such an
object, a sense mysteriously grasped in thinking, as it were a picture in which
all the rules are laid up, 'would itself be another sign, or a calculus to
explain the written one to us'. [Wittgenstein (1974a), p.40.] ... To understand a
sign, i.e., for it to 'live' for one, is not to grasp something other than the
sign; nor is it to accompany the sign with an inner parade of objects in
thought. It is to grasp the use of the sign itself." [Hacker (1993a), pp.167-68.
Italic emphases in the original. Link added.]
Hacker then refers his readers to the following passage of
Wittgenstein's:
"As for wishing or
understanding being merely the expression of the wish or the thought, the
ordinary objection to it is that no mere sign is the thought. The thought
interprets the sign. Thinking is not speaking or reading the symbols. Such an
objection is rooted in the view that thinking, or some process in the mind,
accompanies the symbols. Now is this supposed process something amorphous, a
state having duration while the sentence is said, written, or heard? Perhaps it
is something articulate, so that understanding a sentence consists of a series
of interpretations for each word. The process would be translatable into a
sentence, so that we could derive the sentence from the process or the process
from the sentence. But this only adds one phenomenon to another." [Wittgenstein
(1979a), pp.53-54.]
But,
the "addition of one phenomenon to another" gets us nowhere, for the whole point
of this 'explanation of meaning', or of 'thought', was to bypass an appeal to
phenomena, since phenomena can't determine the sense of what we say or how we mean; nor can words
or 'signs', since they, too, are just lifeless phenomena (again, as Voloshinov
himself agrees).
Of course, to suppose otherwise --, i.e., to imagine that words,
or their 'inner representations', determine their own meaning independently of the use to which humans
put them in material contexts -- would be to fetishise them, as noted
above.
Indeed, this would be
tantamount to believing that words (or their 'inner representations') enjoy a social life
of their own anterior to,
and explanatory of, the linguistic communion that takes place between human beings.
If words (etc.) did in fact acquire their own meanings, piecemeal, in
that manner, and those meanings followed words (etc.) about the place like
a shadow, then the idea that language is a social phenomenon would assume
entirely different implications. In that case, discourse would still be social, but
that would be because words (etc.) were the social beings here -- which would
in turn mean that they
had passed that property on to our use of language! Humanity would be
social because our words already were!
We are now in a position to understand
why that is so: the supposition that a word (or, at least,
its physical embodiment, its 'inner representation', perhaps) can motivate a
human agent (causally or in any other way)
to regard it as the repository of its own meaning -- so that inferences can be
made from ink marks on the page (or from 'images', 'ideas', and 'representations' in the
head) to super-empirical truths about
'Being', or whatever -- would be to misconstrue the products of
the social relations among human beings (i.e., words) as if they were their own autonomous
semantic custodians, as creators and carriers of meaning themselves. In
effect, this would be to anthropomorphise words, treating them as if they had
their own history, social structure and mode of development. In this way, the
social nature of language would reappear in an inverted form as an expression of
the social life of words (etc.). Humanity would be atomised, linguistic signs
(etc.) socialised! [See also
here.]
But, this is precisely the corner into which Voloshinov
has painted himself. He clearly wants to maintain that signs are social, but his
theory holds that our understanding is an 'inner' process, hidden away from
social practice. In the end, despite his claims to the contrary, and his
obvious intentions, for him, it is the social relation among the
signs we have in our
heads that determines meaning, not the social relation among human beings
that decides what our words mean:
"In the first instance,
to
understand means to refer a particular inner sign to a unity consisting of
other inner signs, to perceive it in the context of a particular psyche....
"Self-observation
(introspection) is the understanding of one's own inner sign.... We do not see or
feel an experience we understand it. This means that in the process of
introspection we engage our experience into a context made up of other signs we
understand.A sign can only be illuminated with the help of another sign."
[Voloshinov (1973), pp.35-36. Bold
emphases alone added.]
The following
parable, taken from
Jonathan
Swift's
Gulliver's Travels, lampoons the
idea that we could say things simply with objects -- or just with their names --, or
even that understanding is facilitated by comparing sign with sign (as
Voloshinov believed), or to put
it more pointedly, with anything "in our heads", which are just more "phenomena":
"We next went to the School of
Languages, where three Professors sat in Consultation upon improving that of
their own country. The first Project was to shorten
Discourse by cutting Polysyllables into one, and leaving out Verbs and
Participles, because in reality all things imaginable are but Nouns.
"The other, was a Scheme for entirely
abolishing all Words whatsoever; and this was urged as a great Advantage in
Point of Health as well as Brevity. For it is plain, that every Word we speak is
in some Degree a Diminution of our Lungs by Corrosion, and consequently
contributes to the shortening of our Lives. An Expedient was therefore offered,
that since Words are only Names for Things, it would be more convenient for all
Men to carry about them, such Things as were necessary to express the particular
Business they are to discourse on. And this Invention would certainly have taken
Place, to the great Ease as well as Health of the Subject, if the Women in
conjunction with the Vulgar and Illiterate had not threatened to raise a
Rebellion, unless they might be allowed the Liberty to speak with their Tongues,
after the manner of their Ancestors; such constant irreconcilable Enemies to
Science are the common People. However, many of the most Learned and Wise adhere
to the New Scheme of expressing themselves by Things, which hath only this
Inconvenience attending it, that if a Man's Business be very great, and of
various kinds, he must be obliged in Proportion to carry a greater bundle of
Things upon his Back, unless he can afford one or two strong Servants to attend
him. I have often beheld two of those Sages almost sinking under the Weight of
their Packs, like Pedlars among us; who, when they met in the Streets, would lay
down their Loads, open their Sacks, and hold Conversation for an Hour together;
then put up their Implements, help each other to resume their Burthens (sic),
and take their Leave.
"But for short Conversations a Man may
carry Implements in his Pockets and under his Arms, enough to supply him, and in
his House he cannot be at a loss: Therefore the Room where Company meet who
practise this Art, is full of all Things ready at Hand, requisite to furnish
Matter for this kind of artificial Converse." [Gulliver's Travels, The
Voyage to Balnibari,
Chapter 5. Some paragraphs merged.]
It could be objected that showing one another
various objects isn't at all the same as comparing sign with sign. But,
Voloshinov argued that anything could serve as a sign:
"Signs are particular,
material things; and, as we have seen,any item of nature, technology, or
consumption can become a sign, acquiring in the process a meaning that goes
beyond its particularity. A sign does not simply exist
as a part of reality -- it reflects and refracts another reality. Therefore it
may distort that reality or be true to it, or it may perceive it from a special
point of view. Every sign is subject to the criteria of ideological evaluation
(i.e., whether it is true, false, correct, fair, good, etc.). The domain of
ideology coincides with the domain of signs. They equate with one another.
Wherever a sign is present, ideology is present, too. Everything ideological
possesses semiotic value." [Voloshinov (1973), p.10. Bold
emphasis alone added.]
In that case, no one who agrees with Voloshinov can
legitimately object to the above lampoon by Swift -- which reveals the absurdity of
thinking that comprehension can be achieved by 'comparing signs'. In the above,
it is clear that signs are the agents and we are the patients.
There are other fatal implications for
Voloshinov's theory expressed in and by the above
passage; they will be unpacked as this Essay unfolds. [On this, see
for example, Note
41a.]
24.
This idea first appeared in published form in Wittgenstein (1972),
pp.39-41; 4.02-4.03 (this links to a PDF), which appeared in 1921. The idea also
saw light of day in his unpublished 'Notes on Logic' [Wittgenstein (1913b)], and in
Frege's unpublished work (written in 1914),after discussions with Wittgenstein, as well as in an essay
of his, published in
1923. [(Frege (1914), p.225; Frege (1923).]
This idea has since been seized upon by Chomsky, the Chomskyans, and several
Analytic Philosophers, who transmogrified it into a theory of meaning (Michael
Dummett) or a theory of the computational nature of the 'human mind'
(Chomsky, et al) -- on this see, for example, Hacker (2013a), pp.136-44.
Although Hacker is rightly critical of the misapplication of this idea, he tends
to run-together Wittgenstein's and Frege's versions of it with that of the above
theorists, and, as a result, risks undermining the distinction between the
meaning of a word and the sense of a proposition. There is no suggestion in
Wittgenstein's use of this idea that language needs a 'theory of meaning' (or,
indeed, a 'transformational grammar') so that speakers are able to utter, or form, indicative
sentences. However, Hacker neglected to make this point, which might suggest to
some that he is in fact criticising Wittgenstein, when it is reasonably certain
he isn't. Nor does
Hacker deny the obvious point that we all understand sentences we have
never read or heard before, even while we all fail to understand words we have never
encountered before. Nevertheless, readers of Hacker's work will be forgiven concluding the opposite!
He is so
intent on refuting dominant, contemporary theories of language and meaning that his usual
clarity of expression appears to have deserted him. [See also Baker and Hacker (1984a),
which contains a systematic demolition of all such theories. On the latter, see
Note 31.]
Concerning a different, but related, Fregean idea: on the origin of the linguistic version of The Context Principle, see
Land (1974), Chapter Four. While others
toyed around with this precept, it
assumed an entirely different significance in Frege's work, and yet another
again in Wittgenstein's. Here are Frege's comments:
"[N]ever ask for the meaning
of a word in isolation, but only in the context of a proposition." [Frege
(1953), p.x.]
"[W]e ought always to keep
before our eyes a complete proposition. Only in the context of a proposition
have the words really a meaning." [Ibid., p.71. Cf., also pp.73, 116.]
Plainly, Frege thought this is the case
since we typically
say things in sentences, which fact about us reveals what we mean by our words. Otherwise we
would typically say things using single words.
However, as pointed out later on in this
Essay, this is far too narrow a conception of meaning.
It also encourages the conflation
of the meaning of a word with the sense of a proposition, something Wittgenstein
had been at pains to distinguish, at least in his 'early' and 'middle' periods.
[On this see, Dummett (1956), Dummett (1981a), pp.6-7,
192-96, 495-505, Dummett (1981b), pp.360-427, Dummett (1991a), pp.180-240. For a
different view of Frege, see Baker and Hacker (1984b), pp.194-230, which is heavily
criticised in Dummett (1984, 1988).]
Wittgenstein quoted this Fregean principle (or alluded to it) several times in
both his earlier and later work:
"Only propositions have
sense; only in the nexus of a proposition does a name have meaning."
[Wittgenstein (1972), §3.3, p.25]
"An expression only has
meaning in a proposition." [Ibid., §3.314, p.27.]
"For naming and describing do
not stand on the same level: naming is a preparation for a description. Naming
is so far not a move in the language-game -- any more than putting a piece in
its place on the board is a move in chess. We may say: nothing has so far
been done, when a thing has been named. It has not even got a name except
in the language-game. This is what Frege meant too, when he said that a word had
meaning only as part of a sentence." [Wittgenstein (1958), §49, p.24.
Italic emphases in the original.]
On the connection between Frege and
Wittgenstein and the different slant they put on this principle, see Baker and
Hacker (1984b). Cf., also Baker and Hacker (2005a), pp.159-87, and
Reck (1997). [Also see,
Note 29andNote 86, below.]
In
fact, Voloshinov had his own context principle of sorts -- cf., Voloshinov
(1973), p.36. But, as we have seen (in
Note 23), the sort of context to which he refers is that of other dead signs. For
Wittgenstein, on the other hand, our words come to life when we use them,
typically in sentences.
25.
Naturally, this means that those involved in any 'conversation' using this word would need to
know what sort of word was being employed here (i.e., whether it was the
name of, or the name for something, whether it was a verb (describing or
expressing an
activity, state, performance (etc.)), or whether it was a modifier of some
sort (adjective, adverb, etc.), and so on). As Wittgenstein put it, they would
have to know what "station" the word occupied in language. [On this see, Harrison
(1979), Hanna and Harrison (2004). Concerning the distinction between a name
for and a name of something, see
here.]
Nevertheless, these comments (about what
speakers must know) should be read in the light of Note 20, above.
On what speakers do
know, see Devitt (2006). [However, the reader mustn't assume that I agree with everything Devitt says about
language and 'mind', particularly his insistence that 'thought'
precedes language --, a doctrine that seems at odds with much of the rest
of his book! On that, see here.]
Incidentally, "Bogomil"
is the name of a religious sect that operated in the
Balkans in the
Middle Ages.
[On this, see Lambert (1992).]
25a.
This isn't to deny that we also speak of the meaning of what is said, or of the
sentences used to say it. However, that doesn't undermine the distinction between the
meaning, and the sense, of an indicative sentence.
26.
This is a seriously neglected area in the Analytic Philosophy of Language. Despite
this,
there is a reasonably accessible account of some of the background to this in Hacking (1975).
A
more detailed picture can be found in Hallett (2008). Cf., also Skorupski
(1997). Although this distinction is widely recognised, as far as I know there
is no work that covers this topic in either a comprehensive or a satisfactory manner. [On
this, see Note 29 and
Note 30, below.] If anyone knows any different, please
e-mail me.
26a. It could be pointed out that we build
our sentences out of the meanings of the words they contain. In that way, our
sentences are actually constructed out of linguistic atoms. Hence, the distinction
between the meaning of a word and the sense of a sentence doesn't undermine
atomist theories of meaning, as was alleged in the main body of this Essay. On
that, see Note 29, below.
27.
These two aren't unrelated. That theme will be explored at greater length in
Essay Twelve Parts Two, Three, Four and Seven (summary
here).
28.
Has anyone ever tried to explain (successfully!), say, to a child, or to a foreign language
speaker, what a certain word means that they have never encountered before by appealing to the
intentions of the
individual
using it? That is, as opposed to explaining why a certain individual
might have said what they did, after the accepted meaning of that word has been
explained?
Of course, if the meaning of sentences and
words were both context-, and intention-dependent, interlocutors wouldn't bother
putting sentences together, they would just mouth
single words at one another, the meanings of which could be ascertained from
the intentions so to utter them. Hence, instead of
saying "I'm hungry", children would only ever come out with "Food".
Instead of
saying:
"Let us take an example…. [:]
I'm hungry conjures up a general concept. When, however, we look at
different contexts in which the phrase might be used, we see how the evaluative
accent changes everything. A child saying this to her mother might be
indirectly a request for the mother to get her something, an enquiry about what
there is to eat, or a statement that she just feels like something to eat.
One adult saying it to another might mean that it's time for lunch and be a
suggestion that they go somewhere to eat…. In each case the context is not
merely the gloss on the meaning but constitutes different meanings --
different inevery aspect…." [Holborow (1999), p.28. Bold emphasis added.]
Holborow could say…, er, well, what?
"Different", perhaps --, or "controversial"?
However, the complex ways we articulate
language should have suggested to those enamoured of Voloshinov's work that the meaning of
a word and
the sense of a sentence can't be given an occasionalist slant howsoever one
tries to do it.
29.
Such an account need not be general, nor need it be an explanation
of language (except in an elucidatory sense outlined
here) -- it
might simply describe it perspicuously (and that is so for reasons
outlined in Essay Twelve
Part One).
Other considerations that are relevant here
were well summarised by
David
Wiggins (the reader shouldn't conclude that I endorse everything it contains
-- particularly its indiscriminate use of "meaning" and "sense" in connection
both with words and sentences):
"It has been maintained by
Ryle
and
others that words have sense in an only derivative manner, that they are
abstractions rather than extractions from sentence-sense. There is
something we must acknowledge and something we must reject in this doctrine.
What we must concede is that when we specify the contribution of words we
specify what they contribute as verbs or predicates or names or whatever, i.e.,
as sentence-parts, to a whole sentence-sense. Neither their status as this or
that part of speech nor the very idea of words having a sense can exist in
isolation from the possibility of words' occurrence in sentences. But this is
not yet to accept that words do not have sense as it were autonomously.
And they must. If our entire understanding of word-sense were derived by
abstraction from the senses of sentences and if (as is obviously the case) we
could only get to know a finite number of sentence senses directly, there
would be an infinite number of different ways of extrapolating to the senses of
sentences whose meanings we have to work out. But we do in fact have an agreed
way of working them out. This is because word-senses are autonomous items, for
which we can write dictionary entries.
Quine
has put the point in dispute so elegantly and concisely that it is
enough to quote him:
'The unit of communication
is the sentence not the word. This point of semantical theory was long obscured
by the undeniable primacy, in one respect, of words. Sentences being limitless
in number and words limited, we necessarily understand most sentences by
construction from antecedently familiar words. Actually there is no conflict
here. We can allow the sentences full monopoly of 'meaning' in some sense,
without denying that the meaning must be worked out. Then we can say that
knowing words is knowing how to work out the meanings of sentences containing
them. Dictionary definitions are mere clauses in a
recursive definition
of the
meanings of sentences.'" [Wiggins (1971), pp.24-25; quoting Quine (1981),
pp.75-76. I owe this reference to Mark Platts; cf., Platts (1997), pp.23-24.
Link added.]
The
above passage was included because of the concise way it expresses the idea
discussed in the main body of this Essay. However, it demonstrates that there is no room here for 'meaning atomism' -- the idea
that we assign meanings to our words atomistically. Even so, I will pass no opinion on whether or not we build our
sentences in the manner indicated. Nevertheless, all of us
make do with a vocabulary that is already part of our social network of
communication, the meaning of the overwhelming majority whose words we had no hand in establishing.
Of course, we must distinguish here between
(a) Words as 'units of meaning' (understood as outlined above) and (b) Words
supposedly learnt atomistically, by each
individual in their own way, 'in their own head'. Expressed concisely in this manner, (a)
isn't
inconsistent with anything written here, whereas (b) is. More on this
below.
Admittedly, there are widespread
disagreements among
Analytic Philosophers about how to interpret the distinction between
word-meaning and sentence-sense, but that doesn't affect the distinction itself.
[On this, see Dever (2008), and
Szabó (2017); the first of these might prove rather challenging to anyone
not familiar with modern logic, the second perhaps much less so.]
Some might object that the present author's
views here clearly form part of a
philosophical theory, which puts them at odds with the
her own stated aims.
That isn't so. The views expressed here are
in fact part of an
elucidatory "form of representation", and as such form no part of a philosophical theory
(on this, see Essay Thirteen Part
Two, when it is published in 2025). That is why I said earlier:
Such an account need not be
general, nor need it be an explanation of language
(except in an elucidatory sense outlined
here) -- it might
simply
describe it perspicuously (and this is so for reasons outlined
Essay Twelve
Part One).
However, on this,
also see the comments I have
added
here.
Back
in 1979, I was privileged to hear the late Michael Dummett read a paper at
Bedford College in London (later re-named Regent's College, which is now
part of
Regent's University) that broke entirely new ground in this area. To the
best of my knowledge his paper hasn't been published -- although, Dummett
(1987) is the closest I know of to the content of that earlier paper. Readers
are directed there for more details. Anyone not familiar with modern logic might
find parts of that paper somewhat daunting.
30.
Clearly, Voloshinov's theory threatens to erode the difference between words and sentences, as
Wittgenstein noted:
"For naming and describing do
not stand on the same level: naming is a preparation for a description. Naming
is so far not a move in the language-game -- any more than putting a piece in
its place on the board is a move in chess. We may say: nothing has so far
been done, when a thing has been named. It has not even got a name except
in the language-game. This is what Frege meant too, when he said that a word had
meaning only as part of a sentence." [Wittgenstein (1958), §49, p.24.
Italic emphases in the original.]
31.
For example, the otherwise excellent book, Language, Sense and Nonsense
[Baker and Hacker (1984a)] is seriously compromised by its authors' adherence to
a sophisticated version of the same sort of confusion that afflicts Voloshinov's
work. Clearly, this isn't the place to enter into why Baker and Hacker rejected
(in all but name) the distinction outlined in the main body of this Essay. Unfortunately, these
two then proceeded to compound this error in Baker and Hacker (1983a), pp.57-79, 145-70. As we will
see, this serious defect can in fact be traced back to certain aspects of
Wittgenstein's work.
It is worth pointing out here that the most
forthright defence of "Occasionalism" (certainly in the 'Wittgensteinian camp')
is to be found in the writings of
Charles Travis;
cf., Travis (1989, 2000, 2006). Travis's work might seem to lend powerful
support for the sort of analysis of language championed by Voloshinov. However,
even if Travis were correct (and there is good reason to suppose he isn't; on
that, see Bridges (2010)), it would be extremely difficult to recruit his
concept of meaning (etc.) to the cause defended by Voloshinov. This isn't just because Travis's
work is extremely difficult to read and almost impossible to comprehend (Jerry
Fodor described Travis's writing style as analogous to "slugs mating" -- Times
Literary Supplement, Number 5127, 2001, p.3), but because his brand of
"Occasionalism" is light years away from that endorsed by Voloshinov and the
comrades mentioned above. Moreover, if Travis were right (again, there is good
reason to suppose he isn't -- see
below), much of contemporary Cognitive Science (including those
aspects Marnie Holborow and John Parrington appear to endorse) can't be
correct. Of course, readers are invited to ascertain these facts for themselves
by consulting Travis's work -- and good luck with that one!
31aQuestions multiply rapidly: Is "theme"
the property of a sign, or an utterance? What is the "entity", or the "object",
mentioned at the
beginning of this passage (re-quoted below)? Is it the point of an utterance (i.e., why it was said, or
what was hoped to be achieved by uttering it)? Or is it
who or what it was directed toward (e.g., the listener perhaps, or the topic of
conversation)? Is "theme"
thus the significance of an
utterance or of this non-descript
'object'?
And, in view of its transient existence, is "theme" a sort of
trope? If each sign has its own "theme", as it seems each utterance, do
"themes" combine and act in consort in some way like instruments
in an orchestra? Is each utterance (or "verbal
performance") a sign, or is it each word?
"Let us agree to call the
entity which becomes the object of a sign the theme of the sign.
Each fully fledged sign has its theme. And so every verbal performance has its
theme….
"The theme of an ideological
sign and the form of an ideological sign are inextricably bound together and are
separable only in the abstract…." [Voloshinov (1973), p.22. Bold emphasis added.
Italic emphasis in the original.]
"A definite and unitary
meaning, a unitary significance, is a property belonging to any utterance as
a whole. Let us call the significance of a whole utterance its
theme. The theme must be unitary, otherwise we would have no basis for
talking about any one utterance. The theme of an utterance is individual and
unreproducible, just as the utterance itself is individual and unreproducible.
The theme is the expression of the concrete, historical situation that
engendered the utterance. The utterance 'What time is it?' has a different
meaning each time it is used, and hence, in accordance with our terminology, has
a different theme, depending on the concrete historical situation ('historical'
here in microscopic dimensions) during which it is enunciated and of
which, in essence, it is a part." [Voloshinov (1973), p. 99. Bold emphases added. Italic
emphasis in the original.]
Here,
"theme" appears to belong to a whole utterance (perhaps as its
occasional meaning, part of that meaning, or its 'significance'),
while at the same time it is also the "entity" that becomes the "object" of a
sign, and an "expression of the concrete, historical
situation that engendered the utterance"!
With the best will in the world, it isn't easy to make much sense of this, and
it is little use searching through the writings of those who look to
Voloshinov for inspiration in this area, since they totally ignore these
difficulties! What is worse, they never even ask questions of this highly
obscure passage.
But,
when we read further, the mystery only deepens, for Voloshinov says:
"...Only active understanding
can grasp theme -- a generative process can be grasped only with the aid of
another generative process." [Ibid., p.102.]
According to this, "theme" isn't:
"...the entity which becomes
the object of a sign...", [Ibid., p.99.]
but a
"generative process". Voloshinov is all over the place!
32.
It must be acknowledged, however, that Parrington did admit the following:
"The second fundamental
feature of Voloshinov's theory of consciousness which defines it as materialist
lies is [in?] the way it sees language as providing the individual consciousness with
a link to the outside world. Thus it is a view of consciousness that is grounded
in reality. The determining role of society is stressed continually throughout
Marxism and the Philosophy of Language. Having said this, what Voloshinov's own
work lacks is the
empirical scientific
evidence that proves such a connection exists. Fortunately, such evidence is
available in the extensive studies on child development carried out by Vygotsky
at around the time Voloshinov was writing. Although Vygotsky approached the
question of consciousness from a very different angle, he came to very similar
conclusions about the role of language in shaping consciousness." [Parrington
(1997), p.133. Bold emphasis added.]
What little evidence Parrington and others
actually offer in support of Voloshinov will be examined presently.
[Vygotsky's 'evidence' will also be reviewed in a later
re-write of this Essay.]
Now, it could be
argued that Voloshinov is simply analysing familiar facts, of which we are all
aware, in order to make
their significance plain. But, what "familiar facts" are these? [Several
possible candidates will be considered as this Essay unfolds.]
Some might complain here that although I
criticise Voloshinov for omitting evidence in support of his theory, the present
author's work (especially this particular Essay) is surely subject to the same
criticism. Why is this a defect of Voloshinov's work while it isn't of Rosa
Lichtenstein's? "What evidence do you offer in support of what you say?" it
could be asked.
Of course, this would be a valid criticism if
I were offering here a scientific theory of language, but since I'm not, it isn't.
Moreover, my arguments largely depend on familiar words (readily
understood by all competent users of the English language), not on technical
terms and obscure concepts lifted from Traditional Thought and a priori
Psychology. Any technical words that have been used in passing are
either being employed in order to expose their defects, or they can be (and
often have been) explained either in ordinary terms or by means of several links to Internet
resources that do this job for me.
It could be objected that this response is
inconsistent with an
earlier claim.
In fact, what was said earlier was this:
The views expressed
here are in fact part of an
elucidatory "form of representation"
(and as such form no part of a philosophical theory...).
So, I am not offering a
scientific, or even a philosophical, theory, merely an intervention in an ongoing debate about
language, on a par with the things I say later about evolution, psychology, or,
in several earlier Essays, about Mathematics and Physics.
However, whenever I do make scientific or factual claims, I invariably
substantiate them with links or references -- Essay Seven
Part One being a prime example of
such.
If anyone thinks differently, they
should
e-mail me, and, if correct, I will apologiseprofusely
and provide the missing evidence, or withdraw the relevant claim. Would that DM-fans did the same!
32a.
It could be countered that scientists do in fact study extremely ephemeral sub-atomic
objects and events all the
time, so why is this a problem for Voloshinov? However, quite apart from the
fact that
we do not yet know what we are supposed to be looking for (this will soon
become apparent as the argument unfolds), the problem is that each "theme",
and each occurrence of it, is
totally unique. This isn't the case
with sub-atomic particles, which can be studied in repeatable experiments.
Indeed, many of these particles are absolutely identical, as we saw
here. [This
topic will also be addressed presently in the main body of the Essay. In addition, see
Note 34, below.]
33.
Except, it is more like a transcendental assertion.
[A
Transcendental Argument is one that seeks to show what must be the case for
something else to be true, or to exist.]
34.
Even if a perfect (or as near perfect as possible) electronic (etc.) copy of
each conversation were recorded for study this would still fail to alter the
deflationary conclusions reached in the main body of this Essay -- indeed, as Voloshinov
himself incautiously admits:
"The utterance 'What time is
it?' has a
different meaning each time it is used, and hence, in accordance with our
terminology, has a different theme, depending on the concrete historical
situation ('historical' here in microscopic dimensions) during which it
is enunciated and of which, in essence, it is a part." [Voloshinov (1973), p.99.
Bold emphases added.]
Unless these "microscopic dimensions" can
also be
captured and reproduced -- and how would that be possible for anything other than a minor
deity? --, "theme" plainly can't be recorded. And, this isn't a conclusion that
the present author has imposed on Voloshinov's work, for he admits this himself:
"[T]heme must be unitary,
otherwise we would have no basis for talking about any one utterance. The theme
of an utterance is
individual andunreproducible, just as the utterance itself is
individual and unreproducible." [Ibid., p.99. Bold emphasis added.]
Hence,
ex hypothesi, "theme" can't be reproduced in an experiment and so can't become the subject of
a scientific
enquiry into the actual circumstances of a single conversation that has taken
place in the
entire course of human history. [On this, see Note 36b.]
In which case, this half-baked concept must have
itself been imposed on reality,
not read from it -- once more, contrary to the claim that
dialecticians never do this.
Naturally, if this were so, it would
cripple scientific research. Indeed, it would make
communication impossible, too.
This
confirms the fact that in any given speech community, human beings do
not proceed as Voloshinov suggests -- and neither do they interpret
each another's speech. They might interpret the words used by a foreign language
speaker, but not a homophonic (same language) interlocutor. This isn't to deny that we sometimes
struggle to comprehend the odd or confusing things people sometimes say (in our
own language), but we don't do so typically.
It
could be argued that written language should be treated in an entirely
different way to the spoken word. However, Voloshinov himself equates the
two:
"Any word used in actual
speech possesses not only theme and meaning in the referential, or content,
sense of these words, but also value judgement: i.e., all referential contents
produced in living speech are said or written in conjunction with a specific
evaluative accent. There is no such thing as a word without
evaluative accent." [Voloshinov (1973), p.103. Italic emphasis in the
original. Bold emphasis added.]
And,
on the same page, Voloshinov quotes a long passage from
Dostoyevsky to make his point clear. In that passage, six artisans use a
certain noun (which is in fact an expletive) in six different ways. Voloshinov then
says:
"All six 'speech
performances' by the artisans are different despite the fact that they all
consisted of one and the same word...." [Ibid., p.104. Quotation marks
altered to conform with the conventions adopted at this site.]
And yet, Voloshinov is here referring to the
written word, in which case, even he sees that there is no intrinsic
difference between the written and the spoken word. [To be sure, they were
quoted words he was analysing,
but they were still written quoted words.]
Moreover, Voloshinov wrote at length in
Chapter Two about the limitations of "abstract objectivism" (making several valid points along the way, many of which could well be directed against Chomsky's
theory), but his main complaint about philologists who study the written
word isn't that
the written word is different from speech, but that:
"The philologist-linguist
tears the monument [by this, Voloshinov means written signs on monuments from
the past or the present, it seems -- see p.72, RL] out of the real domain and
views it as if it were a self-sufficient, isolated entity....
"The dead language the
philologist studies is, of course, and alien language...." [Ibid., p.73.]
This perhaps explains why Voloshinov
went on to say:
"A book, i.e., a verbal
performance in print, is also an element of verbal communication." [Ibid.,
p.95. Italic
emphasis in the original.]
In which case, we may conclude that there is
for Voloshinov no intrinsic difference between the spoken and the written word.
The point he wished to make was that active examples of both were still
sociological events or phenomena, which would be misunderstood or distorted if they
were isolated from the stream of life in which they had originally been embedded.
36.
To be sure, Voloshinov says that "theme" is indivisible -- but how does he know?
Scientists used to think atoms were.
Anyway, as we are about to
discover, things aren't quite so simple.
The observation, in the main body of
the Essay -- i.e., "unless we are to suppose that each "theme" is timed to
coincide with the beginning and end of each speech act" --, raises several serious problems
of its own. What, for
example, is the minimum linguistic unit that is capable of
generating (or expressing) a single "theme"? To be sure, Voloshinov appears to
want to pair "themes" with utterances (or "signs") -- which option is unavailable
to him anyway, as will be argued presently in the main body of this Essay --
and single words.
"Let us agree to call the
entity which becomes the object of a sign the theme of the sign. Each
fully fledged sign has its theme. And so every verbal performance has its
theme." [Voloshinov (1973), p.22. Bold emphasis added.]
"A definite and unitary
meaning, a unitary significance, is a property belonging to any utterance as
a whole. Let us call the significance of a whole utterance its theme…."
[Ibid., p.99. Italic emphasis in the original.]
In the first passage, "theme" is the "object
of a sign". In the second, it is the "significance of a whole utterance". As we
will see later, by "sign" Voloshinov means practically anything that can be
counted as one, and that includes single words:
"Thus, side by side with the
natural phenomena, with the equipment of technology, and with articles for
consumption, there exists a special world -- the world of signs.
"Signs are particular,
material things; and, as we have seen, any item of nature, technology, or
consumption can become a sign, acquiring in the process a meaning that goes
beyond its particularity. A sign does not simply exist
as a part of reality -- it reflects and refracts another reality. Therefore it
may distort that reality or be true to it, or it may perceive it from a special
point of view. Every sign is subject to the criteria of ideological evaluation
(i.e., whether it is true, false, correct, fair, good, etc.). The domain of
ideology coincides with the domain of signs. They equate with one another.
Wherever a sign is present, ideology is present, too. Everything ideological
possesses semiotic value." [Ibid., p.10. Bold
emphasis alone added.]
But, even if the second option were either possible,
or indeed available, we would still be no
further forward unless we are told precisely what an "utterance" is in
such a context. Is it a word, a syllable, a
phoneme, a phrase, a clause, a
sentence, a paragraph, an entire speech, a book, a conversation...?
Some may object that Voloshinov is quite
clear that utterances can be any or all of the above. But, this response
simply compounds the problem, for we have already been told that any or all of
these can represent different "themes", and that each may in fact express
several "themes". [On that, see
here.] So, for example, does an
uttered sentence express a single "theme", or is there a different "theme" for
each word? On this, see Note 36a1, below
36a.
It is worth recalling that according to Voloshinov "theme" isn't what is
said, but the "object" of what is said (or, perhaps, the
"entity" that becomes the "object" of a "sign") --
that is,
as far as can be ascertained. [Although he
changes his mind about this almost on
almost every page!] Such "objects" can of course change -- as dialecticians themselves
would be the first to acknowledge.
"Let us agree to call the
entity which becomes the object of a sign the theme of the sign.
Each fully fledged sign has its theme. And so every verbal performance has its
theme….
"The theme of an ideological
sign and the form of an ideological sign are inextricably bound together and are
separable only in the abstract…." [Voloshinov (1973), p.22.
Italic emphasis in the original.]
36a1.
It could be objected that each utterance is a specific token of a certain
sentence type, individuated, or relativised, to a specific time and place. That would allow us to pair-off
"themes" and "utterances", one-one. However, that interpretation is contradicted by other
things Voloshinov says; for example:
"The utterance 'What time is
it?' has a different meaning each time it is used, and hence, in accordance
with our terminology, has a different theme, depending on the concrete
historical situation ('historical' here in microscopic dimensions)
during which it is enunciated and of which, in essence, it is a part." [Ibid.,
p.99. Bold emphasis added.]
But, if the utterance "What time is it?" is a
specific token, and not a type, individuated or relativised to a time and
place, it can't
in fact
be used again. Unique utterances conceived this way can't be used at another
time or place. This shows that Voloshinov understood an utterance to be a
type sentence, capable of being used many times. If so, an utterance can't
be paired with a "theme", after all, since "themes" are unique.
[It is worth recalling that only type
sentences can be used over and over as speakers or writers instantiate its many
tokens. If Voloshinov meant tokens here, since they are occasion-specific, they
couldn't be used again -- unless they were tokens of a certain type.]
However, as we will soon see, in the main body
of this Essay, things aren't
even this straight-forward!
36b.
To be sure, Voloshinov says that the "theme" of an utterance is
"unreproducible", but that doesn't mean that a given "theme" might not be
pairable with any one of the example
types given in the main body of this Essay, since "unreproducible" doesn't necessarily mean "un-pairable"
(or even "inexpressible by"). It could mean (a) "totally unique", or (b) "not now capable of being produced". But, even these are equivocal. For
example, the original
Penny Black
is unreproducible (by definition and by fact), since that particular stamp is no
longer made, but that doesn't mean it can't be paired with other reproducible stamps
or envelopes in the here-and-now. Or, consider the late
Maria
Callas's voice: it is totally unique; no one is able to mimic it exactly.
But, that hasn't stopped it being reproduced on countless records and CDs.
In that case, since, Voloshinov doesn't say
what he meant by "unreproducible", the options presented in the main body of
this Essay can't be ruled out.
37.
Even if we were to appeal to neurological evidence (in the shape of brain scans,
etc.) to try to identify which
'thought' was associated with which "theme", that would be of no
assistance. No matter how sophisticated the science, we would still be forced to
identify the supposedly relevant brain waves, sets of electrical impulses, or
bursts of neurological activity, by reference to the reports that subjects made
of the thoughts they supposedly instantiated at that time, or with which they were
later correlated, etc. Hence, even hard-nosed
scientists will have to individuate 'thoughts' first by the use of everyday criteria
(which was the point of referring to what a subject might report), not by the employment of physical evidence. In that
case, since "thought" would be used to interpret the relevant scientific
evidence, the latter would be parasitic on the former, not the other way round.
That would still be the case even if each subject was told what to think, or was
induced into thinking something indirectly by having to watch a film, or read a
book or a card. Here the use of everyday criteria would be transferred to the
decisions made by the experimenter, not the subject. And even if this were
automated in some way, by the use of computers (etc.), those criteria
would now be those applied by the programmer. It is impossible to edit out of
the picture human input or control at some point (and hence the application of
such criteria), even in the sciences (as was argued in detail in Essay Three
Part Two).
[Also see Note 38.]
It
could be argued that scientists might be able to compile a well-confirmed dictionary
or manual connecting these reports with certain events in the brain so that
they wouldn't have to appeal to such reports in the future. When, for example,
event E1
occurs in the CNS, they could read-off from their manual that the subject involved was
thinking, say, "I'm hungry", and so on. In that case, we could learn
to individuate
thoughts in this way.
But,
even if this were possible, it wouldn't help, for we have already been told
that "theme" is totally unique and unreproducible. If
so, the scientists involved in
drawing-up this manual would be faced with the task of correlating neurological
events with something that is intrinsically unique. And if that is the
case, they wouldn't be able to repeat a single such reading, since, ex hypothesi,
the "theme" it supposedly represented will never occur
again! Hence, if Voloshinov were correct, they wouldn't be able to confirm
the accuracy of this hypothetical manual; nor would they be able to distinguish
any of these supposed
correlations from sheer coincidence.
Indeed, they wouldn't even be able to build
such a manual in the first place; Voloshinov's theory rules that out, too:
"The theme
of an utterance is
individual andunreproducible, just as the utterance itself is
individual and unreproducible." [Voloshinov (1973), p.99. Bold emphasis added.]
Since scientists
also use language,
what they record is no less subject to the very same equivocations as the
utterances and thoughts of their subjects.
[On this, see Note 38.] Down that road I fear lies another infinite
regress...
Independently of the above, even this aspect of neuroscience is
contested territory. On the correlations that neurophysiologists claim to have found,
and what critics have had to say about them, see
Note
53,and here.
38.
This is all quite independent of the fact that physical properties like these
would only be identifiable by the "theme" (or the utterance) they supposedly
accompanied. In order for observers to be able to say where one set of
identifying physical properties ends and another begins, some sort of reference
would have to be made to the "theme" that they instantiated,
otherwise any number of minor physical changes might be inadvertently roped in.
In that case, once more, "theme" would have to be individuated first
so that it could be individuated second! Hence, in order to pair-off a "thought"
with a "theme", so that the latter might be individuated, "theme" would have to
be individuated first! There is no way out of this circle.
Moreover, this process (or at least its investigation) would be parasitic on the
ordinary (reproducible) meanings of any of the words used,
which would, of course, render the whole exercise pointless. [On this, see Note 39, below.]
Anyway, as we
have seen, an
ascent into formal, technical or mathematical language in order to help solve, or eradicate,
any of the associated philosophical 'problems' would be no less futile. That is primarily because
they depend on
ordinary criteria for their construction and expression -- i.e., they depend on those
involved being able to recognise symbols, reproduce them at will, write them down
accurately,
double check what has been written down, explain themselves to others (if required), etc., etc.,
-- if
they are to be rendered comprehensible by anything recognisably human.
Moreover, such technical 'fixes' invariably re-duplicate the 'problems' they had
originally been introduced to solve. In this case, of course, among other
things, the reproducibility factor here renders formal devices useless in the
study of "theme", which is why the phrase was italicised.
"Theme" is "unreproducible", so we are told.
[On the
futility of appealing to formal devices to solve philosophical 'problems', see Essay
Five and
Essay Twelve Part One. See also,
Hacker (2007b).]
39.
Similar difficulties would also bedevil any attempt to explain what might have
been meant (or "themed") by any of the words a speaker uttered -- i.e., whether
or not that attempt itself was aimed at interpreting what had just been said by
someone else, or whether it formed part of a study of sets of words linked to
context-types in general -- including, if you will, all that is written in
Holborow's book, as well as in Parrington and
Doherty's articles.
In short, the transient nature of "theme"
will unavoidably scupper any attempt made by anyone to explain any other "theme",
since everything the above comrades might have said or written on this topic,
for example, would possess its own
ephemeral and
unreproducible "theme", and so on, ad infinitem.
[To be sure, some might think that Voloshinov
thought to
distinguish the spoken from the written word, but that 'difficulty' has already been
covered.]
In which case, this would put an end
to the scientific study of "theme", for, as noted above, not only would the
latter be forever inaccessible, the writings of those studying 'it' would also be
subject to the very same constraints; their "themes" would be forever
inaccessible, too. And, if Voloshinov is to be believed, without access to its "theme", a body of text or an
utterance would be incomprehensible:
"…The theme must be unitary,
otherwise we would have no basis for talking about any one utterance. The theme
of an utterance is individual and unreproducible, just as the utterance itself
is individual and unreproducible. The theme is the expression of the concrete,
historical situation that engendered the utterance. The utterance 'What time is
it?' has a different meaning each time it is used, and hence, in accordance with
our terminology, has a different theme, depending on the concrete historical
situation ('historical' here in microscopic dimensions) during which it is
enunciated and of which, in essence, it is a part.
"It follows, then, that the
theme of an utterance is determined not only by the linguistic forms that
comprise it -- words, morphological and syntactic structures, sounds, and
intonation -- but also by extraverbal factors of the situation. Should we miss
these situational factors, we would be as little able to understand an utterance
as if we were to miss its most important words. The theme of an utterance is
concrete -- as concrete as the historical instant to which the utterance belongs.
Only an utterance taken in its full, concrete scope as an historical
phenomenon possesses a theme. That is what is meant by the theme of an
utterance." [Voloshinov (1973), pp.99-100. Bold emphases alone added.]
But, that would render Voloshinov's book
incomprehensible since no one (now) has access to its "theme", even if it once
had one!
40.
Any attempt to play the usual 'get-out-of-jail-free' card here (along the lines
that dialectics is meant to be contradictory, so no wonder Voloshinov
said inconsistent things -- or, perhaps even: these criticisms ignore the "identity-in-difference"
[IED] between "theme" and meaning) would be to no avail, either. The first tactic
was called the "Nixon Defence"
in Essay Eight Part One (the reader is referred back to that discussion for more
details),
and the second was neutralised in
Essay Six.
However, with respect to the second line of
defence (the IED-ploy), while there might indeed be a difference between meaning
and "theme", since we are still unclear what "theme" is,
it is far from
obvious what the alleged identity relation is supposed to operate between!
In fact, meaning seems to share
none of the characteristics of "theme", so these two can't be identical.
[More on this
later.]
41.
Of course, two can play at that game. In view of all the difficulties,
confusions and equivocations we have encountered in relation to Voloshinov's ideas, one is tempted to respond by
pointing that those who wish to employ "theme" to explain
understanding in effect want to switch on a light after having
smashed the light bulb, destroyed the switch, ripped out the wiring, cut the
mains, razed the power station, bulldozed the wind farm, flooded the coal mine,
demolished the dam, set fire to the oil
refinery, blown up the gas field and extinguished the Sun.
41a0.
An appeal to approximate understanding must fail, too. If E1 were correct, we
wouldn't even understand the phrase "approximate understanding"!
E1: Well, we never really
understand one another, do we?
41a.
In fact, as we have already seen,
Voloshinov is torn between a Marxist interpretation of language as a
sociological and communicational phenomenon, and the bourgeois individualist,
traditional view that language is
individual and primarily representational. So, in
places he says things like this:
"Signs can arise only on
interindividual territory. It is territory that cannot be called 'natural'
in the direct sense of the word [Added in a footnote: Society, of course, is
also a part of nature, but a part that is qualitatively separate and
distinct and possesses its own specific systems of laws.]: signs do not
arise between any two members of the species
Homo sapiens.
It is
essential that the two individuals be organised socially, that they
compose a group (a social unit); only then can the medium of signs take place
between them.The individual consciousness not only cannot be used to explain
anything, but, on the contrary, is itself in need of explanation from the
vantage point of the social, ideological medium.
"The individual
consciousness is a social-ideological fact. Not until this point is
recognised with due provision for all the consequences that follow from it will
it be possible to construct either an objective psychology or an objective study
of ideologies....
"The only possible objective
definition of consciousness is a sociological one. Consciousness cannot be
derived directly from nature, as has been and still is being attempted by naive
mechanistic materialism and contemporary objective psychology (of the
biological, behaviouristic, and
reflexological
varieties). Ideology cannot be derived
from consciousness, as is the practice of idealism and psychologistic
positivism.
Consciousness takes shape and being in the material of signs created by an
organised group in the process of its social intercourse. The individual
consciousness is nurtured on signs; it derives its growth from them; it reflects
their logic and laws. The logic of consciousness is the logic of ideological
communication, of the semiotic interaction of a social group....
"All that has been said above
leads to the following methodological conclusions: the study of ideologies
does not depend on psychology to any extent and need not be grounded in it.
As we shall see in greater detail in a later chapter, it is rather the reverse:
objective psychology must be grounded in the study of ideologies. The
reality of ideological phenomena is the objective reality of social signs....
Individual consciousness is not the architect of the ideological superstructure,
but only a tenant lodging in the social edifice of ideological signs."
[Voloshinov (1973), pp.12-13. Bold
emphases alone added. Spelling
modified to conform with UK English.]
"Thus every sign...is
social." [Ibid., p.34.]
"[T]he sign and its social
situation are inextricably fused together. The sign cannot be separated from
the social situation without relinquishing its nature as a sign." [Ibid., p.37. Italic
emphasis in the original.]
In the above passages the social takes
precedence. However, on other occasions he says things like the following:
"Any ideological product is
not only itself a part of reality (natural or social), just as is any physical
body, any instrument of production, or any product of consumption, it also, in
contradistinction to these other phenomena, reflects and refracts another
reality outside itself. Everything ideological possesses meaning: it
represents, depicts or stands for something lying outside itself. In other words
it is a sign. Without signs there is no ideology. A physical body
equals itself, so to speak; it does not signify anything but wholly coincides
with its particular, given nature. In this case there is no question of
ideology.
"However, any physical body
may be perceived as an image; for instance, the image of natural inertia and
necessity embodied in that particular thing. Any artistic-symbolic image to
which a particular physical object gives rise is already an ideological product.
The physical object is converted into a sign. Without ceasing to be a part of
material reality, such an object, to some degree, reflects and refracts another
reality.
"A sign does not simply exist
as a part of reality -- it reflects and refracts another reality. Therefore it
may distort that reality or be true to it, or it may perceive it from a special
point of view…." [Ibid.,
pp.9-10.]
"Idealism and psychologism
alike overlook the fact that understanding itself can come about only within
some kind of semiotic material (e.g., inner speech), that sign bears upon sign,
that consciousness itself can arise and become a viable fact only in the
material embodiment of signs...understanding is a response to a sign with
signs." [Ibid., p.11.]
"After all, meaning can
belong only to a sign; meaning outside a sign is a fiction. Meaning is the
expression of a semiotic relation between a particular piece of reality and
another kind of reality that is stands for, represents or depicts....
"...[T]he semiotic material
of the psyche is pre-eminently the word -- inner speech." [Ibid.,
pp.28-29.]
"In the first instance,
to
understand means to refer a particular inner sign to a unity consisting of
other inner signs, to perceive it in the context of a particular psyche....
"Self-observation
(introspection) is the understanding of one's own inner sign.... We do not see or
feel an experience we understand it. This means that in the process of
introspection we engage our experience into a context made up of other signs we
understand. A sign can only be illuminated with the help of another sign."
[Ibid., pp.35-36.]
"The ideological sign is
made viable by its psychic implementation just as much as psychic implementation
is made viable by its ideological implementation. Psychic experience is
something inner that becomes outer and the ideological sign, something outer
that becomes inner. The psyche enjoys extraterritorial status in the organism.
It is a social entity that penetrates inside the organism of the
individual person. Everything ideological is likewise extraterritorial in the
socioeconomic sphere, since the ideological sign, whose locus is outside the
organism, must enter the inner world in order to implement its meaning as
sign." [Ibid., p.39.]
"Any genuine kind of
understanding will be active and will constitute the germ of a response. Only
active understanding can grasp theme -- a generative process can be grasped only
with the aid of another generative process.
"To understand another
person's utterance means to orient oneself with respect to it, to find the
proper place for it in the corresponding context. For each word of the utterance
that we are in process of understanding, we, as it were, lay down a set of our
own answering words. The greater their number and weight, the deeper and more
substantial our understanding will be.
"Thus each of the
distinguishable significative elements of an utterance and the entire utterance
as a whole entity are translated in our minds into another, active and
responsive, context. Any true understanding is dialogic in nature….
Understanding strives to match the speaker's word with a counter word….
[Ibid., p.102.In all of the above, bold emphases alone added.]
Here, signs are representational, and the
language of meaning and understanding is internal and individual. The
social has been lost.
At this point, Voloshinov's theory is (in
some respects) reminiscent of Leibniz's 'windowless
monads'. The latter were supposed to be hermetically sealed off from one
another and, by implication, the rest of the universe. They had their entire
nature and history written into them by 'god' -- along with all their apparent
interactions and relations with other monads ("apparent" since, as noted above, they
actually possessed none of these interactions and relations, but only acted as
if they did). So, they were fundamentally isolated and
atomic in nature (in a logico-metaphysical sort of sense), while being paradoxically social (since
that was written into their contracts by 'god', so to speak, supposedly creating
a Harmonious Whole, "the best of all possible worlds"). So, even
though they reflected
(logically, not physically) the
rest of the universe, they were in fact completely isolated from it. Voloshinov's theory suffers from an
analogous form of philosophical schizophrenia -- he clearly wants
his theory to be social, but the way he spells out the details means it is fundamentally a-social.
It might be argued that these two views don't in fact clash, they
complement each other. On that, see Note 23, above.
It could now be objected that this
interpretation of Voloshinov is incorrect,
as Parrington points out:
"The psyche is thus not
an internal but a boundary phenomenon. Or as Voloshinov put it, 'Individual
consciousness is not the architect of the ideological
superstructure, but only a tenant lodging in the social edifice of ideological
signs'." [Parrington
(1997), pp.129-30, quoting Voloshinov (1973), p.13, and not p.39 as
Parrington has it.]
In which case, the individual psyche isn't an "internal"
phenomenon,
contrary to the argument presented above, it is a "boundary phenomenon".
Or so it could be argued.
And yet, despite what Parrington asserts, I
been unable to find anywhere in Voloshinov's book where he says the
psyche isn't an internal
phenomenon (and Parrington nowhere quotes Voloshinov to the effect that it is a
"boundary phenomenon"). Indeed, and quite the contrary, Voloshinov says things like this:
"Idealism and psychologism
alike overlook the fact that understanding itself can come about only within
some kind of semiotic material (e.g., inner speech), that sign bears upon sign,
that consciousness itself can arise and become a viable fact only in the
material embodiment of signs...understanding is a response to a sign with
signs." [Voloshinov
(1973), p.11.]
"...[T]he semiotic material
of the psyche is pre-eminently the word -- inner speech." [Ibid.,
pp.28-29.]
"In the first instance,
to
understand means to refer a particular inner sign to a unity consisting of
other inner signs, to perceive it in the context of a particular psyche....
"Self-observation
(introspection) is the understanding of one's own inner sign.... We do not see or
feel an experience we understand it. This means that in the process of
introspection we engage our experience into a context made up of other signs we
understand. A sign can only be illuminated with the help of another sign."
[Ibid., pp.35-36.]
"The ideological sign is
made viable by its psychic implementation just as much as psychic implementation
is made viable by its ideological implementation. Psychic experience is
something inner that becomes outer and the ideological sign, something outer
that becomes inner. The psyche enjoys extraterritorial status in the organism.
It is a social entity that penetrates inside the organism of the
individual person. Everything ideological is likewise extraterritorial in the
socioeconomic sphere, since the ideological sign, whose locus is outside the
organism, must enter the inner world in order to implement its meaning as
sign." [Ibid., p.39.In all of the above, bold emphases
alone added.]
Not much wiggle room
there; the psyche is
"internal" for Voloshinov (even if it can become "outer" --
whatever
that means -- it certainly isn't a "boundary phenomenon").
It could be
countered that Voloshinov
explicitly says the following:
"Psychic experience is
something inner that becomes outer and the ideological sign, something outer
that becomes inner. The psyche enjoys extraterritorial status in the organism.
It is a social entity that penetrates inside the organism of the
individual person. Everything ideological is likewise extraterritorial in the
socioeconomic sphere, since the ideological sign, whose locus is outside the
organism, must enter the inner world in order to implement its meaning as
sign." [Ibid., p.39. Italic emphases in the original; bold emphases
added.]
Even so, while the above says that the psyche:
"enjoys extraterritorial
status in the organism. It is a social entity that penetrates inside the organism of the
individual person...",
and:
"Psychic experience is
something inner that becomes outer and the ideological sign, something outer
that becomes inner",
Voloshinov still begins with the psyche as
an internal phenomenon, whatever it might later become. While the psyche deals with sociological signs, the
latter only become comprehensible when they
have been processed internally by the individual concerned.
[On this, see Note 23and Note 41b. In fact,
as we discovered
earlier, Voloshinov's theory of signs and "inner
speech" prevents anyone from understanding either or both!]
Again this would be like saying that monetary
value is a social phenomenon, but then immediately qualifying this by arguing that it is
really up to each individual to assign their own value to money. The former is completely
undermined by the latter.
41b.
This is a consequence of Voloshinov's belief that understanding is in fact a
relation between inner signs:
"Idealism and psychologism
alike overlook the fact that understanding itself can come about only within
some kind of semiotic material (e.g., inner speech), that sign bears upon sign,
that consciousness itself can arise and become a viable fact only in the
material embodiment of signs...understanding is a response to a sign with
signs." [Voloshinov (1973), p.11. Italic emphasis in the original.
Bold emphasis added.]
"In the first instance, to
understand means to refer a particular inner sign to a unity consisting of
other inner signs, to perceive it in the context of a particular psyche....
"Self-observation
(introspection) is the understanding of one's own inner sign.... We do not see or
feel an experience we understand it. This means that in the process of
introspection we engage our experience into a context made up of other signs we
understand. A sign can only be illuminated with the help of another sign."
[Ibid., pp.35-36. Italic emphases in the original. Bold emphases added.]
How
the individual "psyche" manages to do all this is left entirely mysterious. All
we have before us are dead signs that
are somehow related to one another. As Wittgenstein noted, what gives them life is their
social
use by human beings, not their alleged 'inner processing'. [On this, see Note 23, and Stroud (1991, 1996).]
42.
Indeed, when we are attending to
something,
extraneous sounds often impair any attempt to do so. Moreover, a
source that we could neither turn down nor switch off, and which is present all the time
we were awake would seriously impair our capacity to attend to anything at all,
as well as our ability to communicate or learn anything.
It could be objected that this totally
misconstrues the nature of "inner speech". But, how do we know? We don't even
know what "inner speech" is yet, let alone if it exists!
We have
already seen that this ancient idea goes
back at least as far as Plato; in its more modern form, it received its most
significant boost in the work of Vygotsky.
However, a major recent study of
the latter's work [i.e., Frawley (1997)] doesn't even so much as itemise the
(supposed) evidence that confirms the existence of "inner speech", while another
advocate of this 'concept' [i.e., Ushakova (1994)] all but admits that
Vygotskians not only can't agree what "inner speech" is, they fail even
to agree
over how it manifests itself! To be sure, other contributors to the same
study as Ushakova [i.e., those found in Lantolf and Appel (1994)] discuss the
empirical research they have devoted to the study of "inner speech" (i.e.,
in relation to second language acquisition), but they all simply assume
this phenomenon exists (even while they all appear to disagree over exactly what
'it' is!), presenting no evidence at all that it does! In this respect, their
'research' resembles that of other psychologists and scientists who waste their
time delving into
Parapsychological phenomena. This 'discipline' bears all the hallmarks
of a pseudo-science.
[I will add a few comments on Sokolov
(1975) to a later re-write of this Essay.]
43.
If
internal proximity were sufficient for understanding to take place
then a post box, for example, would be able to understand the mail it holds --
and a packet of three the condoms it contains. [Recall that inner processes are
no less physical than letters or condoms.]
Of course, with words there is an
obvious difference: they have meaning. But, internal proximity alone
can't lend
to words a meaning which it denies to postal items or diverse prophylactic products. [On
this, see Note 23.] Anyway, we
have already seen that Voloshinov argues that anything can be a sign:
"Signs are particular,
material things; and, as we have seen,any item of nature, technology, or
consumption can become a sign, acquiring in the process a meaning that goes
beyond its particularity. A sign does not simply exist
as a part of reality -- it reflects and refracts another reality. Therefore it
may distort that reality or be true to it, or it may perceive it from a special
point of view. Every sign is subject to the criteria of ideological evaluation
(i.e., whether it is true, false, correct, fair, good, etc.). The domain of
ideology coincides with the domain of signs. They equate with one another.
Wherever a sign is present, ideology is present, too. Everything ideological
possesses semiotic value." [Voloshinov (1973), p.10. Bold
emphasis alone added.]
That
can only mean that items of mail and prophylactic products must be 'signs', too.
Moreover, Voloshinov himself
acknowledges that internal proximity can't lend to words a meaning:
"Meaning does not reside
in the word or in the soul of the speaker or in the soul of the listener.
Meaning is the effect of interaction between speaker and listener produced
via the material of a particular sound complex." [Ibid., pp.102-03. Bold
emphasis alone added.]
To be sure, there is a further difference
here:
letters and condoms aren't
processes. Nevertheless, the only point being made here is that 'inner
proximity' isn't sufficient for an object to have meaning. Nor is the fact
that something is an 'inner process'. [I would go further, and claim that it
isn't even necessary.
On occasion, Voloshinov seems to agree with this; see the quotation above.] The digestive system, for instance, works 24 hours a day, involving
inter alia
many different processes, but that doesn't imply that the contents of the stomach have a meaning,
or that a layer of
gastric
epithelia understands what a recently consumed kebab is trying to tell it.
It could be countered that the processes in
the CNS are far more
complex, and of a totally different order to those that take place in the
digestive tract. Maybe so, but the mere fact that they are 'inner' doesn't
imply that the processes in the CNS have 'meaning', which is all that
this Note seeks to establish.
[On
Wittgenstein's point that signs and images can't lend to words their meaning since they, too, are dead, see, for example,
Stroud (1991, 1996), and here.]
Complexity doesn't
affect this point,
either, for no natural phenomena can lend meaning to our words -- again, as Voloshinov himself
recognised. [On
that, see Note 23, once more.]
43a.
At this juncture, it is worth drawing the reader's attention (once again) to the
ancient (quasi-mystical) picture that underlies the idea that "inner speech" is
immediate to 'consciousness', while its 'outer' twin isn't. [On that, see Note
1.]
43b.
It could be argued that these processes take place in the subconscious/unconscious, or they
are processes about which we aren't actually aware.
In fact, all that that
response succeeds in achieving is the replacement of the 'little-man-in-the-head'
sort of scenario with a 'little invisible/hidden/secret/silent-man-in-the-head'
sort of scenario. This is just the 'ghost
in the machine' again -- and with a gag order.
To be sure, Cognitive Scientists in general appeal to hidden processes and structures in the
CNS to account for
'thought' or language, even while they object to their theories being
characterised in this way -- that is, they take great exception to the claim
that they surreptitiously appeal to an homunculus to account for our
capacity to think or talk, etc. I will say no more about this here since I
discuss this topic in more detail
here, and in Note 44,
Note 47, and Note 86, and
throughout much of the rest of this Essay.
43c.
This is an adaptation of and argument found in Harman (1967), neatly summarised
for us by
Michael Devitt:
"If a speaker's
competence in a language consists in having
knowledge-that of its
rules, then assuming RTM [The
Representational Theory of Mind -- RL], she must
represent those rules. That representation must itself be in a language. What is
it to be competent in that more basic language? If we suppose that the more
basic language is the same as the original language then we are caught in a
vicious circle. If we suppose that it is some other language ('Mentalese'
perhaps), then its rules also have to be represented. This requires a still more
basic language. And so on. The only way to avoid a vicious circle or an infinite
regress is to allow that we can be competent in at least one language directly,
without representing its rules. Why not then allow this of the original
language, the one spoken?" [Devitt (2006), p.92.]
43d.
The
difficulty some might still have in accepting this contention might be connected with a
failure to see that understanding is a publicly performed, checked and monitored activity, not a privatised
or atomised 'inner skill'. [On that, see the references listed in Note 86.]
It
could be objected that we don't always understand one another. Indeed, but if
Voloshinov were correct, we couldn't understand one another. [I have
responded to that objection,
here.]
44.
As noted above, representationalists have a tendency to deny one or all of these allegations,
often pouring scorn on the accusation that their theory relies on just such an
homunculus, or that elements of cognition are 'represented' somewhere in the
CNS.
[On the disingenuous tactics
adopted by
representationalists, see Devitt (2006). See also
Note 47, and the references listed in
Note 86, below. In fact, Devitt (2006) contains a sophisticated demolition of
this entire way of looking at language and mind. On this, see also Note 43c, above.]
It could be objected
that the interpretation aired here appears to suggest that dialecticians view
understanding (etc.) as a passive sort of activity (hinted at by the use
of the sedentary couch potato analogy in the main body of this Essay, etc.),
when they manifestly don't do this. In fact, they stress the active,
interactive, practical
side of understanding.
Nevertheless, even if this 'internal spectator' were presented as active,
it would merely
resemble an inner keep fit fanatic. Either way,
on this view, meaning would still fail to be social, but private, based
on a hidden, secret, and individualised skill.
Indeed, "practice" in such a scenario would be confined to whatever takes place inside
each cranium!
It
could be argued that this is a gross caricature of the dialectical connection
between practice and cognition. But, that isn't so, and for reasons outlined in
Essay Three Part Two. Readers are
directed there for more details.
It
could now be objected that this misrepresents Voloshinov's theory, since it is
quite clear that he believes meaning emerges in the relationship between speakers,
and that this is a sociological phenomenon. Perhaps so, but the other things he
says (outlined in in Note 23 and
Note 41a) carry the opposite
implication.
45.
This inner individual only appears to be 'human' because of the
fetishisation of
language upon which this fable has traditionally been predicated. [Again, see
Note 23 on this.] There is in fact no way this 'inner
being' is human in any recognisable sense of that word, since it enjoys
no social life, and yetit magically possesses powers of comprehension way
beyond the relatively feeble abilities we possess -- since, according to
Voloshinov, we can't comprehend anything without it!
On this aspect of
traditional theory, see Hacker (1993g, 2013), and
Kenny (1984b).
Of course, given the scenario depicted by,
for example, Daniel Dennett (outlined by Alex Callinicos in Note 54,
below), our heads would
appear to be populated by countless rather dim 'human' surrogates. It is to be
hoped they have all been unionised for their own good! Such dimwits can't be
relied upon to defend, let alone recognise, their own interests.
[More on this in Note 2, above, and Note 46, below.]
45a.
An excellent example of the 'homunculus fallacy' is to be found in Kurzweil
(2013). As Colin McGinn points out:
"There is another glaring problem with
Kurzweil's
book: the relentless and unapologetic use of homunculus language. Kurzweil
writes: 'The firing of the axon is that pattern recognizer shouting the name of
the pattern: "Hey guys, I just saw the written word 'apple'.'"' Again:
'If, for example, we are reading from left to
right and have already seen and recognized the letters "A," "P," "P," and "L,"
the "APPLE" recognizer will predict that it is likely to see an "E" in the next
position. It will send a signal down to the "E" recognizer saying, in effect,
"Please be aware that there is a high likelihood that you will see your 'E'
pattern very soon, so be on the lookout for it." The "E" recognizer then adjusts
its threshold such that it is more likely to recognize an "E."'
"Presumably (I am not entirely sure) Kurzweil would agree that such descriptions
cannot be taken literally: individual neurons don't say things or predict things
or see things -- though it is perhaps as if they do. People say and predict and
see, not little bunches of neurons, still less bits of machines. Such
anthropomorphic descriptions of cortical activity must ultimately be replaced by
literal descriptions of electric charge and chemical transmission (though they
may be harmless for expository purposes). Still, they are not scientifically
acceptable as they stand.
"But the problem bites deeper than that, for two reasons. First, homunculus talk
can give rise to the illusion that one is nearer to accounting for the mind,
properly so-called, than one really is. If neural clumps can be characterized in
psychological terms, then it looks as if we are in the right conceptual ballpark
when trying to explain genuine mental phenomena -- such as the recognition of
words and faces by perceiving conscious subjects. But if we strip our
theoretical language of psychological content, restricting ourselves to the
physics and chemistry of cells, we are far from accounting for the mental
phenomena we wish to explain. An army of homunculi all recognizing patterns,
talking to each other, and having expectations might provide a foundation for
whole-person pattern recognition; but electrochemical interactions across cell
membranes are a far cry from actually consciously seeing something as the letter
'A.' How do we get from pure chemistry to full-blown psychology?
"And the second point is that even talk of 'pattern recognition' by neurons is
already far too homunculus-like for comfort: people (and animals) recognize
patterns -- neurons don't. Neurons simply emit electrical impulses when caused
to do so by impinging stimuli; they don't recognize anything in the
literal sense. Recognizing is a conscious mental act. Neither do neurons read
or understand -- though they may be said to simulate these mental acts.
"Here I must say something briefly about the standard language that neuroscience
has come to assume in the last fifty or so years (the subject deserves extended
treatment. [McGinn ignores the fact that Bennett and Hacker have already done
this (see the reference below) -- RL.] Even in sober neuroscience textbooks we
are routinely told that bits of the brain 'process information,' 'send signals,'
and 'receive messages' -- as if this were as uncontroversial as electrical and
chemical processes occurring in the brain. We need to scrutinize such talk
with care. Why exactly is it thought that the brain can be described in these
ways? It is a collection of biological cells like any bodily organ, much like
the liver or the heart, which are not apt to be described in informational
terms. It can hardly be claimed that we have observed information
transmission in the brain, as we have observed certain chemicals; this is a
purely theoretical description of what is going on. So what is the basis for the
theory?
"The answer must surely be that the brain is causally connected to the mind and
the mind contains and processes information. That is, a conscious subject
has knowledge, memory, perception, and the power of reason -- I have
various kinds of information at my disposal. No doubt I have this information
because of activity in my brain, but it doesn't follow that my brain also has
such information, still less microscopic bits of it. Why do we say that
telephone lines convey information? Not because they are intrinsically
informational, but because conscious subjects are at either end of them,
exchanging information in the ordinary sense. Without the conscious subjects and
their informational states, wires and neurons would not warrant being described
in informational terms.
"The mistake is to suppose that wires and neurons are homunculi that somehow
mimic human subjects in their information-processing powers; instead they are
simply the causal background to genuinely informational transactions. The brain
considered in itself, independently of the mind, does not process information or
send signals or receive messages, any more than the heart does; people do, and
the brain is the underlying mechanism that enables them to do so. It is simply
false to say that one neuron literally 'sends a signal' to another; what it does
is engage in certain chemical and electrical activities that are causally
connected to genuine informational activities.
"Contemporary
brain science is thus rife with unwarranted homunculus talk, presented as if it
were sober established science. We have discovered that nerve fibres
transmit electricity. We have not, in the same way, discovered that they
transmit information. We have simply postulated this conclusion by falsely
modelling neurons on persons. To put the point a little more formally: states of
neurons do not have propositional content in the way states of mind have
propositional content. The belief that London is rainy intrinsically and
literally contains the propositional content that London is rainy, but no state
of neurons contains that content in that way -- as opposed to metaphorically or
derivatively (this kind of point has been forcibly urged by
John Searle
for a long time).
"And there is theoretical danger in such loose talk, because it fosters the
illusion that we understand how the brain can give rise to the mind. One of the
central attributes of mind is information (propositional content) and there is a
difficult question about how informational states can come to exist in physical
organisms. We are deluded if we think we can make progress on this question by
attributing informational states to the brain. To be sure, if the brain
were to process information, in the full-blooded sense, then it would be
apt for producing states like belief; but it is simply not literally true that
it processes information. We are accordingly left wondering how electrochemical
activity can give rise to genuine informational states like knowledge, memory,
and perception. As so often, surreptitious homunculus talk generates an
illusion of theoretical understanding." [McGinn (2013), quoted from
here. Italic emphases in the original, bold emphases and links added;
quotation marks altered to conform with the conventions adopted at this site.
Spelling modified to UK English.]
Although McGinn is correct in what he says about "homunculus
talk", he, too, remains locked into the
Cartesian Paradigm, since he
tells us that "the brain is causally connected to the mind and the mind contains
and processes information". He also talks about replacing anthropomorphic
language with "literal descriptions of electric charge and chemical
transmission", and that neurons can "simulate" some of our psychological
capacities, even though he then goes on to say:
"The mistake is to suppose
that wires and neurons are homunculi that somehow mimic human subjects in their
information-processing powers...." [Ibid.]
But, McGinn's mistake here is that he
assumes such language can be replaced with "literal descriptions", just as
he seems oblivious of the fact that this entire way of talking is based on
conceptual confusion and linguistic distortion -- as this Essay seeks to
show.
For an exhaustive corrective to this way of
conceptualising our psychological lives, see Bennett and Hacker (2008, 2021), Bennett, et al (2007),
and Hacker (2013a). See also
Kenny (1984b).
46.
How on earth this homunculus could possibly 'represent' or 'explain'
sounds, smells, colours and feelings either to us or to itself, before it
turned to us, we will
have to pass over in silence. In fact, this picture would seem to fall foul of the 'Harman
Objection', summarised in Note 43c, only now modified along
the following lines:
"If a speaker's competence in
a language consists in having a set of diminutive homunculi do all the
cognitive work for her, then assuming RTM [The
Representational Theory of Mind
-- RL],
each of these diminutive human beings must represent various items to
themselves, too. These representations must be in a more basic language. If we
suppose that the latter is the same as the original language then we are caught
in a
vicious circle. On the other hand, if we suppose that it is some other
language ('Mentalese'
perhaps), then its rules will also have to be represented. But, this requires a
still more basic language, and even more basic homunculi, and so on. The
only way to avoid a vicious circle or infinite regress is to allow that we can
be competent in at least one language directly, without representing its rules,
and without these inner 'friends' to do the work for us. Why not then allow this
of the original language, the one spoken?" [Devitt (2006), p.92,
modified.]
Nevertheless,
Daniel
Dennett reckons he has a reply to this, which will be examined in
Note 54, below.
47.
The existence of 'inner viewers' like this seems to be implied by all forms of
Cognitive Theory --, unless, that is, its specialised vocabulary is to be interpreted
metaphorically. But, even then, what is the 'cash value' of these
metaphors (to use
William
James's happy phrase). For example, if someone describes a man as a pig, the
'cash value' of this simple metaphor might be that the said individual is
uncouth, has disgusting habits, or that he treats women with no respect,
serially cheating on them, and so on. That being the case: what is the 'cash value' of these
'representational' metaphors? It isn't easy to say.
Otherwise, this is indeed the
Cartesian Paradigm, which still dominates
Cognitive Theory, as noted
earlier.
There is in fact a very useful summary of
ancient, early modern and modern views in this area in Sokolov (1975), pp.11-33. A
recent example of how the Cartesian Paradigm still dominates modern theory in this
area can be
found in Hurford (2007).
As we will discover later, in relation to the evolution
of human 'consciousness' and language, this Paradigm overshadows the entire
discipline.
In view of the fact that those who employ these metaphors seem
to intend them to be taken literally, it is rather ironic that in their endeavour to
construct an 'objective' view of reality -- in this case, one such pertaining to our
cognitive capacities --, theorists find they have to introduce an 'ideal
inner observer' (surely the analogue of the Ideal Viewer required to account for the
'objectivity' of modern Physics) to whom such things must be "represented".
Hence,
on this account, science is only able to produce an 'objective' view of reality
-- 'inner' and 'outer' -- by rendering both thoroughly Ideal!
To paraphrase
Anthony Kenny:
those who dote on this sort of talk (i.e., about "inner speech") seem overly impressed by our ability to talk to
ourselves.
[On
understanding in general, see Baker
and Hacker (2005a), pp.357-85, and Baker and Hacker (2005b), pp.305-56.]
On Wittgenstein's criticism of
such 'inner signs', see, for example, Note
23 above, as well as Stroud (1991, 1996).
49.
This resembles, somewhat uncomfortably,
the way that the Ancient Greeks named their 'gods' and then developed an entire
mythology to explain the relation between them!
To be sure, DM-theorists might want to
add a rider here that all this inter-relating should be done "dialectically" --
as if that word were itself a magic wand
that can somehow neutralise the
fetishisation upon which it
depends, and which gave it life. [On that, seeNote 23
andNote 40, above.]
50.
This doesn't amount to a criticism of science but of the metaphysical objectification that
all too often accompanies it in ancient and early modern thought -- in the latter
case, and more pointedly, the world-view bequeathed to us by
Descartes.
[On this, see
Read (2008), as well as Note 1. This
'world-view', which is still central to the fetishised interpretations of 'thought',
'language',
'mind' and 'consciousness' covered in the main body of this Essay, has been
critically evaluated in many of the books and articles listed
in Note 86, below.]
It could be argued that it is what these
words refer to that is of interest to scientists. So, for them, for
example, "thought" refers to what goes on in the head when we think. Maybe
they do so argue, but that response is itself subject to the rebuttal posted
here.
It is also worth pointing out
(yet again) that
these comments aren't directed at our ordinary use of terms like "consciousness"
and "mind", but at the philosophical employment of typographically identical words.
On that, see here and
here.
51.
This isn't to suggest that scientists (or even philosophers) don't theorise
about such 'inner processes', only that it is invariably the case that (in
practice) an appeal to material evidence (i.e., evidence that has been
derived from, and is based on, 'external' reality) takes precedence, while these
'inner events', or 'processes', remain irrelevant in this regard. [That
observation will be
explained presently.]
However, as noted above, when it comes to
reifying 'mental processes' (as 'inner events' or 'structures'), scientists
are as prone as anyone else to make the same mistakes when they try to
'philosophise' about the implications of their work, or when they indulge in
their frequent 'popularisations'
of it.
Indeed, whenever historians, for example, attribute aims,
motives or intentions to long dead individuals, they manifestly don't consider
first the (inferred) 'events' in any of their subjects' heads as evidence
for whatever psychological states they might want to attribute to them -- even
if some attempt to infer what these 'events' might be later re-surfaces as part of a speculative
or re-constructive biography of one of more of these subjects, etc., -- unless, that is, there were clear
evidence of some sort of pathology of the brain in the one being studied.
Only then would they consider the hypothetical state of their subjects' brains
to be relevant in the first instance.
On the other hand, if they were to
speculate about such 'inner' states first, before they had examined the
relevant 'external' evidence (in the form of documents, letters, biographies,
artefacts, testimonies, etc.), no one would take them seriously. Their work would
be viewed as fanciful, at best. Historians examine the aforementioned documents, artefacts, testimonies, letters,
etc., first, not last. Only after considerable stage-setting has taken
place would a historian, for instance, presume to speculate about any alleged 'inner
processes' in the CNS or brain of any such character from the past. And that is true, not just of long dead individuals, it
is also true of those alive today. So, if someone wanted to write a biography
of, say,
Richard Dawkins, she wouldn't begin by speculating about what goes
on, or has gone on, in his head first. She would do exactly what historians do: examine
documents, letters, relevant books and articles, consult Dawkins's work, and
conduct interviews with relatives, friends, acquaintances, colleagues -- and
Dawkins himself -- and so on.
Only after considerable stage-setting would a competent biographer even begin to
speculate or conjecture about Dawkins's psychological make-up or motivation.
This shows that "outer criteria" are decisive,
even here -- i.e., when it comes to assessing the character and
psychology of historical, and even contemporary, figures -- and not
just in our everyday dealings with one another. Indeed, this is something
Voloshinov himself sort of half accepts -- see, for example, pp.36-37 of
Voloshinov (1973) -- although, as Note
23 points out, he ruins
this observation with all that talk about "inner speech" and "inner signs".
The above comments apply equally well to the
work of scientists; readers are is referred to
comments I made
about this in Essay Three Part Two. [In the aforementioned Essay, those
comments were advanced in relation
to 'abstraction', but they equally apply to "inner processes" in general.]
51a."But, what about the imaginative, or theoretical, reconstruction of such processes?"
someone might ask. But, even there the motivation to compose an
imaginative, or theoretical, reconstruction can only base itself on (i) the linguistic false moves outlined in the main body of this
Essay and (ii) the aforementioned Platonic-Christian-Cartesian Paradigm
-- and nothing more.
Again,
it could be objected that what goes on
in our heads surely can't be irrelevant. Over the last hundred years or so
neuroscientists have studied the
CNS extensively, and
have mapped and located regions of the brain that control the structures and processes
which the
author of this Essay seems to think are chimerical.
In fact, the existence of the neurological
processes that scientists study hasn't been questioned -- only their
interpretation. What has been called "chimerical" are the occult processes
that have been imported into Philosophy and Science from the Platonic-Christian-Cartesian Paradigm. Several other, but more
targeted, counter-arguments
will be outlined presently.
51b.
This topic leads naturally into a consideration of Wittgenstein's comments about rule-following.
I don't want to become side-tracked here, so the reader is directed to the
following books and papers: Bloor (1997), Floyd (1991), Kripke (1982), Kusch
(2006), Williams (1999) -- but especially Robinson (2003c), and the material posted here.
There
are scores of discussions of this topic on-line. For example,
here,
here and
here (the last of these links to a PDF).
51c. The term "Superscience"
(or "Super-Truth") was
introduced in an earlier Essay,
where it denoted attempts made by Traditional Philosophers to
concoct a
priori solutions to 'problems' they themselves had invented. These
'solutions' (and associated theses) turn out to be 'true' independently of the way the
world happens to be, and, moreover, relate to a hidden world underlying 'appearances', accessible to thought
alone. This approach to 'philosophical knowledge' was called a "Superscience" since it mimics the sciences in that purports to
deliver knowledge about reality (typically expressed in
indicative sentences), but which knowledge surpasses in logical form anything
that the sciences have ever, or could ever, deliver. [In fact, I borrowed
this term from certain Wittgensteinian authors; for example, Hacker (1987). More details
can be found in Essay Twelve Part One.
I have summarised the argument
here.]
52. This
doesn't, of course, mean that nominalisations per se are illegitimate, merely that
these specific nominalisations constitute the
only 'evidence' there is that these spurious 'inner' objects and processes actually
exist. [On this See Note 1.] In
other words, they were the direct creation of 'word-magic' and nothing more.
53.See
also Vul, et al (2009), the interviews
here
and
here, the summary
here, as well as the discussion
here. A short rebuttal of the many criticisms that the Vul paper attracted can be
found here,
with a more recent reply
here. [This links to a PDF.]
Also see
Vul and Kanwisher (2010) -- (this also links to a PDF) --,
and Prinz (2006). There is also much more material available at Edward Vul's
site, here.
There is another paper by
Yarkoni and Braver available
here on such correlations.
Here is a recent
New Scientist report on some of this work:
"Doubts raised over brain scan findings
"Jim Giles
"Some of the hottest results in the nascent field
of social neuroscience, in which emotions and behavioural traits are linked to
activity in a particular region of the brain, may be inflated and in some cases
entirely spurious.
"So say psychologist
Hal Pashler
at the University of California, San Diego, and his colleagues, who examined
more than 50 studies that relied on functional magnetic resonance imaging (fMRI)
brain scans, many published in high-profile journals, and questioned the authors
about their methods.
"Pashler's team say that in most of the studies,
which linked brain regions to feelings including social rejection, neuroticism
and jealousy, researchers interpreted their data using a method that
inflates the strength of the link
between a brain region and the emotion or behaviour.
"The claim is disputed by at least two of the
critiqued groups. Both argue that Pashler has misunderstood their results and
that their conclusions are backed by other studies.
"In many of the studies, researchers scan
volunteers' brains as they complete a task designed to elicit a particular
emotion. They then divide the images from the scans into cubes called voxels,
which can each contain millions of neurons, and attempt to correlate the
activity of particular voxels with emotional changes reported by the volunteers.
"The problem arises when researchers attempt to
calculate the strength of this correlation. This has to be done in two stages.
The first is to identify regions in which the correlation between voxel activity
and the emotion exceeds a certain threshold. In the second stage, the
researchers assess the strength of the correlation in that region.
"Pashler recommends that two independent sets of
scans be used in these two stages. If the same set is used for both, there is an
increased risk of misinterpreting random noise as a genuine signal.
"Yet in almost 30 of the papers Pashler's team
analysed, researchers used the same scans to identify the voxels of interest and
determine the final correlation. This inflates the correlation above its true
value, and has the potential to produce apparent links between emotions and
brain regions when none exists, Pashler's team claims.
"To demonstrate their point, the team used this
technique to search for correlations in simulated brain scan data. This appeared
to reveal statistically significant correlations, when in fact there were none.
The
critique
has been peer-reviewed and accepted for
publication in
Perspectives on Psychological Science.
"Many researchers are not surprised by Pashler's
conclusions, as the more rigorous analysis that his team recommends requires
more data. It is expensive to run fMRI scans and difficult to find volunteers.
"Nikolaus Kriegeskorte at the US National
Institute of Mental Health in Bethesda, Maryland, is trying to gauge the number
of neuroscience papers that use this method, because he also believes it is
problematic. And in 2007, researchers withdrew part of an fMRI paper published
in Nature Neuroscience (DOI:
10.1038/nn0107-1) after another
researcher showed that random noise could have produced the reported
correlation.
"The researchers criticised by Pashler strongly
contest his team's conclusions. Tania Singer at the University of Zurich in
Switzerland says Pashler's survey questions were 'ambiguous and incomplete',
causing him to misjudge the way her group corrected for random noise. Matthew
Lieberman of the University of California, Los Angeles, says that even if his
2003 study in Science (DOI:
10.1126/science.1089134) inflates the
strength of the link between feelings of rejection and physical pain, the link
itself stands, as studies using different methods also identified it." [New
Scientist 2691, 14/01/2009. Links in the original.
Quotation marks altered to conform with the conventions adopted at this site.
See also a
similar article by Alison Abbott in Nature.]
Once
more, my quoting or referencing the above sources doesn't imply I agree with the
unexamined assumption which they all seem to accept: that there are such things
as 'mental processes' in the brain,
or which arise from that organ. In fact, as this Essay shows, the opposite is in
fact the case -- or, rather, the supposition that there are such 'mental
processes' is radically confused.
It
could be objected that we do in fact use the word "mental" to refer to hidden
process in the brain, for example, when we refer to mental arithmetic. If
so, much of the above material is completely misguided. To be sure, this is a legitimate use
of the word, but we have yet to see the proof that this is done 'in the brain',
or that it refers to a 'process in the
CNS'.
It
could now be asked of the author: "Well, what do you suppose is going on in the
brain? Do these structures and processes have nothing to do with
consciousness, thought, emotion, etc.?"
The
problem with this response that it is based on the assumption that there is anything
'mental' there to be correlated with any activity in regions of the brain, or
in the whole brain itself. Since the spurious concepts, and words,
"consciousness" and "thought", as they are used in the Cognitive Sciences (and
the Philosophy of Mind), for example, are the result of the nominalisation and reification of certain verbs, there is about as much here to correlate as there would be between the price of
coffee and the life and loves of a
Harpy.
Despite this, the above
objection might
continue: "But what relation is there between these
neurological or psychological structures and processes and
how we are able to think, or with what we are capable of deciding? Surely, they
must have something to do with our psychological lives and makeup. An empty head
wouldn't be able think at all!"
This problem is
connected with the social nature of language. As will be argued in more detail
in Essay Twelve, since language serves primarily as a means of communication, its capacity to represent
the world isn't too impressive, especially in areas far removed from everyday
life and common understanding. That is
why theorists find they have to construct models, employ metaphors and analogies,
indulge in 'thought experiments', use 'scare' quotes all the time, invent technical terms
and neologisms, rely on a distorted or confused use/misuse of ordinary words,
nominalise and reify verbs to order in their endeavour to 'represent' nature and the 'mind'.
[And, what is
worse, they have
yet to break free from the Platonic-Christian-Cartesian Paradigm.]
When it comes to comprehending 'what goes on in the brain', this predicament is
all the more intractable. That is because (i) The dice have already been loaded
by the almost universal adoption of the aforementioned Paradigm and the use of
cognitive models that haven't advanced
much beyond
bourgeois individualism of the 17th
and 18th centuries, and
because (ii) The social nature of discourse and the words we use to give
expression to our
psychological lives can't be applied to what goes on in the brain
without mis-using them or undermining their social nature (as we
have seen).
The bottom line is
that it isn't up to philosophers to decide about these issues, and it isn't up to me,
either. However, that doesn't mean we have to accede, or give ground, to the
impromptu and amateur Metaphysics that passes itself off as 'Brain Science' or
'Philosophy of Mind' these days -- nor indeed the vastly inferior 'dialectical'
version of either or both.
And, of course, if an
individual literally had an empty head, they couldn't enjoy any sort
of psychological life since they would be dead.
Update October 2013: According to
a
leader article in the New Scientist, Neuroscience seems to be in total disarray:
"The idea of putting a dead salmon in a brain
scanner would be funny if it were not so serious. When Craig Bennett of the
University of California, Santa Barbara, tried it in 2009, he wasn't expecting
to find anything -- he was just doing test runs on the machine. But when he
looked at the data he got a shock. The fish's brain and spinal column were
showing
signs of neural activity.
"There was no such activity, of course. The
salmon was dead. But the signal was there, and it confirmed what many had been
quietly muttering for years: there's something fishy about neuroscience.
"When fMRI brain scanners were invented in the
early 1990s, scientists and the general public were seduced by the idea of
watching the brain at work. It seems we got carried away. The field is plagued
by false positives and other problems. It is now clear that the majority --
perhaps the vast majority -- of neuroscience findings are as spurious as brain
waves in a dead fish (see "Hidden
depths: Brain science is drowning in uncertainty").
"That seems shocking, and not just because
neuroscience has appeared to be one of the most productive research areas of
recent years. Some of those dodgy findings are starting to make their way into
the real world, such as in ongoing debates about the use of
fMRI evidence in court.
"Some historical perspective is helpful here,
however. The problems are not exclusive to neuroscience. In 2005, epidemiologist
John Ioannidis published a bombshell of a paper called 'Why most published
research findings are false'. In it he catalogued a litany of failures that
undermine the reliability of science in general. His analysis concluded that at
least half, and possibly a large majority, of published research is wrong.
"Ioannidis might have expected anger and denial,
but
his paper was well received.
Scientists welcomed the chance to debate the flaws in their practices and work
to put them right.
"Things are by no means perfect now. Scientists
are under immense pressure to make discoveries, so negative findings often go
unreported, experiments are rarely replicated and data is often 'tortured until
it confesses'. But -- thanks in no small part to Ioannidis's brutal honesty --
all of those issues are now out in the open and science is working to address
them. The kerfuffle over neuroscience is just the latest chapter in a
long-running saga." [New
Scientist220, 2939, 18/10/2013, p.3.
Links in the original. Quotation marks altered to conform with the conventions adopted
at this site.]
The Leader, however,
finishes on an up-beat note:
"Genetics went through a similar 'crisis' about a decade ago and has since
matured into one of the most reliable sciences of all. The fact that
neuroscience is facing up to its problems is the sign of a young discipline
growing up. Some of the flashy discoveries about brain areas 'for' love or
religion will go the way of genes 'for' intelligence, or whatever. But
neuroscience will be more nuanced and powerful for it." [Ibid. Quotation
marks altered to conform with the conventions adopted at this site.]
It is difficult to
share the unfounded optimism on display here. As
Wittgenstein pointed out,
and as has been argued in this Essay, Neuroscience has been crippled by profound
conceptual confusion (again, this is a consequence of buying into the
Cartesian Paradigm), unlike Genetics. The main article
mentioned (and linked to) in the Leader gives the game away:
"Amid these concerns, it might seem as if our
understanding of the brain is set to disappear in a fog of uncertainty, and you
will find many observers in the popular press who are now bashing 'neuromania'.
But it's important not to forget the advances of the last century. And while the
tough conclusions of Ioannidis and his colleagues are certainly reason to
reassess our knowledge, their insights should only lead to more fruitful efforts
in uncovering the mind's mysteries." [Ibid.,
p.33. (Unfortunately, this article is only available to subscribers.)
Quotation marks altered to conform with the conventions
adopted at this site. Link in the original.]
Here 'the mind' is
plainly an object of some sort different in name alone from the 'soul' of
Cartesian Metaphysics. No solution is going to be found for this pseudo-problem, since, as the above shows, the entire discipline is still in thrall
to Plato and Descartes. These hardy "ruling ideas" refuse to
wither and die.
Lenin committed himself to the following
(Cartesian) view of 'matter and mind':
"[T]he sole
'property' of matter with
whose recognition philosophical materialism is bound up is the property of
being an objective reality, of existing outside our mind." [Lenin (1972),
p.311.]
"Thus…the concept of matter…epistemologically
implies nothing but objective reality existing independently of the human
mind and reflected by it." [Ibid.,
p.312.]
"[I]t is the sole categorical, this sole
unconditional recognition of nature's existence outside the mind and
perception of man that distinguishes dialectical materialism from relativist
agnosticism and idealism." [Ibid.,
p.314.]
"If energy is motion,
you have only shifted the difficulty from the subject to the predicate, you have
only changed the question, does matter move? into the question is energy
material? Does the
transformation of energy take place outsidethe mind, independently of
man…or are these only ideas?...." [Ibid.,
p.324.]
"The fundamental characteristic of
materialism is that it starts from the objectivity of science, from the
recognition of objective reality reflected by science." [Ibid.,
pp.354-55.
Italic emphases in all of the above are those in the original; bold emphases added.]
I
have commented on these ideas in Essay Thirteen
Part One; the
reader is directed there for more details. There is was argued that,
according to the above, since (a)
The 'mind' can't be outside itself, and (b) Only those objects and
processes which are outside the 'mind' can be regarded as material,
therefore (c) The 'mind' can't be material! Paradoxically, Lenin is committed to
the idea that 'the mind' is non-material!
Alongside
other DM-fans, Lenin plainly pitched his tent well-and-truly inside the Cartesian Paradigm;
all he had was a different 'solution' to the pseudo-problem of the relation
between 'mind' and matter.
The
pressure on certain comrades to come up with a materialist-sounding theory
to explain
'consciousness' (etc.) continues to affect their judgement (for instance, here). For example, this is how Alex Callinicos
characterised Daniel Dennett's work in this area:
"...Dennett
comes at the argument from a specific angle. He is by background a philosopher
of mind, the author of a number of well known books, notably Brainstorms
(1978) and Consciousness Explained (1991). As the title of the latter
book indicates,
Dennett is hardly the kind of
analytical philosopher
happy to
concentrate on the exquisitely precise examination of details of linguistic
usage. He isn't afraid to take on big subjects.
"There is,
moreover, a unifying theme that connects Darwin's Dangerous Idea with
Dennett's earlier writings. In the latter he has been concerned to develop what
might be described as a non-reductionist materialist theory of the mind. In
other words,
he has sought to find a way of treating the mind as a natural phenomenon,
whose activities are continuous with those in the physical world, while at the
same time recognising that human beings are 'intentional systems' whose
behaviour cannot be explained without ascribing to them beliefs, desires and
other mental states. Dennett accepts that these states cannot be reduced to
corresponding physical states, but he wants to avoid treating them (as many
contemporary philosophers still do) as partaking of some mysterious
'mindstuff' fundamentally different from the physical world.
"Dennett has
attempted to clarify the issues involved
by drawing on
Artificial Intelligence (AI), a discipline that came into existence to
try to understand the computers it helped to create. He argues that AI can help
us to understand the human mind in two ways. First, to the extent that computers
do things that are analogous to what minds do, they show that mental activity is
best understood less in terms of the physical hardware it depends on (the brain
and nervous system in the case of humans), but rather in terms of the functions
which it realises. AI suggests that these functions do not necessarily have to
be performed by physical organisms like us. Computers must, at least for some
purposes, be treated as intentional systems.
"Secondly, AI
explains how computers perform their functions by analysing them into component
sub-systems each of which undertakes tasks which require less intelligence than
those of the computer as a whole, and each of which is in turn composed of
progressively smaller and less intelligent sub-systems. Dennett suggests we
think of computers as composite beings made up of homunculi (tiny men):'The
highest level design breaks the computer down into a committee or army of
intelligent homunculi with purposes, information and strategies. Each homunculus
in turn is analysed into smaller homunculi, but, more important, into less
clever homunculi'.
"Dennett
believes that AI can throw light on the apparent mystery of how intentionality
-- all the complexity and richness of human mental life -- can somehow emerge
from brute, mindless matter. The analogy of the computer, composed of
progressively less intelligent sub-systems (increasingly stupid homunculi),
shows that there is no sharp dividing line between mind and matter but a series
of continuous gradations which blur this distinction. The mind itself
straddles the boundary between the mental and the 'merely' physical since,
like the computer, it is composed of a number of sub-systems each of which
displays less intentionality than the system as a whole:
'In an
organism with genuine intentionality -- such as yourself -- there are, right
now, many parts, and some of these parts exhibit a sort of semi-intentionality,
or mere as if intentionality, or pseudo-intentionality -- call it what you like
-- and your genuine, full-fledged intentionality is in fact the product (with no
further miracle ingredients) of the activities of all the semi-minded and
mindless bits that make you up.... That is what a mind is -- not a miracle
machine, but a huge semi-designed, self-redesigning amalgam of smaller machines,
each with its own design history, each playing its own role in the "economy of
the soul".'
"This view of
mind is brilliantly developed in Consciousness Explained. Here in
particular Dennett seeks to refute the conception of mental life inherited from
Descartes according to which there is 'a special centre in the brain' which is
the focus of consciousness. He offers in its place 'the Multiple Drafts model of
consciousness', according to which:
'all
varieties of perception -- indeed all varieties of thought or mental activity --
are accomplished in the brain by parallel, multitrack processes of
interpretation and elaboration of sensory inputs. Information entering the
nervous system is under continuous "editorial revision".'
"There is no
specific point which can be identified as 'the moment of consciousness'; what we
call consciousness is simply the effect of all these simultaneously occurring
processes. The mind must be seen as a 'Joycean machine', 'a cobbled together
collection of specialist brain circuits', which functions by 'yoking together
these independently evolved specialist organs in common cause and thereby giving
their union vastly enhanced powers'.
"But this
theory, Dennett argues, is only one strategy which materialist explanations of
the mind can pursue. 'There are two paths to intentionality,' he writes. That
pursued in books like Brainstorms and Consciousness Explained
represents what he calls the 'synchronic path'. In other words, it uses AI to
help offer a static analysis of how human brains as they exist now, as organs of
a certain living species, perform complex mental functions. But there is another
route to the same goal: 'The Darwinian path is diachronic, or historical, and
concerns the gradual accretion, over billions of years, of the sort of Design --
of functionality and purposiveness -- that can support an intentional
interpretation of the activities of organisms (the "doings" of "agents").'
"In other
words, the theory of evolution offers the possibility of writing a natural
history of the mind, of explaining how all the rich diversity of mental life
gradually emerged from a physical world from which, originally, it was utterly
absent." [Callinicos
(1996), pp.100-03. Bold emphases and italics added.
Quotation marks altered to conform with the conventions adopted at this site. Among other works of Dennett's, Callinicos is
referring to Dennett (1978, 1989, 1991, 1995).]
However, despite what these two might say,
it will be readily apparent that Dennett and Callinicos have bought into the
Cartesian Paradigm, since they both
attempt to explain the connection
between the 'mind' and the
CNS. Just like Lenin,
and countless others, all they offer
is a different 'solution' to this spurious problem. Apart from the traditional
jargon they both use to frame this 'problem', they
leave several other clues
that confirm they are indeed trapped in this Paradigm, one
of the most revealing of these is where Callinicos informs us that Dennett's aim is to:
"...refute the
conception of mental life inherited from Descartes according to which there is
'a special centre in the brain' which is the focus of consciousness. He offers
in its place 'the Multiple Drafts model of consciousness'...". [Ibid. Bold
emphasis added.]
So, for Dennett, the 'mind' no longer operates
through the
Pineal gland -- as it did for Descartes -- it does so right across the brain,
a picture Callinicos appears to endorse. These two don't begin by questioning the origin
of this view of 'the mind' in mystical Platonism, Christianity
and the misuse of ordinary
language (in fact, concerning the latter, they both reject that approach --
see below), and regard
the 'problem of consciousness' as one they can solve with what turn out to be another set of inappropriate metaphors and analogies!
But, if 'the mind' (as it is depicted by Philosophers) is nothing more than a
faint echo of the Christian-Platonic view of the 'soul',
motivated by little other than the
nominalisation and
reification of a handful of psychological verbs (etc.),
then there is in fact nothing to unravel, re-connect or 'solve', and thus no
need for these "stupid
homunculi", in the first place.
In fact, the earlier quotation from Button,
et al, is even more apposite:
"As to the widespread
disparagement of attempts to resolve philosophical problems by way of appeals to
'what we would ordinarily say', we would proffer the following comment. It often
appears that those who engage in such disparaging nonetheless themselves often
do what they programmatically disparage, for it seems to us at least arguable
that many of the central philosophical questions are in fact, and despite
protestations to the contrary, being argued about in terms of appeals (albeit
often inept) to 'what we would ordinarily say...'. That the main issues of
contemporary philosophy of mind are essentially about language (in the
sense that they arise from and struggle with confusions over the meanings of
ordinary words) is a position which, we insist, can still reasonably be proposed
and defended. We shall claim here that most, if not all, of the conundrums,
controversies and challenges of the philosophy of mind in the late twentieth
century consist in a collectively assertive, although bewildered, attitude
toward such ordinary linguistic terms as 'mind' itself, 'consciousness',
'thought', 'belief', 'intention' and so on, and that the problems which are
posed are ones which characteristically are of the form which ask what we
should say if confronted with certain facts, as described....
"We have absolutely nothing
against the coining of new, technical uses [of words], as we have said. Rather,
the issue is that many of those who insist upon speaking of machines' 'thinking'
and 'understanding' do not intend in the least to be coining new, restrictively
technical, uses for these terms. It is not, for example, that they have decided
to call a new kind of machine an 'understanding machine', where the word
'understanding' now means something different from what we ordinarily mean by
that word. On the contrary, the philosophical cachet derives entirely from their
insisting that they are using the words 'thinking' and 'understanding' in the
same sense that we ordinarily use them. The aim is quite
characteristically to provoke, challenge and confront the rest of us. Their
objective is to contradict something that the rest of us believe. What the 'rest
of us' believe is simply this: thinking and understanding is something
distinctive to human beings..., and that these capacities set us apart from the
merely mechanical.... The argument that a machine can think or understand,
therefore, is of interest precisely because it features a use of the words
'think' and 'understand' which is intendedly the same as the ordinary use.
Otherwise, the sense of challenge and, consequently, of interest would
evaporate.... If engineers were to make 'understand' and 'think' into technical
terms, ones with special, technical meanings different and distinct from
those we ordinarily take them to have, then, of course, their claims to have
built machines which think or understand would have no bearing whatsoever upon
our inclination ordinarily to say that, in the ordinary sense, machines do not
think or understand." [Button, et al (1995), pp.12, 20-21. Italic
emphases in the original. Quotation marks altered to conform with the conventions adopted at this site.]
In which case, Callinicos's admission that
Dennett eschews any consideration of
"what we say" is part of the
problem, not part of the solution:
"...Dennett comes at the
argument from a specific angle. He is by background a philosopher of mind, the
author of a number of well known books, notably Brainstorms (1978) and
Consciousness Explained (1991). As the title of the latter book indicates,
Dennett is hardly the kind of analytical philosopher happy to concentrate on the
exquisitely precise examination of details of linguistic usage. He isn't afraid
to take on big subjects." [Callinicos
(1996), p.100. Bold emphasis alone added.]
These "big subjects" are, of course, based on
a distortion of language, which, according to Marx can be avoided if:
"The
philosophers...dissolve their language into the ordinary language, from which it
is abstracted, in order to recognise it, as the distorted language of the actual
world...." [Marx
and Engels (1970), p.118.]
On
AI and Dennett, see the following:
Bennett and Hacker (2008), pp.237-63, Bennett and Hacker (2021), pp.470-91, Button,
et al (1995), pp.88-95, Hutto (1997), Robinson (2003b), and Shanker (1986c, 1987a, 1987b,
1987c, 1988, 1995,
1996b, 1997, 1998). On what computers still can't do, see Dreyfus
(1992). Cf., also Note 2,
Note 61, Note 97 and the references listed in Note 86, below. See also, Fodor
(1998c). For a general criticism of this way of viewing the 'mind', see Putnam
(1988). On Dennett's appeal to evolution and 'design', see
here.
Dennett's reference to increasingly
'stupid homunculi' is also susceptible to the arguments quoted
here and
here.
[There is more than a
faint echo
in Callinicos's argument of Engels's first 'Law' (i.e.,
Q«Q);
that is, as new levels of complexity arise we witness a change of "quantity"
into "quality". I have shown (here
and here) that this
isn't even an example of Engels's 'Law', whatever else we decide to make of
it! On Trotsky's use of this 'Law', see
below.]
Nevertheless, some Marxists might still
object that if there is no material basis for "mental events", "mental
processes" or "mental states" (for example, in the
CNS), this might allow
space for an immaterialist or Idealist theory of the mind/'soul'.
Second, as we have
also seen, these phrases
(i.e., "mental events", "mental processes" and "mental states")
are
hoplessly obscure.
Third, the factors that have generally
motivated materialist worries of this sort stem from an
ancient
tradition that misidentified, distorted, and misconstrued
both the
nature of our psychological make-up and the language we use to give expression
to it. That tradition not only
fetishised
psychological words -- employing along the way specially concocted
terms like "mind", "consciousness", "thought" --, it
is predicated on the
representational theory of language and mind. As such, it ignored and thus
helped undermine the communitarian nature of discourse, relying on a
ruling-class,
Dualist view of one or both. [When I use the word "mind" I am, of course,
employing it in its everyday sense, not as philosophers and mystics have used
it.]
Some might take exception to this and claim
that they aren't Dualists. However, as noted above, it is quite
clear that Dialectical Marxists, almost to a
clone, have bought into the
Cartesian Paradigm,
concerned as they are to try to explain the connection between the 'mind' and
the
CNS. Here is
Callinicos, again:
"Dennett
believes that AI can throw light on the apparent mystery of how intentionality
-- all the complexity and richness of human mental life -- can somehow emerge
from brute, mindless matter. The analogy of the computer, composed of
progressively less intelligent sub-systems (increasingly stupid homunculi),
shows that there is no sharp dividing line between mind and matter but a series
of continuous gradations which blur this distinction. The mind itself
straddles the boundary between the mental and the 'merely' physical since, like
the computer, it is composed of a number of sub-systems each of which displays
less intentionality than the system as a whole....
"This view of
mind is brilliantly developed in Consciousness Explained. Here in
particular Dennett seeks to refute the conception of mental life inherited from
Descartes according to which there is 'a special centre in the brain' which is
the focus of consciousness. He offers in its place 'the Multiple Drafts model of
consciousness', according to which:
'all
varieties of perception -- indeed all varieties of thought or mental activity --
are accomplished in the brain by parallel, multitrack processes of
interpretation and elaboration of sensory inputs. Information entering the
nervous system is under continuous "editorial revision".'
"There is no
specific point which can be identified as 'the moment of consciousness'; what we
call consciousness is simply the effect of all these simultaneously occurring
processes. The mind must be seen as a 'Joycean machine', 'a cobbled together
collection of specialist brain circuits', which functions by 'yoking together
these independently evolved specialist organs in common cause and thereby giving
their union vastly enhanced powers'. [Callinicos
(1996),
pp.100-03. Bold emphases added.]
Callinicos isn't alone. DM-fans who have
written on this topic invariably argue along similar lines.
According to this 'dialectical' view of human
cognition and action, the Cartesian
picture that there is something called "the mind", which isn't material, and which operates
through the
Pineal gland -- a favourite spot, incidentally, for mystics to locate the
'third
eye' -- has been replaced by a series of "modules", "structures" or "processes", which can't be reduced to
their material base (so they aren't material either: that was established as
Lenin's view,
above, and in Essay Thirteen
Part One), but which are no longer localised, as they
were for Descartes.
The details may have changed, but the form has remained the same. And, as
we have seen throughout this Essay, Voloshinov and his epigones
have also adopted
different versions of the same Cartesian myth.
Here, for example, are our old friends, Woods and Grant:
"What
we call 'mind' is just the mode of existence of the brain. This is an
immensely complicated phenomenon, the product of many millions of years of
evolution. The difficulty in analysing the complex processes that occur
within the brain and nervous system, and the equally complex interrelations
between mental processes and the environment, has meant that a proper
understanding of the nature of thought has been delayed for centuries. This
has enabled idealists and theologians to speculate on the allegedly mystical
nature of the 'soul,' conceived as a non-material substance which deigned to
take up temporary residence in the body. The advances of modern neurobiology
mean that the idealists are finally being driven from their ultimate refuge.
As we begin to unlock the secrets of the brain and nervous system, it becomes
progressively easier to explain the mind, without recourse to supernatural
agents, as the sum total of brain activity." [Woods
and Grant (1995), p.286. Quotation
marks altered to conform with the conventions
adopted at this site. Bold emphases added.]
Again, these two don't begin by
questioning Descartes's division of the phenomena into their 'mental' and
physical 'substances'; as we saw with other DM-fans, they have a different solution to the 'problem'
Descartes manufactured. But, this artificial division still remains. To be sure,
they deny there is any need to appeal to a non-material 'soul', but all they
have done is replace the 'soul' with another mysterious entity -- the "mode of
existence of the brain" --, since they don't explain what this could possibly
mean (nor can they -- no one can
without anthropomorphising that organ).
Instead of exposing the bogus nature of the traditional
Platonic-Christian-Cartesian Paradigm
as the misguided product of inappropriate metaphors, distorted language and mystical
theology that is it, they merely offer us their own mysterious alternative to it.
The above two then repeat several of the
errors we met earlier in this Essay:
"Using examples of child
behaviour, he [Vygotsky -- RL] explained why children spend a lot of time
talking aloud to themselves. They are rehearsing the habits of planning that
they would later internalise as inner speech. Vygotsky showed that this inner
speech underpinned the human ability to recollect and recall memories. The
human mind is dominated by an inner world of thoughts, stimulated by our
sensations, which is capable of generalisation and perspective. Animals also
have memories, but they seem to be locked into the present, reflecting the
immediate environment. The development of human inner speech allows humans to
recall and develop ideas. In other words, inner speech played a key role in
the evolution of the human mind." [Ibid.,
pp.293-94. Bold emphases added.]
As we
have seen, all this talk of an "inner world of thought" is Cartesian in all but
name -- while managing to be incoherent into the bargain.
Here is Engels himself, endorsing
Feuerbach's view (albeit with certain reservations), along the same
lines:
"The course of evolution of
Feuerbach is that of a Hegelian -- a never quite orthodox Hegelian, it is true
-- into a materialist; an evolution which at a definite stage necessitates a
complete rupture with the idealist system of his predecessor. With irresistible
force, Feuerbach is finally driven to the realization that the Hegelian
premundane existence of the 'absolute idea', the 'pre-existence of the logical
categories' before the world existed, is nothing more than the fantastic
survival of the belief in the existence of an extra-mundane creator; that the
material, sensuously perceptible world to which we ourselves belong is the only
reality; and that our consciousness and thinking, however supra-sensuous they
may seem, are the product of a material, bodily organ, the brain. Matter is not
a product of mind, but mind itself is merely the highest product of matter.
This is, of course, pure materialism. But, having got so far, Feuerbach stops
short. He cannot overcome the customary philosophical prejudice, prejudice not
against the thing but against the name materialism....
"In the second place, we
simply cannot get away from the fact that everything that sets men acting must
find its way through their brains -- even eating and drinking, which begins as a
consequence of the sensation of hunger or thirst transmitted through the brain,
and ends as a result of the sensation of satisfaction likewise transmitted
through the brain. The influences of the external world upon man express
themselves in his brain, are reflected therein as feelings, impulses, volitions
-- in short, as 'ideal tendencies', and in this form become 'ideal powers'....
"As all the driving forces
of the actions of any individual person must pass through his brain, and
transform themselves into motives of his will in order to set him into action,
so also all the needs of civil society -- no matter which class happens to be
the ruling one -- must pass through the will of the state in order to secure
general validity in the form of laws....." [Engels
(1888), pp.596-97,
615. Quotation marks
altered to conform with the conventions adopted
at this site. Bold emphasis added.]
This idea even crops up in Das Kapital:
"My dialectic method is not
only different from the Hegelian, but is its direct opposite. To Hegel, the
life process of the human brain, i.e., the process of thinking, which, under
the name of 'the Idea,' he even transforms into an independent subject, is the
demiurgos of the real world, and the real world is only the external, phenomenal
form of 'the Idea.' With me, on the contrary, the ideal is nothing else than
the material world reflected by the human mind, and translated into forms of
thought." [Marx
(1996), p.19. Quotation marks
altered to conform with the conventions adopted
at this site. Bold emphases added.]
To be sure, Marx is referring to Hegel's
ideas here, but he clearly endorses his view of"the life process of the human
brain, i.e., the process of thinking"; he merely wishes to reverse the direction
of analysis, or perhaps even of causation:
"[T]he ideal is nothing else
than the material world reflected by the human mind, and translated into forms
of thought." [Ibid.]
That, too, is Cartesian in all but name -- as
are these comments of Trotsky's:
"...[I]t
is precisely dialectical materialism
that prompts us to the idea that the psyche could not even be formed unless it
played an autonomous, that is, within certain limits, an independent role in the
life of the individual and the species.
"All the same, we approach
some sort of critical point, a break in the gradualness, a transition from
quantity into quality: the psyche arising from matter, is 'freed' from the
determinism of matter, so that it can independently --- by its own laws --
influence matter." [Trotsky (1986), p.106. Quotation marks altered to
conform with the conventions adopted at this site. Bold emphases added. "Psyche" is clearly
Trotsky's word for 'mind' or 'consciousness'.]
Here
also is Spirkin:
"While idealism creates a
gulf between reason and the world, materialism tries to discover the unity
between the two by inferring the spiritual from the material. In
materialism, the interpretation of consciousness is based on its recognition as
a function of the human brain, the essence of which lies in the reflection and
constructive-creative transformation of the world. Historical-materialist
theory maintains that it is impossible to analyse consciousness in isolation
from other phenomena of social life. From the very beginning consciousness has
been a social product and it will remain so as long as human beings exist. The
human brain embraces the potentials evolved by human history, the inherited
abilities that are realised through training and education and the whole
assembly of social influences, and through exposure to world culture. The
brain becomes the organ of consciousness only when a person is drawn into social
life and assimilates historically evolved forms of culture. The essential
purpose of consciousness is to give people a true orientation in the world, the
ability to know and transform it by means of reason. When we say that a person
is conscious of something, we mean that he understands the meaning of what he
has perceived or remembered and takes into consideration the possible
consequences of his actions and can be held responsible for them to society and
himself.
"Human consciousness is a
form of mental activity, the highest form. By mental activity we mean all mental
processes, conscious and unconscious, all mental states and qualities of the
individual. These are mainly processes of cognition, internal states of the
organism, and such attributes of personality as character, temperament, and so
on. Mental activity is an attribute of the whole animal world.
Consciousness, on the other hand, as the highest form of mental activity, is
inherent only in human beings, and even then not at all times or at all levels.
It does not exist in the newborn child, in certain categories of the mentally
ill, in people who are asleep or in a coma. And even in the developed, healthy
and waking individual not all mental activity forms a part of his consciousness;
a great portion of it proceeds outside the bounds of consciousness and belongs
to the unconscious phenomena of the mind. The content of the activity of
consciousness is recorded in artefacts (including language and other sign
systems), thus acquiring the form of ideal existence, existence as knowledge, as
historical memory. Consciousness also includes an axiological, that is to say,
evaluative aspect, which expresses the selectivity of consciousness, its
orientation on values evolved by society and accepted by the individual --
philosophical, scientific, political, moral, aesthetic, religious, etc. It
includes the individual's relation both to these values and to himself, thus
becoming a form of self-consciousness, which is also social in origin. A
person's knowledge of himself becomes possible thanks to his ability to relate
his principles and orientation to the stand points of other people, his ability
to consider these stand points in the process of communication. The very term
'consciousness', that is to say, knowledge acquired together with others, points
to the dialogical nature of consciousness." [Spirkin
(1983), pp.154-55. Quotation marks
altered to conform with the conventions adopted at this site. Bold emphases added.]
Here, too, is Cornforth:
"Mental functions are
functions of highly developed matter, namely, of the brain. Mental processes are
brain processes, processes of a material, bodily organ.
"The essential feature of
mental processes is that in and through them the animal continually builds up
most complicated and variable relations with its surroundings. When we
perceive things we are relating ourselves to external objects through the
perceptual activity of the brain. And when we think of things, we are
relating ourselves to external objects through the thought activity of the brain....
"Adopting the scientific
approach to the study of consciousness, Marxism therefore denies the idealist
theory that when we perceive, feel or think there are two separate processes
going on -- the material process of the brain and the mental process of
consciousness. Only one process is involved, namely, the material
process of the brain. Mental processes are simply one aspect of the process
of the functioning of the brain...." [Cornforth (1974), pp.22-24. Bold
emphases added; italic emphases in the original.]
Once more, the above comrades
don't even
think to challenge the Idealist dichotomy ('mind'/brain); just like Lenin,they simply have a different
'solution' to it.
[The neo-Cartesian views of
several more Dialectical Marxists will be added at a later date. In the meantime, see
Note 55a, below.]
So, as noted several times, all that the above authors have to offer is a different solution
to the same pseudo-problem, the relation between 'mind' and matter. As we saw
was the case with
Dennett, Callinicos and several others, the above theorists fail to question the origin of this myth -- in
Platonic mysticism and the
misuse of ordinary language --, but simply regard the 'problem' of
'consciousness' as something they are well equipped to solve. By accepting the
problematic laid down by Descartes,
along with his
jargon, they have only succeeded in trapping themselves within this Ancient Paradigm. As Anthony Kenny noted
(quoted earlier):
"Descartes view of the nature
of mind endured much longer than his view of matter. Indeed among educated
people in the West who were not professional philosophers it is still the most
widespread view of the mind. Most contemporary philosophers would disown
Cartesian dualism but even those who explicitly renounce it are often profoundly
influenced by it. Many people, for instance,
go along with Descartes in identifying the mental realm as the realm of
consciousness. They think of consciousness as an object of introspection; as
something we see when we look within ourselves. They think of it as an
inessential, contingent matter that consciousness has an expression in speech
and behaviour. Consciousness, as they conceive it, is something to which each of
us has direct access in our own case. Others, by contrast, can only infer to our
conscious states by accepting our testimony or making causal inferences from our
physical behaviour." [Kenny (1992), p.2. Bold emphases added; paragraphs
merged.]
This view has
largely remained unchallenged for
centuries:
"...[V]irtually no
philosopher doubted, from the time of
Locke
until roughly 1914, that, whatever concepts and ideas were, they were clearly
mental objects of some kind. And no large-scale and comprehensive demolition job
was done against this particularly wide-spread and influential philosophical
misconception until
Wittgenstein
produced his
Philosophical Investigations...." [Putnam (1975b), p.7.]
What is even more ironic is that the above
comrades attempt to 'solve' this pseudo-problem with a series of dialectical gyrations that are even more
obscure than the conundrum Descartes bequeathed to humanity!
Fourth, there is no need for materialists to
concede an inch of ground to the Idealists in this area since there is in fact a materialist
account of psychological phenomena already available to us. The latter relies on
concepts drawn from
HM, or which are consistent with it
-- i.e., on the psychological vocabulary found in ordinary language.
The
pay-off is that this approach undermines immaterialism far faster than DM
succeeds in perpetuating it.
[Further details are given below; see
also,
Note 86.]
55. On
this, see Note 54 above. Moreover,
as we saw in Essay
Twelve Part One, such a theory
would collapse into incoherence anyway. [In relation to that, see the references
itemised in Note 86, below.]
55a. Often this will include a direct
or indirect reference to
Engels's first 'Law' in support of this inference. However, as we saw in Essay Seven
Part One, that 'Law' is far too
insubstantial to support the weight of a
Mayfly on a
crash diet, let alone this a priori
dogma.
An
excellent recent illustration of this reliance on the 'First Law' can be found in this quotation from
Mike
Macnair:
"The phrase 'transition from
quantity to quality' has the disadvantage of Hegelianism. But it has the helpful
aspect of drawing attention precisely to the fact that crisis emerges out of
prior, gradual processes, which are perfectly identifiable while they are going
on, even if they may happen not to be noticed until the outbreak of crisis....
"Materialism, in the sense of
Marxist materialism, has more than one level. The most basic level is that it is
unnecessary to suppose the existence of god or gods, a 'world-spirit', the
Hegelian self-moving Idea, spirits, the existence of the soul, the élan vital,
or an immaterial homunculus 'consciousness' which sits in the human body and
drives it as a motorist drives a car. The phenomena can be adequately explained
by the methods of the sciences without any such suppositions. The ideas in my
head are electro-chemical phenomena in my brain which are part of an embodied
consciousness, which has developed through the physical (Darwinian) and social
evolution of the human species. The words I am writing are -- as I write them --
electrical patterns in the computer; when they are printed they will be patterns
of ink on the printed page. They are just as material as trees, etc." [Quoted
from
here. Bold emphases added.
Quotation marks altered to conform with the conventions adopted at this
site.]
Alas, in the very same breath as rejecting the
need to appeal to an homunculus, McNair refers his readers to concepts that only work
if there is indeed one of these 'beasties' lodged in his head!
To be sure, McNair doesn't explicitly
link Engels's 'Law' with 'consciousness', but there is little in his article to suggest he would disagree with this comment
of Chris Harman's:
"Such developments do not
involve just quantitative changes. As the growth of labour, the growth of
sociability and the growth of language reinforced each other, encouraging the
selection of a whole range of new genes, new networks of nerve cells wouldemergein the brain, making possible whole new ranges of interaction
between people and the world around them….
"So there has to be a
recognition of how
quantity turns into quality, of how through successive changes animal life
gave birth to that new form of life we call 'human', which has a dynamic of its
own, shaped by its labour and its culture not by its genes…." [Harman
(1994), pp.102. Bold emphases added.]
Indeed, Trotsky himself seems to have
endorsed this point:
"...[I]t is precisely
dialectical materialism that prompts us to the idea that the psyche could not
even be formed unless it played an autonomous, that is, within certain limits,
an independent role in the life of the individual and the species.
"All the same, we approach
some sort of critical point, a break in the gradualness, a transition from
quantity into quality: the psyche arising from matter, is 'freed' from the
determinism of matter, so that it can independently --- by its own laws --
influence matter." [Trotsky (1986), p.106. Quotation marks altered to
conform with
the conventions adopted at this site. Bold emphasis added.]
Once more, all four are clearly operating within the
Cartesian Paradigm, arguing that
the alleged link between 'mind' (or, in Trotsky's case, the "psyche") and matter needs to be explained. These comrades just assume there is
something called "the mind" which has to be related to "the body";
they don't even think to question this ancient
myth. Once again, all they have to offer is their own solution to this
intractable 'problem'. And yet, if there is no such thing as 'the mind', the
'problem' simply evaporates.
[Again, when I use the word "mind" I am employing it in its everyday sense, not
as philosophers and mystics have used it.]
Sure, they reject dualism, but
Descartes's concepts have been transposed and then translated into
what they take to be materialist terms. No attempt has been made to reject
up-front Descartes's distortion of language, or question his paradigm root-and-branch. The terms of the
entire problematic have been set by Descartes's Paradigm; the attitude and
approach adopted by DM-theorists toward that Paradigm is
consistent with what I alleged of them in
Essay Two:
As
will soon become apparent, for all their claims to be radical, when it
comes to Philosophy
DM-theorists are
surprisinglyconservative -- and universally incapableof
seeing this even after it has been pointed out to them!
[An
excellent example of this phenomenon, and one that has
been highly influential
on how DM-theorists receive and then respond to such criticism, has been posted
here.]
At a
rhetorical level, this philosophical conservatism is camouflaged behind what at first sight
appears to be a series of disarmingly modest denials --,
which are then promptly flouted.
The
quotations given below (in
Essay Two and in
Note 1) show that DM-theorists are keen to
deny that their system is wholly, or even partly,
a priori, or that it has been dogmatically imposed on
the world, not read from it. However, the way that dialecticians themselves
phrase their theories contradicts these seemingly modest-looking claims,
revealing that the opposite
is in fact the case....
However, unlike
dialecticians, Traditional Metaphysicians were quite open and honest about what
they were doing; indeed, they brazenly imposed their a priori theories on
reality and hung the consequences.
But,
because dialecticians have a novel (but nonetheless defective) view of both
Metaphysics and FL (on that, see
here and
here), they are oblivious of the
fact that they are just as eager as Traditional Theorists have always been to
impose their ideas on the world, and equally blind to the fact that in so-doing
they are aping the
alienated thought-forms of their class enemy, whose society they seek to abolish.
Naturally, this means that their 'radical' guns were spiked beforethey were even loaded; with such weapons, is it any wonder that DM-theorists
fire nothing but philosophical blanks?
[FL = Formal Logic.]
DM is
a conservative theory precisely because its adherents have imported and
then adopted the
distorted methods,
a priori
thought-forms, theories and meaningless jargon of Traditional Philosophy.
Finally, some might conclude that the approach
adopted in the Essay means that Ms Lichtenstein is a
Behaviourist. Nothing could be further from the truth. On that, see
Hacker (1993e).
56.
These allegations are further substantiated below, as well as
here.
56a0.
The evidence that there are
indeed such
sub-vocal movements has been controversial at least since the days of
Watson.
[There is a wealth of detail in Sokolov (1975); alas, this book is now nearly
forty years out-of-date.]
However, communication and surveillance technologies are now being built around this phenomenon,
so it is difficult to conclude it is illusory. [On that,
see
here and
here.] If this is indeed a genuine phenomenon, then "inner speech" (if this
is what it is!) will simply be 'outer speech' with the volume set close
to
zero, as noted above. In other words, it will be another category of overt
behaviour, as Watson had claimed. Even so, it is far from clear whether or not scientists
involved in this research are
studying the same phenomenon as Watson. Moreover,
Skinner
(working in the same (Behaviourist) tradition as Watson)
later denied this phenomenon could even account for thinking. [Skinner (1992), pp.434-35.]
On
that, see here.
[Unfortunately, this link is now dead! It used to link to a site that debunks many of the misrepresentations of
Behaviourism with which we have become familiar since
Chomsky's famous 'refutation' of Skinner.]
That 'refutation' is reappraised in Richelle (1995),
and shown to be far less conclusive than many have supposed. [On Richelle's book, see
here. (This link is also dead!) ] See also Julià (1983) and MacCorquodale (1970), whose argument has
been neatly summarised by a friend of mine, as follows:
"MacCorquodale among other
things characterized Chomsky's review as 'an amalgam of some rather out-dated
behaviouristic lore including reinforcement by drive reduction, the extinction
criterion for response strength, a pseudo-incompatibility of genetic and
reinforcement processes, and other notions which have nothing to do with
Skinner's account.' In other words MacCorquodale charged that Chomsky was
basically attacking a straw man which had little to do with Skinner's actual
views, in particular failing to distinguish the ways in which Skinner's radical
behaviourism differed from other forms of behaviourism." [Jim Farmelant, quoted from
here. Spelling altered to conform
with UK English.]
56a. Indeed,
how we even know that "inner speech is 'predicative' if we have no direct
access to it (and no one has ever heard it) remains something of a mystery.
56b. Further ruminations on this topic will take us too far into a consideration of Wittgenstein's
famous
Private Language Argument [PLA]. More on that in another Essay.
However, it is
possible to argue that the PLA was prefigured in
Marx's
work, as indeed it was, in Part Two, Chapter Three of Voloshinov's own book --, i.e.,
pp.83-98. This may perhaps be from where Wittgenstein derived some of these ideas, via
Sraffa. I have said more about that,
here.
57.
For somerecent work on "inner speech", see
Note 42, above.
58.Again, it is worth pointing out that
it isn't easy to see what
Parrington means by "inner speech" -- since he forgot to tell us!
On soliloquy, cf., Ryle (1982b) and Squires
(1974). See also, Hacker (1993b, 1993c, 1993d), Hark (1990), and Johnston (1993).
In addition, see the references in Note 86,
below.
59.A very useful summary of these and other
senses of "meaning" can be found in Audi (1999), pp.545-50 (written by
Brian Loar).
This isn't, of course, to deny the sort of complexity
that interests linguists. On the latter, see Cruse (2000), pp.46-63.
60.
It has to be said: this is also true of Wittgenstein! He only examined a
limited range of meanings of "meaning", and in his later work conflated the sense
of a sentence with its meaning. The same is the case with many Wittgensteinians,
too.
61.
The alleged connection between thought and language is far too large a topic to
address adequately here; I will attempt to do so in a later Essay. Anyway,
Marxists
have in general adopted the view that these two seemingly disparate 'phenomena' are intimately linked, i.e., that
language is the embodiment of 'thought'.
"Language is the immediate actuality of thought.Just as
philosophers have given thought an independent existence, so they were bound to
make language into an independent realm. This is the secret of philosophical
language, in which thoughts in the form of words have their own content. The
problem of descending from the world of thoughts to the actual world is turned
into the problem of descending from language to life." [Marx
and Engels (1970), p.118. Bold emphasis added.]
[There is a very good summary of the views expressed by Marx and other Marxists
(concerning the relation between language and thought) in the opening pages
of Sokolov (1975), pp.1-3.]
The problem with all this is that it relies
on a very narrow understanding of our use of the word "thought". This then
motivates
comrades to postulate the existence of "inner speech" to account for this
restricted understanding, transforming 'thought' into a process that takes place
in the head. Indeed, Sokolov falls into that trap, too.
[On this, see Bennett and Hacker (2021), pp.187-93, 384-90. See also,
Hacker (1993a), pp.143-82, and Teichmann (2015). In this Essay I have alluded
in general to this narrow,
traditional concept (i.e., 'thought') using 'scare quotes'.]
However, typical uses of the verb form of this word include the following:
T1: NN thought that the
strike would start on the 5th.
T2: I think that you are too
quick to condemn the organisers of this demonstration.
T3: The shop stewards thought
about management's latest offer for about ten seconds, and then rejected it.
T4: John thought about his
brother James.
T5: I need to think through
what you said.
T6: The strike committee
thought over the proposal to widen the strike.
T7: I can't think who you
mean.
T8: I'd hate to think why he
said that.
T9:
NN would like to think your application for this post is serious.
Here,
this verb seems to take on the following forms:
(1) "NN
(or one or more unnamed individuals) thinks that
p" (where
"p" is a propositional variable, and "NN" stands for a
Proper Name),
(2) "NN
(or one or more unnamed individuals) thinks about/through/over F", and "NN
(or one or more unnamed individuals) φ-ies to
think who/why F" (where "F" is a noun
or verb phrase and "φ" a verb phrase).
This
isn't to suggest that the above exhausts the verbal forms the use of this word can
assume, far from it. For more detail on this, see Hacker (1993b, 1993c,
1993d), and Ryle (1971e, 1971f, 1971g, 1982a). On our use of phrases like "the
mind", see Hacker (1993f), as well as these videos:
Video Five: Peter Hacker
-- Resolving The
'Mind/Body' Problem
Video Six: Peter Hacker
On The Mind
Neuroscience And Free Will
[As well as a video of Hacker's mentioned
earlier. Nevertheless, it is worth
noting that Ryle makes far too many
concessions to the idea that there is, or can be, 'non-linguistic thought'; but this
isn't the place to dissect his ideas. Hacker's
work is a necessary corrective, however.]
There
are, of course, other uses of "thought" and its cognates. We can speak about "human thought" in
general, and we can say things like "She put much thought into her reply", "Your thoughts on this would be most welcome", "A thought just crossed
my mind", or even "He gave no thought to her feelings".
However, this nominal use of this word is often clearly metaphorical. Do thoughts really cross minds? How fast? In which
direction? Do they walk, run or hitch a ride along a nerve? Or are we not
merely adverting to the fact that something had occurred to the individual
concerned? And when
someone welcomes your thoughts, are they really asking that you reveal to them a part of
your
brain, or even its inner goings-on via a brain scan or brain probe? Are they not rather expressing a desire
to hear what you have to say? Or, maybe, that they value your opinion? Moreover, when we
put much thought into something (such as a reply), is the 'thought' involved packed in there tightly, or
loosely? And how is such thought to be measured (so that we know there is indeed
much packed in
there)? Are replies really containers? Or are we not
saying that the work in question showed great care and attention to detail? Are we perhaps alluding to the time spent on it, the superior quality of its mode of expression,
style
and clarity? And
when we refer to "human thought" are we referring to what goes on in our heads?
Why then do we say that libraries are repositories of human thought? Are we not
in fact referring to our collective intellectual achievements and practices?
[There
are many more examples of the deconstruction of nominalised words
connected with human thought to be found in Hacker's and Ryle's work, referenced above.]
Nevertheless, the idea that the word "thought" names something that goes on in
our heads is based on a defective analogy drawn between with sentences like these:
T9: NN had a thought in her
head.
T10: NN had a tooth in her
head.
From
T9, we may obtain:
T11: NN thought that...
or,
perhaps:
T12: NN thought about...
But,
we can't obtain the following from T10:
T13: NN toothed that...
or,
T14: NN toothed about...
This
shows that the nominal form of "thought" doesn't work like an ordinary
count noun,
and that the verb form (expressed in T11) is its primary form. In that case,
"thought" doesn't name anything in the head (or anywhere else),
unlike "tooth".
The nominal use of "thought" is a result of the
nominalisation of the verb form (hence, the derivation of T11 and T12, say, from T9 is
unexceptional). And that,
naturally, helps account for the complex or metaphorical nature
of the nominal form of this word.
Of
course, we could imagine a poet, say, using T13 or T14 figuratively, but no
literal sense can be made of either as they stand.
Someone might object, arguing that to tooth something is to bite it. Maybe so,
but that can't be obtained from T10, as T11 can be obtained from T9 -- no sense
can be made of "NN toothed that a cake", or "NN toothed about a sandwich."
This
result shouldn't worry Marxists since it puts the active form of the word
first, emphasising its social nature.
Be this as it may, Holborow seems to be a little more
secure in her grasp of the intimate connection between language and 'thought'
than Parrington appears to be. [Cf., Holborow (1999), pp.31-32.]
For example, Parrington says things like the
following:
"Voloshinov was primarily
concerned to develop a Marxist theory of language, but his theories have a
crucial importance for understanding human consciousness because they also deal
with two of psychology's central questions -- the relationship between the
individual and society and the relationship between thought and language."
(p.123)
"Both Chomsky and
Pinker
are similarly incapable of explaining the influence of social change on language
and thus of the development of language over time. Both these flaws are linked
to a much deeper and more fundamental one -- the separation between individual
thought and social language that we found in the earlier theories of language.
Voloshinov's work is as relevant today as it was in the 1920s because it
bridges the gap between individual thought and social language." (p.125)
"We saw that in many theories
of language there was a split between social language and the individual
consciousness. This was particularly reflected in
Saussure's (sic) formulation which counterposed the fixed language of society to
the 'wilful, intellectual act' characterising the speech of the individual. A
similar split characterises the theories of Chomsky and his followers. Thus
Chomsky refers to 'competence' versus 'performance' in speech. These views of
language are themselves a consequence of a lack of awareness about the
material nature of the connection between thought and language." (p.129)
"If inner speech is the
link between thought and language as Voloshinov argues, how can we learn
about its shape and form?... Thus there is a gap between thought and words
but words remain the only possible way of expressing ourselves fully to others.
Thought which remains unexpressed remains immature and eventually dies out...."
(Loc cit., pp.134-35)
[Bold emphases and links added.]
The above looks like it distinguishes thought from
language; indeed, it stresses the "gap" between them that "inner
speech" is supposed to "bridge". If 'thought' were solely a feature of
our use of
language, then, of course, no bridge would be needed.
"Voloshinov provided a number
of key insights which allow us to overcome the gap between social language and
individual consciousness. Firstly, on the material basis of thought. What
concepts such as Pinker's 'mentalese' do is merely obscure the fact that both
thought and language have the same material basis, that is, words." (p.129)
"Language, therefore, is not
just an expression of otherwise independent and fully formed thought, but rather
is a necessary form of the thought's realisation." (Loc cit.,
p.135)
[Ibid.]
And yet, it is difficult to see how the
"material basis" for thought can be "words" if there is also a "gap between
thought and words". Moreover, if a thought can be "unexpressed" and can eventually "die out", thought must be possible without language. From
this it seems that Parrington accepts the
Platonic-Christian-Cartesian Paradigm,
that thought precedes language, and that language is a vehicle by means
of which we
express thoughts formed independently of, and anterior to, language. If this social medium
(language) is independent of thought, then thought can't be a social product,
but the product of an atomised individual. In which case, this is just a
contemporary
version of the
early bourgeois idea that language is only a medium of communication when
it allows us to exchange ideas formedindependently of it. As noted
earlier:
It was established in Essay Twelve Part Seven
(not yet published) that a particular theory of language has dominated 'Western'
(and, indeed, 'Eastern') thought for over two thousand years. This approach sees the primary role of discourse (in fact, in many cases, its only
role) as representational, and hence that it acts solely as a vehicle for thought
(or an outer expression of 'inner thought'),
not as a means of communication. In fact, if discourse was
ever (reluctantly) seen as a means of
communication, it was then often regarded as a vehicle for communicating thoughtsalready arrived at
independently of, and prior to, social interaction.
Although Parrington appears to reject the
idea that language is an individualised phenomenon, and thus that it is primarily
representational (in, for example, his discussion of the ideas of
Ferdinand de Saussure --
pp.123ff -- although we will soon see that several other things he says
about language reveal that he has in fact also retreated, and accepts the idea that language isn't a
social phenomenon, after all), this seems to commit him to the theory that
thought is representational, and that language feeds off this
atomised, or individualised, component of human psychology.
"The foremost philosopher of language at the time Voloshinov was writing was the
Swiss linguist Ferdinand de Saussure, often described as the founder of modern
linguistics. Saussure believed that the job of linguists was not to study the
history of language but its structure, and that structure should be studied in
its own terms and for its own sake. Expressing this belief in a famous
statement, Saussure said:
'In separating language from speaking we are at
the same time separating: (1) what is social from what is individual; and (2)
what is essential from what is accessory and more or less accidental. Language
is not a function of the speaker; it is a product that is passively assimilated
by the individual. It never requires premeditation and reflection enters in only
for the purpose of classification.... Speaking, on the contrary, is an
individual act. It is wilful and intellectual.'
"Saussure had a reason for making such a separation between language and actual
speech. By considering language as a system set aside from the individuals who
actually practise it, he was able to make some important generalisations about
the nature of language which had been impossible before. One of Saussure's
primary concepts was that words are arbitrary signs. This lack of a physical
connection with the objects in the outside world that words specify has the
important consequence that it provides the conditions for the meaning of words
to change over time. Another fundamental concept that Saussure introduced was
the idea that the linguistic value of a word in any given context is determined
by the other words with which it occurs, and that this follows certain rules of
grammar (the way words are strung together) and phonetics (the way words sound).
This interaction he compared to that which exists between the different pieces
in a game of chess. As well as his own personal discoveries, Saussure's work was
of more general importance in opening up the possibility of studying the
structural constraints and the rules which govern language by considering it as
an abstract system. This has indeed been the main preoccupation of linguistics
over the course of the 20th century.
"Voloshinov was quite aware of the significance of Saussure's work. But he
argued that his achievements had come at a price. According to Voloshinov, one
of the problems of making a 'comparison of language to a system of mathematical
signs', as Saussure did, was that this can develop into an interest solely in
'the inner logic of the system of signs itself, taken as in algebra, completely
independently of the ideological meanings that give the signs their content'.
Voloshinov pointed out that Saussure's formulation failed to explain how
language as a social form changes over time and from where individual creativity
comes.
"Voloshinov counterposed Saussure's ideas to those of an earlier philosopher of
language,
Wilhelm von Humboldt
whose theory of language can be viewed as the polar
opposite to that of Saussure. Inspired by
Romantic
philosophy, von Humboldt believed that the essence of language is precisely
that creative act of the individual which Saussure had rejected as a subject for
study. However, this meant that von Humboldt was unable to view language as a
social system. He thus found it impossible to explain why particular individuals
say the things they do, assuming as he did that individual creativity was given.
His approach ended up both emptying language of its ideological content and
neglecting its intrinsically social nature. According to Voloshinov, both
Saussure and von Humboldt had falsely separated the individual from society.
Although Saussure recognised language as a social phenomenon, his method
prevents us studying it in a social context." [Parrington (1997),
pp.123-24. Links added.]
And yet, as we have seen (especially
here), Voloshinov's work in the end also treats
language as an individual, not a social, phenomenon. So, Parrington's comments about
thought (alongside other things he says about language, examined in the main
body of this Essay) confirm the allegation made above that Parrington has
also bought into this bourgeois individualist idea.
Although, it is perhaps better to say that
while Voloshinov and his epigones say they accept the social nature of
language, they have clearly failed to think through the implications of that
commitment.
As I
have argued in Essay Twelve Part One:
Furthermore, an even less well appreciated corollary of this view of discourse is that language is
primarily a vehicle of communication, not of representation.
Added in and Endnote: The attack on the social roots of language -- replacing
a Marxist commitment to this idea
with a mystical belief that language in effect contains a secret code
which is capable of reflecting the underlying 'Essence' of Nature, and which has somehow
also been stitched into the fabric of reality, so that the one can 'reflect' the
other
-- helped motivate the belief that
language is primarily
representational (as we will see in the next two Parts of Essay Twelve
-- summary here).
According
to this ancient doctrine, language itself contains hidden clues -- clues that can only be
accessed, or 'understood', by the elite, their ideologues, their hangers-on, lackeys, or by specially-trained
'thinkers'. Cosmic verities like this lie
way beyond the grasp of ordinary humans -- so the story goes --, trapped as they
are in a world of 'commonsense', dominated by ordinary language and 'formal
thinking'. This
'Divine Code' was thought to have been written into, or actually was, the
'primary language' given
to Adam by God
-- but, similar myths are also found in other religions and cultural traditions. Much of
Hermetic,
Neo-Platonic,
Alchemical
and Kabbalistic mysticism is
largely based on this dogma.
[On this, see Bono (1995), Eco (1997), and Vickers (1984b). This topic
will be explored more fully in Essay Fourteen Part One (summary
here), and
other Parts of Essay Twelve.]
Signs,
or 'hidden messages', were believed to be written in the stars,
too, or in
sacred books, tea leaves, the flight of birds, the organs and entrails of slaughtered animals -- or,
indeed, in its more recent incarnation, encrypted somehow in our central nervous system as a
"transformational grammar"
("unbounded
merge") or
"language of thought".
In DM-circles, this doctrine surfaces as part of the a
priori dogma that thought is dialectical because reality is dialectical
(which 'profound secret' is, alas, hidden from those who refuse to see, or who
just do not
"understand" dialectics). Hence, DM can
be called an "Algebra of Revolution", which works because it alone is tuned into the "pulse of reality"
--
or, perhaps even: because reality
'dances'
to its tune.
It is undeniable that some Marxists have acknowledged the
(perhaps limited)
applicability of the former corollary -- that language is conventional --, but hardly any
(perhaps non at all) have considered the full implications of the second -- that
language isn't primarily
representational. Certainly Marx and Engels failed to do this, and so have
subsequent
Marxists. Indeed, much of what they have written on this topic -- especially about
'abstraction', 'cognition' and knowledge -- suggests that the
opposite is in fact the case.
In this regard once more, dialecticians aren't alone. Until recently, little critical attention has
been paid to the traditional view that language
is primarily representational, i.e., that it enables
human beings to re-present the 'objective' world in "thought", the "head", the "mind",
"consciousness", or "cognition" first, before communication can begin.
The
underlying assumption has rarely been questioned (again until recently): that
is, that it is only after language users have learnt to picture reality
to themselves that they are then able to communicate their thoughts to others, and that observation applies equally well to those who at least
give lip service to the idea that the primarily role of language lies in communication. This means that, despite what
they might say, the social nature of language is
seen by the vast majority of Marxists as a consequence of the isolated (but later pooled) cognitive
resources of individuals, an expression of their attempt to share the 'contents' of
their 'minds', their 'abstractions',with their listeners, not the other way round.
It seems to many (even on the revolutionary left) that here at least
we have an example of private (mental) production
coupled with public gain, for on this view, it is the isolated activity of lone abstractors
that
powers cognition, and this supposedly helps drive the social advancement
of knowledge after these abstractions have somehow been pooled, or
shared....
This approach thus relegates meaning to the private
domain of the 'mind', something that each individual brings to
language --, perhaps as an expression of their own biographies or the
ideological parameters that constrain them....
As we saw in Essay Three
Part Two, post-Renaissance
thinkers (Rationalists and Empiricists alike) took the public domain (where meaning is created), inverted it, and
then projected
it back into each individual skull, re-configured there as the
social
relations among ideas, or 'concepts'!
This resulted in the systematic
fetishisation of language and thought, leading to the conflation of the
'objective' world with the subjective contents of the 'mind'. ["Fetishised",
since, as we will see, words are seen as agents, here.] The outer, social
world was thus re-located in each individual head, the latter seen as primary, the former as secondary
(or non-existent, in some cases!). In this way, the social was
privatised, internalised, and hence neutralised. No wonder then that
modern philosophy soon descended into full-blown Idealism, with Kant complaining that
it was scandal that philosophers had so far failed to prove the existence of the 'external' world!
Small
wonder, too, that Dialectical Marxists felt they had to invert things once more --
allegedly putting them 'back on their feet' -- all the while failing to
note that their theory of language and cognition actually prevents them from doing
precisely that.
More recently, this ruling-class thought-form has re-surfaced in
several new
disguises: sometimes as the inter-relationship between neurons (as they
'communicate' with one another), supposedly controlled by the oppressive
power of the gene -- which now seems to operate as a sort of surrogate inner
Bourgeois
Legislative and Executive Authority --; sometimes as computational device (or at
least a device that helps 'the mind' write or use the 'software').
Given this view, while human beings might be born free of language, everywhere
they
are imprisoned by linguistic chains manufactured and controlled by an
inner surrogate 'state' (comprised of genes, 'modules', or assorted 'neural
nets')....
In fact, this is one ideological inversion that has remained
upside down (but in different forms), not just for hundreds but for thousands
of years, and which is largely the source of the other 'inverted ideas'
cobbled-together by Traditional Philosophers and dialecticians alike. Inverted now, as in a camera obscura, these rotated
concepts cloud
the thoughts of all those whose brains they have colonised -- which, of course, helps
explain why the ideas of the ruling-class always rule.
[More details can be found in Essay Twelve
Part One,
here.]
Dave McNally,
for example, had this to say in his highly selective summary of
Voloshinov's ideas:
"The very first pages of
Marxism and the Philosophy of Language are organized around the opposition
between natural phenomena and the world of signs. Natural bodies do not signify,
they do not take on special meanings by reference to something else, says
Voloshinov. A tool is a tool, a loaf of bread is a loaf of bread. Each is
self-identical, a natural object, not a sign. Human reality is this twofold,
consisting of natural phenomena and signs: 'side by side with the natural
phenomena, with the equipment of technology, and with articles of consumption
there exists a special world -- the world of signs' [Voloshinov (1973),
pp.9-10.] Now, as a point of departure this would be a helpful distinction. It
would enable Voloshinov to use the distinction natural phenomena/sign to point
out that signs do not occur in a purely natural way. To acquire meaning they
must be socially organised to that end, they must become part of a sign system.
But from here, it would be necessary to show how 'natural' phenomena -- the
material structure of the world, biological history, the human larynx, voice,
hands, and so on -- are essential to human sign systems. It would be necessary,
in other words, to show that signifying activity is conditioned by the
'natural,' just as the human-natural world is socially mediated." [McNally
(2001), p.119. Quotation marks altered to conform with the conventions adopted at
this site. Spelling adapted to UK English; italic emphases in the
original.]
But, don't we say things like "Those clouds
are a sign of rain", or "Those footprints are a sign there are deer in the
vicinity". To be sure, they are signs for us, but weren't clouds a sign of rain
before we evolved?
Be this as it may, McNally goes on to point
out that this distinction is problematic since it is not only far too rigorous,
it is static, as well. [McNally, it seems, prefers a 'dialectical' relationship, here.]
Even so, he still maintains the view that for Voloshinov:
"[S]igns...are social not
individual; a fully formed psychic life is a social one." [Ibid., p.120.]
However, as we have seen (here
and here), while Voloshinov gestures
at a social understanding of language and signs, his view of both rapidly
lapses into a traditional view that language is an individual and
representational phenomenon.
[I will add several more
comments about McNally's work here in a later re-write of this Essay.]
Holborow has similarly been
led astray by another of Voloshinov's inappropriate metaphors: the idea that "signs
are the means of mental processing" (ibid., p.31).
On signs, see Note 23, and
Note 85. [See also Stroud (1991, 1996).]
61a. Some might object and claim they can repeat the same thought
without using these or any other words! However, little content can be given to
that particular counterclaim until far more details are forthcoming. And good
luck to anyone who attempts to fill those details in without using any language.
62.
That is, unless we acknowledge the existence of 'languageless' thoughts, once
more. [On
this, see Note 1.]
It
would take a brave Marxist to go down that route, for it would amount to yet
another damaging concession to Dualism. [On that, see
here, as well as Note 66 and Note 86.]
As, Parrington himself pointed out:
"Language, therefore, is not
just an expression of otherwise independent and fully formed thought, but rather
is a necessary form of the thought's realisation." [Parrington (1997),
p.135.]
Even so, as we saw above, this still leaves it unclear whether
there can be 'pre-linguistic thoughts'. As far as can be ascertained, Parrington
doesn't rule this out, and neither does Voloshinov. [On this, see Note 61.]
62a0. There is a detailed description of how
metaphors and pictures like these have misled Traditional Philosophers in Fischer
(2011a, 2011b).
63.
I.e., in the sense that there are no round squares -- that is, it makes no sense to suppose
there are any, not
that it is false to say that there are. On the other hand, if it were merely false to say there were
such things as round squares, then
it would make sense to suppose that it could be true there were round
squares.
Plainly, that would run together what should be matters of fact with what
conceptually confused sentences. [Why this is so is explored at length in Essay Twelve
Part One. Again, on this, see
Note 67.]
64.
Any more than there is a gap between 'Satan' and 'God', or between Big Foot and
the Tooth Fairy, for science to study.
This might seem to suggest that the issue here is empirical, when it isn't;
on that, see
Note 63.
65. The idea that there could be linguistic
expressions -- or representational surrogates of them -- in our heads falls
foul of the modified Devitt objection
outlined earlier.
Of course, 'consciousness' itself is the
product of yet another reification of language.
However, in this case, this reification produces the spurious 'philosophical
problem' that some declare is among the
hardest problems science has yet to solve. But, and once again, there is
no 'problem' of 'consciousness' that isn't itself the product of the above
linguistic reification (the origin of which was explored
here, and in Note 1).
If we
examine how we
actually use this word (and its associated terms) in everyday life, it is easy to see that
"consciousness" normally only ever
appears in medical or forensic contexts:
C1: "Has the patient regained
consciousness?"
C2: "It's at this point that
he lost consciousness."
Neither of these even remotely suggests an
inner process, one that is incapable of mystifying any but the terminally naive. Both
sentences suggest there are behavioural criteria for the application of the
word, "consciousness".
Is it even possible to regain an
'inner process'? Or to lose one?
[It is also important to distinguish
the verb form of this word -- "conscious" -- from the nominal form --
"consciousness".]
I am,
however, aware that the above comments could
provoke far more hostility
from some quarters than practically anything else in this Essay; but that just
shows to what extent this ancient ruling-class idea
has colonised most minds. Again, as Anthony Kenny, Hilary Putnam and Francisco
Garzón noted (quoted earlier):
"Descartes view of the nature
of mind endured much longer than his view of matter. Indeed among educated
people in the West who were not professional philosophers it is still the most
widespread view of the mind. Most contemporary philosophers would disown
Cartesian dualism but even those who explicitly renounce it are often profoundly
influenced by it. Many people, for instance,
go along with Descartes in identifying the mental realm as the realm of
consciousness. They think of consciousness as an object of introspection; as
something we see when we look within ourselves. They think of it as an
inessential, contingent matter that consciousness has an expression in speech
and behaviour. Consciousness, as they conceive it, is something to which each of
us has direct access in our own case. Others, by contrast, can only infer to our
conscious states by accepting our testimony or making causal inferences from our
physical behaviour." [Kenny (1992), p.2. Bold emphases added; paragraphs
merged.]
"...[V]irtually no
philosopher doubted, from the time of
Locke
until roughly 1914, that, whatever concepts and ideas were, they were clearly
mental objects of some kind. And no large-scale and comprehensive demolition job
was done against this particularly wide-spread and influential philosophical
misconception until
Wittgenstein
produced his
Philosophical Investigations...." [Putnam (1975b), p.7.]
"The
notions of computation
and representation are not just common currency in cognitive science modelling.
To put it mildly, they are the building blocks of the discipline.
Alternative voices from a number of subdisciplines that call into question these
notions have periodically been raised. Unfortunately, after an initial, and
usually short, excitement they remain quiet. Silence is due mainly to two
reasons. On the one hand, the dominant
paradigmoverwhelms competitors
(sometimes due to 'pragmatic' considerations) with data already accounted for
and results to be accounted for, and on the other hand, alternative framings are
repeatedly absorbed and made innocuous. Both reasons are interrelated.
Alternatives raised, by default, carry the burden of proof in such a way thatthe dominant paradigm is the one that chooses what phenomena are in need
of explanation.... Problems start when the what limits the range
of options available when it comes to answering the how." [Garzón (2008),
pp.259-60. Quotation marks altered to conform with the conventions adopted at this site. Bold emphases alone added.]
On this, see the references
given in Note 86, below --
but, particularly Hacker (2012, 2013a);
I have reproduced the opening two pages of the latter,
here.
Also see
Note 66 and Note 69.
66.
This might seem to some to be an example of the present author "poisoning the wells";
that is, warning others off raising questions concerning the material basis of thought
with a mixture of scare tactics and
"keep off" signs. However, the main argument in this part
of the Essay revolves around the suggestion that this entire topic is a
pseudo-problem, predicated on an ancient, mystical dogma
that we are all in effect just embodied souls, that there are
two different features of our make-up that need explaining (or, even, that there is a 'gap'
that has opened up between these two aspects which needs bridging) -- the 'Mind',
or 'Consciousness', and the Body -- which
doctrine itself is the result of centuries of
mysticism, linguistic
reification and
fetishisation.
If,
however, there is no good reason for
materialists to accept the dogma that there are two 'realms' -- an inner,
hidden, 'mental' world, and an 'outer', physical world of events or processes --, then it is
surely pertinent to ask where this notion
came from and why the vast majority of Marxists have given it credence.
Indeed,
it is in just this area that the links between Traditional Dogma and
DM become all the clearer, and which need exposing as a result.
We have already seen that this approach to 'mind'
and 'consciousness' is based on the Platonic-Christian-Cartesian view of
'mind' and body. As
noted above, dialecticians might be
tempted to
reject this allegation out-of-hand, but the fact is that even though they
propose their own distinct solution to this 'problem', they have obviously bought into the
world-view Descartes bequeathed to 'western' thought, along with its obscure jargon and
incoherent terms-of-reference.
[On this, see also here and
here.] But, this entire problematic is based
solely on linguistic confusion and Mystical Platonism; that being so,
there is in fact no problem to address. In turn, this means that we
don't need a 'theory' to 'solve' it.
Were it not for the deleterious effect that
DM has had on the thinking of revolutionaries (super-gluing them to the Cartesian and
Representational view of language and 'mind', compounded by
what is in fact a
metaphysical view of
'reality'), an investigation into the material roots of thought would rightly have
been situated in the open, in a public arena -- where half-way
decent
HM-theorists should
have had their attention focussed anyway.
67.
Of course, this isn't to deny that the
word "thought" can function as a noun. On this, see Note 61.
"[P]hilosophical problems arise when
language goes on holiday."
68a.
It is no surprise then that from this heady brew there
has also emerged (of late) 'rational economic man'
(beloved of 'Rational
Choice Marxists'), a fictional character who
behaves just like the atom of thought that originally went into 'his' creation.
Marxists in general won't be unaware of the rest of this
sorry tale; it is outlined
here and
here for any who
might have allowed themselves to forget.
68b.
As we will see, these moves also
had a profound effect on
Marxist theory itself. [More on this below, and in Essay Twelve Part Seven. Until
that is published, see
here.]
69.Ruling-class ideology has in fact succeeded in doing this,
and to such an
extent that much of the material in this
Essay will horrify and repel many comrades -- at least
many of those who have
made it this far --, or, rather, those who have become used to viewing 'the mind'
through Cartesian spectacles -- so used to
it in fact that much, if not all of this Essay, will be rejected out-of-hand.
After all, how else could ruling ideas always
rule other than by making radically alternative, if not politically dangerous
ideas (like those presented
here) seempatently false? Or, by biasing the debate to such an extent
that the language employed is loaded with Platonic, Christian or Cartesian
jargon, concepts, inappropriate metaphors and misleading analogies, trapping radical minds inside this
dominant Paradigm?
I have quoted
this several times, but it is worth repeating:
"The
notions of computation
and representation are not just common currency in cognitive science modelling.
To put it mildly, they are the building blocks of the discipline.
Alternative voices from a number of subdisciplines that call into question these
notions have periodically been raised. Unfortunately, after an initial, and
usually short, excitement they remain quiet. Silence is due mainly to two
reasons. On the one hand, the dominant
paradigmoverwhelms competitors
(sometimes due to 'pragmatic' considerations) with data already accounted for
and results to be accounted for, and on the other hand, alternative framings are
repeatedly absorbed and made innocuous. Both reasons are interrelated.
Alternatives raised, by default, carry the burden of proof in such a way thatthe dominant paradigm is the one that chooses what phenomena are in need
of explanation.... Problems start when the what limits the range
of options available when it comes to answering the how." [Garzón (2008),
pp.259-60. Quotation marks altered to conform with the conventions adopted at this site. Bold emphases alone added.]
On the other hand, if an attempt is ever made
by Traditional Theorists (or dialecticians!) to substantiate these Neo-Cartesian
doctrines, then that task is often regarded as an
annoying distraction,
required only because of the objections advanced by those of us who have
allegedly adopted an 'anti-scientific' or 'anti-philosophical' bias -- that is, because of the qualms
expressed by those of
us who have the temerity to question and then reject the Platonic-Christian-Cartesian view of
the 'Mind'.
[On this, see the quotation from Button
et al,posted earlier.] That
reaction in turn is because this Paradigm has sunk so deep into the collective
Traditional and Dialectical 'Mind' that it is now unquestioned dogma -- a genuine
"ruling idea".
"The ideas of the ruling class are in every epoch
the ruling ideas, i.e. the class which is the ruling material force of society,
is at the same time its ruling intellectual force. The class which has the means
of material production at its disposal, has control at the same time over the
means of mental production, so that thereby, generally speaking, the ideas of
those who lack the means of mental production are subject to it. The ruling
ideas are nothing more than the ideal expression of the dominant material
relationships, the dominant material relationships grasped as ideas; hence of
the relationships which make the one class the ruling one, therefore, the ideas
of its dominance. The individuals composing the ruling class possess among other
things consciousness, and therefore think. Insofar, therefore, as they rule as a
class and determine the extent and compass of an epoch, it is self-evident that
they do this in its whole range, hence among other things rule also as thinkers,
as producers of ideas, and regulate the production and distribution of the ideas
of their age: thus their ideas are the ruling ideas of the epoch...." [Marx and Engels (1970), pp.64-65, quoted from
here. Bold
emphases added.]
Notice how Marx
pointed out that:
"The class which has the means
of material production at its disposal, has control at the same time over the
means of mental production, so that thereby, generally speaking, the ideas of
those who lack the means of mental production are subject to it.... Insofar,
therefore, as they rule as a class and determine the extent and compass of an
epoch, it is self-evident that they do this in its whole range, hence
among other things rule also as thinkers, as producers of ideas, and regulate
the production and distribution of the ideas of their age...."
[Ibid. Bold emphases added.]
They also
"rule as thinkers, as
producers of ideas, and regulate the production and distribution of the ideas of
their age", and they do so in "its whole range". In which case, the individuals
who were later to become leading revolutionaries (but who had been "subject to"
the full force of this indoctrination
before they became Marxists), can't fail to have had their thinking shaped by the ideas
and thought-forms of the
ruling-class, chief among which is the Platonic-Christian-Cartesian Paradigm.
Which is, of course, why Lenin thought it quite natural to look to
the work of previous thinkers as precursors of the concepts we find in DM, and
its 'materialist theory of mind':
"The history of philosophy and the history of
social science show with perfect clarity that there is nothing resembling
'sectarianism' in Marxism, in the sense of its being a hidebound, petrified
doctrine, a doctrine which arose away from the high road of the
development of world civilisation. On the contrary, the genius of Marx consists
precisely in his having furnished answers to questions already raised by the
foremost minds of mankind. His doctrine emerged as the direct and immediate
continuation of the teachings of the greatest representatives of
philosophy, political economy and socialism.
"The Marxist doctrine is omnipotent because it is true. It is comprehensive
and harmonious, and provides men with an integral world outlook irreconcilable
with any form of superstition, reaction, or defence of bourgeois oppression. It
is the legitimate successor to the best that man produced in the nineteenth
century, as represented by German philosophy, English political economy and
French socialism." [Lenin,
Three Sources and Component Parts of Marxism. Bold emphases alone
added.]
Notice, DM is
"successor" and "a continuation of the
teachings of the greatest representatives of philosophy".
A "continuation"
of, not a break from!
69a.
Indeed, comrades were forewarned by Marx himself about
the source of
ideological distortions such as these:
"For philosophers, one of the
most difficult tasks is to descend from the world of thought to the actual
world.
Language is the immediate actuality of thought. Just as philosophers have
given thought an independent existence, so they had to make language into an
independent realm. This is the secret of philosophical language, in which
thoughts in the form of words have their own content....
"The philosophers
have only to dissolve their language into the ordinary language, from which it
is abstracted, in order to recognise it, as the distorted language of the actual
world, and to realise that neither thoughts nor language in themselves form a
realm of their own, that they are only manifestations of actual life."
[Marx
and Engels (1970), p.118. Bold emphases added.]
"The ideas of the ruling
class are in every epoch the ruling ideas, i.e. the class which is the
ruling material force of society, is at the same time its ruling intellectual
force. The class which has the means of material production at its disposal, has
control at the same time over the means of mental production, so that thereby,
generally speaking, the ideas of those who lack the means of mental production
are subject to it. The ruling ideas are nothing more than the ideal
expression of the dominant material relationships, the dominant material
relationships grasped as ideas; hence of the relationships which make the
one class the ruling one, therefore, the ideas of its dominance. The individuals
composing the ruling class possess among other things consciousness, and
therefore think. Insofar, therefore, as they rule as a class and determine the
extent and compass of an epoch, it is self-evident that they do this in its
whole range, hence among other things rule also as thinkers, as producers of
ideas, and regulate the production and distribution of the ideas of their age:
thus their ideas are the ruling ideas of the epoch....'" [Ibid.,
pp.64-65. Bold emphases added.]
So, there is no excuse!
70.
This last comment is connected with other issues raised in this Essay. One of
these concerns the observation that ruling-class "prize-fighters" -- in most
cases, these used to be Traditional Philosophers, but, more recently they are often scientists and
other 'opinion-formers' who try their hand at a little amateur 'philosophising' -- display a class-motivated tendency to distort, depreciate
or misuse the
vernacular. This they do partly because it is the only way that they
can spin their metaphysical fairy tales (or "cobwebs"
as
Francis Bacon called them), and partly because it is traditional to do so.
[On this, see Essay Twelve Part Seven (summary
here).]
However, this tactic comes with a price-tag
attached:
it renders such 'theories' non-sensicaland incoherent. As Marx himself noted, that is because this approach has to distort
language in order for it even to seem to work. To compound the
problem, figurative discourse
has to be interpreted (apparently) literally in order to facilitate the creation of a
string
of spurious philosophical 'problems'. The complex way that ordinary language
actually functions has to be ignored, altered or (failing that)
fetishised.
This well-trodden path predisposes those who yet to break with traditional
ways of 'philosophising', into accepting the idea
that general words operate as the names of 'abstract'
particulars, concepts or ideas.
Hence, "thought" names 'Thought', "mind" names 'Mind', "to be aware" names
'Consciousness', and so on. Unfortunately, these moves turn sentences into
lists of names, preventing them from saying anything at all. The
pseudo-propositions produced as a result are now incapable of expressing a
sense (true or false), and are thus non-sensical. Distortion of language also
renders them incoherent.
[These allegations were substantiated in detail
in Essays
Three
Part One and Twelve Part One.
We saw in the former Essay that lists say nothing at all, unless they are
articulated with words that don't function as names.]
71.Below, I itemise the many studies that
present a more satisfactory analysis of what we all already know about
thought (Voloshinov and John Parrington included).
72.
Again, it is worth noting that I don't prefer this use of language. It is only
being employed to assist in its demise.
73. Freud's lack of originality was briefly
examined in an earlier Essay.
Nevertheless, it is a little puzzling to see Parrington refer his readers to
Richard Webster's book on Freud [Parrington (1997),
Note 117, p.149] when Webster himself demonstrated conclusively that Freud
fabricated the evidence that John now says Freud used to prove there is
a link between the 'unconscious' and human sexual behaviour (etc.)! [Cf., Webster
(1995).]
According to Webster (and he isn't the only
one who alleges this), Freud was a
complete charlatan.
Hence, this reference to Freud provides Parrington with about as much genuine support as the
average
WMD dossier
lent to Bush and Blair. [See also Note 74.]
Oddly enough, McNally also seems to have a
high opinion of Freud. [Cf., McNally (2001, 2004).]
74.
The fact that Freud didn't discover the "Unconscious" (as Parrington
still seems to believe) has been uncontroversial among Freud scholars for
several generations. [Cf., Ellenberger (1970), and Whyte (1959).]
The philosophical
difficulties associated with this rather vague notion were outlined in Bouveresse
(1995). [There is more on Freud's
fraudulent ideas in Essay Nine
Part One.]
As Bouveresse notes:
"It is less frequently
noticed, however, that [Freud's] vision of consciousness remained utterly
traditional and bound to the idea of consciousness as the internal perception of
'objects' of a certain type -- the paradigm of clear and immediate perception."
[Bouveresse (1995), p.22. Quotation marks altered to conform with the conventions adopted at this site.]
75.
Except, Parrington offers little or no textual evidence for this fanciful
re-definition of "sense"/"theme". In addition, his identification of "meaning"
with the dictionary definition would have surprised Voloshinov.
Despite this, what Parrington goes on to say about "sense" indicates that he
understands the latter to be different from "theme" in several important
respects, a fact that anyone reading both accounts may readily confirm. [On
this, see Note 77.]
Nevertheless, I can find no reference to
"sense" in Voloshinov's book. The pages to which Parrington himself refers
actually contain Voloshinov's discussion of "theme". This suggests that he has
translated Voloshinov's word for "theme" into his own for "sense" --
while
leaving both terms equally obscure!
76.
This peculiar term-of-art (i.e., "values") has been put in 'scare' quotes since
it is widely used in the social sciences (and elsewhere; vote hungry politicians
are particularly fond of employing it, and it now seems to be a permanent feature of
management-speak)
as a catch-all term for such diverse things as beliefs, moral priorities,
preferences, social mores, attitudes and feelings, along with a host of other
trendy terms. As such, it isn't just hopelessly vague, it is
philosophically useless.
Its appearance in Moral Philosophy is as misguided as it is hackneyed.
'Values' have nothing whatsoever to do with the meaning of the words we use in
moral discourse. A much more secure platform from which to launch an
investigation in this area can be found in von Wright (1963). However, I don't propose to develop
those controversial observations any further at this site.
77.
Of course, this all sits rather awkwardly with what Voloshinov himself says
about "theme" (that is, if the latter is indeed the same as Parrington's "sense"):
"Let us agree to call the
entity which becomes the object of a sign the theme of the sign. Each
fully fledged sign has its theme. And so every verbal performance has its
theme…."
"A definite and unitary
meaning, a unitary significance, is a property belonging to any utterance as
a whole. Let us call the significance of a whole utterance its theme….
"The theme is the
expression of the concrete, historical situation that engendered the utterance.
The utterance 'What time is it?' has a different meaning each time it is used,
and hence, in accordance with our terminology, has a different theme, depending
on the concrete historical situation ('historical' here in microscopic
dimensions) during which it is enunciated and of which, in essence, it is a
part….
"It follows, then, that the
theme of an utterance is determined not only by the linguistic forms that
comprise it -- words, morphological and syntactic structures, sounds, and
intonation -- but also by extraverbal factors of the situation. Should we miss
these situational factors, we would be as little able to understand an utterance
as if we were to miss its most important words. The theme of an utterance is
concrete -- as concrete as the historical instant to which the utterance
belongs. Only an utterance taken in its full, concrete scope as an historical
phenomenon possesses a theme. That is what is meant by the theme of an
utterance….
"Theme is a complex,
dynamic system of signs that attempts to be adequate to a given instant of
generative process. Theme is reaction by the consciousness in its generative
process to the generative process of existence." [Voloshinov (1973), pp.22,
99-100. Bold emphases added. Italic emphases in the original.]
We saw
earlier that Voloshinov had an
insecure grasp of his own distinction between "theme" and meaning,
but the above passage seems to divorce "theme" from the sort of idiosyncratic
"feelings" and "associations" that Parrington mentioned -- especially the parts
that have been highlighted in bold. This isn't to say that the above can't be
made consistent with Parrington words, the point is: why bother? Most of Voloshinov's
ideas are so confused that incorporating them into any form of Marxist linguistics would be rather like trying to marry Das Kapital with Lewis Carroll's
Jabberwocky.
78.
Naturally, this means that words that have never been encountered before by a
given individual would have no meaning/"sense" until that person had formed
their own idiosyncratic "sense"/"tone"-for-them in relation
to it. In that case, when such a
person encounters, say, the word "Bogomil"
for the first time they wouldn't
need to have its meaning explained to them (which, of course, would be a useless
ceremony, anyway, given Parrington's theory), they would just need to form, or
generate, some feelings (etc.) about it.
Indeed, anyone who agrees with Parrington, and who doesn't know what "Bogomil"
means, shouldn't follow the link I posted above; they should instead simply
conjure up a
few feelings about this word and all will be become clear!
Naturally, if this
were so, dictionary makers would surely go out of business, to be replaced
by iterant 'feeling-generators' who would make the meaning of words
comprehensible to everyone by…, er, well what? Slaps in the face? Sad stories?
Jokes?
On the other hand, unscrupulous government
agents could manipulate unsuspecting citizens by inventing new words (making
sure that the 'right' feelings and associations had been engendered in each
case in each victim,
1984-style),
in addition to re-defining older words by altering the feelings (etc.) they
once generated (perhaps by the use of drugs, induced pain, or other forms of
aversion
therapy).
It
could be objected that this gets Parrington's view of meaning completely wrong.
In fact, it isn't easy to discern a clear or consistent view of meaning in his
article. For example, on the one hand he says:
"A related feature of inner
speech that Voloshinov pointed to was that it is more concerned with 'sense'
rather than 'meaning'. In this definition, meaning is the dictionary definition
of a word, for instance cat: 'a furry domestic quadruped'. Sense, on the other
hand, refers to the whole set of psychological events aroused by a word, such as
the personal memories of your own pet and its mannerisms, the feel of its fur
and so on. It contains activities, impressions and personal meanings, not just
accepted social definitions. A word acquires its sense from the context in which
it appears; in different contexts, it changes its sense." [Parrington
(1997), p.135. Bold emphasis added.]
Which
sees meaning (not "sense") as relatively fixed. On the other, he says the
following:
"Voloshinov's aim was an
ambitious one. He believed that a truly comprehensive theory of language should
include in its description all the factors outside words that affect their
meaning. Voloshinov pointed out that the only way that words can have meaning
is through being understood. But words are understood by particular speakers
and listeners, who are also speakers, in particular situations occurring in the
real world. Voloshinov argued that by conceiving words as if no one ever spoke
them, other linguists had missed a fundamental aspect of language, which is that
'the speaker's focus of attention is orientated by the particular, concrete
utterance he is making...what is important to him is not that a word is a
stable and always self-equivalent signal, but an always changeable and adaptable
sign'....
"Thus there is a gap between
thought and words but words remain the only possible way of expressing ourselves
fully to others. Thought which remains unexpressed remains immature and
eventually dies out. Language, therefore, is not just an expression of otherwise
independent and fully formed thought, but rather is a necessary form of the
thought's realisation. In summary, inner speech is the fluid interphase where
meaning can start to be formed and shaped, based on the emotional, practical and
social experience of the individual. However, this still leaves us with the
question of the concrete mechanisms whereby this process takes place." [Ibid.,
pp.126, 135-36. Bold emphases added.]
The above plainly sees meaning as variable between
speakers, and perhaps even between occasions of use -- which certainly was
Voloshinov's view. Why say the following if meanings were somehow fixed?
"Voloshinov pointed out that the
only way that words can have meaning
is through being understood. But words are understood by particular speakers
and listeners, who are also speakers, in particular situations occurring in the
real world....
In summary, inner speech
is the fluid interphase where meaning can start to be formed and shaped, based
on the emotional, practical and social experience of the individual." [Ibid.
Bold emphases added.]
In the above, meaning isn't fixed
publicly, but is established 'internally' and individually, by means of "inner
speech...where meaning can start to be formed and shaped, based on the
emotional, practical and social experience of the individual". This is a view of
language which would, of course, make communication impossible.
This sits rather awkwardly with the following
comment of Voloshinov's:
"Meaning does not reside
in the word or in the soul of the speaker or in the soul of the listener.
Meaning is the effect of interaction between speaker and listener produced
via the material of a particular sound complex." [Voloshinov (1973) pp.102-03.
Italic
emphasis in the original.]
However, we have also
seen
that Voloshinov himself is equivocal on this issue, since several other things he says,
and the above comments from Parrington, place both of them in the
traditional camp -- where meaning is a
subjective feature of an individual's "understanding", which is itself
a private 'mental' act, dependent on circumstances.
So, the argument presented here doesn't
constitute an obvious
misrepresentation of Parrington's views. Of course, it is entirely possible that
I have misconstrued what he wanted to say --, but then, he simply invites
it!
78a.Some readers might object here that there is
no way to determine whether or not we all use words the same way, or that we
communicate successfully. The problem is, of course, that the one advancing this objection will have to
use words in the same way as the rest of us or fail to make his/her point -- which,
naturally,
wouldn't then be worth making.
It could be countered that those advancing
the above objection might be
using just those words (or just some words) the same way as the rest of us to
make the point that we do not use all our words the same way.
However, it isn't
essential for communication to work that we all use every single word in
exactly the same way (which position isn't being defended here, anyway --
for example, there is, what
Hilary
Putnam has called a "linguistic
division of labour" (this links to a PDF)) --,
only that communication would fail if we didn't use words in the same way, typically -- or, if there were no way to resolve incidental confusions or failures in
understanding.
Hence, this (hypothetical) objector would now need to indicate which
words he or she feels are being used in different ways by the rest of us, all
the while using other words just as the rest of us do to make that point -- or fail to make
it, once more.
This sceptical approach to word use
can't be extended to cover the
usual employment of words -- and for the same reason --, otherwise the objector will fail to
communicate his/her point, again.
As we saw
earlier, if there
is a moral tale here, it is that it isn't possible to undermine the ordinary use of language while trying to make
sense, or trying to communicate with the rest of us.
It
might be wondered whether we do in
fact communicate successfully with one another. However, anyone foolish enough
to raise that point will have to withdrawn it on pain of not
having communicated it to the rest of us, if what they had just alleged were itself correct.
It
could be objected that we might think we understand one another when we
don't actually manage to do so. Well, if I understand that objection
correctly, that would be sufficient to refute it. On the other hand, if I
don't, then there is nothing to answer. And the same two options face
anyone rash enough to question whether we typically understand one another, or
even whether we do so at all. So, if this hypothetical objector understands
these two options then their objection fails, too. Alternatively, if they
don't, then they haven't sufficiently thought through this problem, and they
should come back when they have.
Someone could now complain that the above response misses the point. The latter concerned the alleged failure to communicate, not this:
X1:
We only think we communicate, but the fact is that we don't.
But,
the above response showed that the second half of X1 self-destructs. In that
case, the first half of X1 merely expresses a bare possibility without
presenting evidence or argument in its favour. So, my response didn't miss the
point.
Is
there anything that can be said for or against such a 'bare possibility'? If
anyone wants to defend it, they should send
me their best shot; I certainly won't do their work for them. Having said
that it is clear that this 'bare possibility' is on a par with the 'bare
possibility' that you, dear reader, are really an
aardvark.
If the possibility that we fail to communicate has been shown to self-destruct
then we can place a probability of zero on this 'bare possibility' being the
case, just as we can do the same with the possibility that you, dear reader, are
indeed an aardvark.
78b.
It could be argued that this assumes what
was to be proved, that we communicate successfully to begin with. However, as
pointed out in Note 78a, anyone who understands that objection has
automatically
refuted it.
For example, Aitchison (1996), Chapter
Two, argues in favour of the 'self-refuting conclusion' -- ironically, hoping to get her point across successfully,
nonetheless!
79.
Is
this what passes for analysis outside of Analytic Philosophy these
days!?
80.
This was examined in detail
earlier in
this
Essay.
80a.
On
ellipsis and
one-word sentences, see Stainton (2006a). [However, the reader mustn't assume
that I agree with everything in Stainton's book; the reverse is in fact the
case. I don't propose to pull on that particular thread here, though.]
81.
This idea is repeated in Sean Doherty's recent article; cf., Doherty (2000),
pp.155-56.
82.
Holborow uses a slightly different spelling of Voloshinov's name in her book
(viz., "Volosinov"). The one found in this passage has therefore been adapted to
the conventions apparent in Parrington's article.
Incidentally, as noted earlier, and with respect to language, this would place Voloshinov in the
representational, not the social/communicational, camp:
"A sign does not simply exist
as a part of reality -- it reflects and refracts another reality. Therefore it
may distort that reality or be true to it, or it may perceive it from a special
point of view…." [Voloshinov (1973),
p.10.]
Indeed, this
appears to fetishise signs: they "reflect and refract...reality". But, as
Wittgenstein noted, signs are as lifeless
as rocks and stones. Exactly how it is possible for one lifeless object to say anything about other
objects is, therefore, entirely mysterious. Attaching the word "intentionality"
to signs (as many Philosophers attempt to do) is no help at all, since that is
just another lifeless 'sign'. On this, see Stroud (1991, 1996).
[See also
this parable.]
83.
The word "refraction"
rather gives the game away one feels, since refracted light is responsible for
changing or
misrepresenting shapes. [To be sure, not every refracting system distorts,
but then that is because corrections have had to be made.] On this, see Note 84.
Also see Hacking (1983).
84.
If there is nothing to "distort" then plainly none can have taken place (in
this respect).
It could be argued that Voloshinov doesn't mean by his
R/R metaphor that things are distorted.
However, he removed all doubt
when he said the following:
"A sign does not simply exist
as a part of reality -- it reflects and refracts another reality. Therefore it
may distort that reality or be true to it, or it may perceive it from a special
point of view…." [Voloshinov (1973),
p.10. Bold emphasis added.]
Do words really do this? If they do, how
might we ever tell? Any
attempt to do so will have to use words! Which is an indirect way of saying that
words don't, or can't, ordinarily do this -- and that isn't just because words are also part of the world, but because
any attempt to state this alleged fact would itself be a distortion!
Indeed, how could anyone tell whether words were
"true" to reality if everything were ideological? In fact, in Voloshinov's system,
there are only 'signs' confronting other 'signs'. Given this theory, even
direct observation of the world aimed at checking the veracity of our 'signs' must
also be mediated by 'signs', so there is in fact no way we could compare our 'signs'
with an unmediated world to see which were and which weren't true to it!
Indeed, for all we know, those pesky signs
could be distorting our reading of Voloshinov, too!
[The idea that
everything social is ideological will be examined in Essay Three Part Four (where views expressed in Lecercle (2006) will
also be addressed).]
However, as far as Marx is concerned, it is
abstraction and the misuse (not the use) of language that creates distortion:
"The philosophers have only
to dissolve their language into the ordinary language, from which it is
abstracted, in order to recognise it, as the distorted language of the actual
world, and to realise that neither thoughts nor language in themselves form a
realm of their own, that they are only manifestations of actual life."
[Marx
and Engels (1970), p.118. Bold emphases added.]
No hint here that
all words are
'ideological'.
It could be argued that Voloshinov was
committed to the view that words only sometimes distort things:
"A sign does not simply exist
as a part of reality -- it reflects and refracts another reality. Therefore it
may distort that reality or be true to it, or it may perceive it from a special
point of view…." [Voloshinov (1973),
p.10. Bold emphasis added.]
However, how anyone might discriminate among
these suspicious 'signs' is left unclear. Which signs are reliable and which
aren't? Can we even trust the 'signs' in this sentence: "Voloshinov is committed
to the view that words only sometimes distort things"? Perhaps they do
this all the time. Maybe they never do it. [I return to this issue in the main
body of the Essay.]
This reminds one of what Rorty had to say
about those signs that do and those that don't misrepresent 'reality':
"Pragmatic Wittgensteinians
tend to be historicist in their metaphilosophical views. They think that the
problems of pre-Kantian metaphysics, the problems that the naturalists [this is
Rorty's term for those who reject Wittgenstein's approach to philosophy -- RL]
have revived, are hangovers from a particular moment in Western intellectual
history. These problems originate not in a clash between common sense and
science, but rather between the immaterialist notions that Christian theology
had inherited from Plato and Aristotle and the mechanistic and materialist
world-picture sketched by Galileo and Newton. That clash was between
metaphysical outlooks, not between metaphysics and a pre-metaphysical
understanding of things.
"This clash produced the
Cartesian notion of ideas as appearances on the stage of an inner theatre, as
well as the Lockean account of words as signs of such ideas. More generally, it
produced a picture of knowledge as the attempt to acquire accurate mental
representations of non-mental reality. Representationalist accounts of the
relation between language and non-language emerged from the attempt to divide
language into assertions that represent real things and those that do not. On
this historicist view, Wittgenstein's importance lies in his having helped
wrench us out of our Cartesian-Lockean mind-set." [Rorty (2010), p.132. Readers
should also note the caveats concerning Rorty's work I have posted
here.]
In short, Voloshinov has backed himself into
philosophical corner because of his unwise acceptance of the
dominant Cartesian Paradigm.
84a.
I do not wish to enter into a discussion of
ideology here; that will be
postponed until Essay Three Part Four. However, see Note 84,
above.
85.
Many of the comments in this Note are tentative, at best. I may have to
revise them extensively.
This isn't to deny that signs (or signals) might not be coded against words, so
that when translated (or decoded) they could be expressed in sentences, etc.
But, even then it is the use of language
that gives life to signs, not the other way round.
Nor is it to deny that signs work in many
different ways. For
example, words can work as signs, say, outside shops or on commodities.
But, this no more means that words are signs than the fact that paper money can
be burnt means it is fuel. Or, no more than the fact that a copy of theMorning Starcan be used to make a paper hat means that The Morning
Star
is a paper hat.
However, from the things Voloshinov says, it
isn't at all clear that he means the same by the word "sign" as we might do in
English, or even in the way that Holborow seems to understand it. Hence, in what
follows, I will be using the word largely as English speakers would normally understand
it. [The pay-off here is that in so doing I can expose a serious confusion lying
at the heart of
Semiotics, where the use of this word is
as ubiquitous as it is problematic.]
Indeed, there is considerable confusion among philosophers over
the use of the word "sign" (including Voloshinov himself, as we will soon see), where it appears to mean anything that human beings might use,
manufacture, draw or co-opt in order
to signify or mean something beyond the sign itself, ranging from human
artefacts to natural portents. [This appears to be the
semiotic
sense of this word, too.]
But,
this erodes the distinction we would
normally want to draw between signs and symbols, signs and pictures, and signs
and natural portents (as in, "Those clouds are a sign of rain", "Red
spots are a sign of measles"). The semiotic
sense of this word also seems to hold that the function of signs is to designate
('reflect', 'mean', 'refer to', 'indicate') something 'outside of itself' (i.e.,
acting as a
"signifier" of the "signified" -- to use the buzz-words --
eliding
the distinction between "signify" and "refer to", "signifier"/"sign" and "name",
"signified" and "object"), which, if this were true of words,
would mean that all words were (directly or indirectly) referential in
some way, which manifestly isn't
the case. For example, what does "if" name or refer to ('outside'
itself)? Or, "nothing"? [On that,
see below.]
In short, Semiotics seems to be little
other than a systematic capitulation to what has come to be known as The
Augustinian Fallacy:
"These words [Wittgenstein is here
quoting Augustine -- RL], it seems to me,
give a particular picture of the essence of human language. It is this: the
words in a language name objects -- sentences are combinations of
such names. -- In this picture of language we find the roots of the following
idea: Every word has a meaning. This meaning is correlated with the word. It is
the object for which the word stands....
"It is important to note that
it is a
solecism
to use the word 'meaning' to signify the thing
that 'corresponds' to a word. That is to confound the meaning of a name with
the bearer of the name. When Mr. N. N. dies, one says that the bearer of
the name dies, not that the meaning dies." [Wittgenstein (2009), §1, p.5e, and §40,
p.24e. (This links to a PDF.) Italic emphasis in the original; Quotation marks altered to
conform with
the conventions
adopted at this site.
Link added.]
Odd though this might seem, and as I have
pointed out here,
it is plausible to argue that Wittgenstein was prompted in this direction
(directly or indirectly) by Voloshinov himself:
"Meaning is a function of the
sign and is therefore inconceivable...outside the sign as some particular,
independently existing thing. It would be just as absurd to maintain such a
notion as to take the meaning of the word 'horse' to be this particular, live
animal I am pointing to. Why if that were so, then I could claim, for instance,
that having eaten an apple, I have consumed not an apple but the meaning of the
word 'apple'." [Voloshinov (1973), p.28.]
[On this,
also see Essay Twelve
Part One, where I
summarise Wittgenstein's objection to this way of picturing words. Cf.,
Baker and Hacker (2005a), pp.43-91, and Baker and Hacker (2005b), pp.1-28.
However,
the best article on this is Robinson (2003c).]
Someone might object and claim that "nothing"
is in fact the name of the most abstract of all concepts -- following Hegel:
"Being, pure being,
without any further determination. In its
indeterminate immediacy it is equal only to itself. It is also not unequal
relatively to an other; it has no diversity within itself nor any with a
reference outwards. It would not be held fast in its purity if it contained any
determination or content which could be distinguished in it or by which it could
be distinguished from an other. It is pure indeterminateness and emptiness.
There is nothing to be intuited in it, if one can speak here of
intuiting; or, it is only this pure intuiting itself. Just as little is anything
to be thought in it, or it is equally only this empty thinking. Being, the
indeterminate immediate, is in fact nothing, and neither more nor less
than nothing.
"Nothing, pure
nothing:
it is simply equality with itself, complete
emptiness, absence of all determination and content -- undifferentiatedness in
itself. In so far as intuiting or thinking can be mentioned here, it counts as a
distinction whether something or nothing is intuited or thought. To
intuit or think nothing has, therefore, a meaning; both are distinguished and
thus nothing is (exists) in our intuiting or thinking; or rather it is
empty intuition and thought itself, and the same empty intuition or thought as
pure being. Nothing is, therefore, the same determination, or
rather absence of determination, and thus altogether the same as, pure being."
[Hegel
(1999), p.82.Italic emphases in the original.]
I will
examine this particular 'argument' in
more detail in Essay Twelve Part Five, but for present purposes it is worth
pointing out that "nothing" appears to function as the name of "being"
-- or they both name the same 'entity', "complete emptiness, absence of all
determination and content". But, it can't
be a name or operate as a name (or name-like word), nor can it be what something is
(as Hegel seems to believe), nor does it refer to a lack or
a privation.
First of all, it can't be the name of a
material object (or process) -- for obvious reasons (even
Parmenides saw
through that
one!) --, in which case, "nothing" can't have been
derived by means of reflection from, or reflection on or about, anything material, and
thus from any sort of
abstraction from it.
[To be sure, the above rather hastily drawn conclusion runs contrary to several attempts that have been made to explain how 'Nothing'
and 'Being' can be derived by a 'process of abstraction'. Those 'attempts' will
also be picked apart in Essay Twelve Part Five.]
Secondly, if "nothing" were a name (or even a
"sign" for any one thing, or even a lack of anything), then the following inferences would work. Compare
the legitimate deduction that employs names in H1-H3 with the inference represented in H4-H6, which uses
"nothing" as if it were a name.
H6: Therefore, something is
bigger than the universe.
[Remember, H4 and H5 don't have to be true
for the inference to be
valid! All that is required is that if the premises were true, the
conclusion would be true, too.]
H1-H3 works because of the proper names it
employs; H4-H6 fails because "something" and "nothing" aren't names
(nor do they refer to anything) -- otherwise,
of course, H4-H6 would have been a valid inference. "Something" and
"nothing" are
quantifiers.
[Anyone who thinks differently will have to
accept the validity H4-H6 (as well as H10-H12 and H13-H15, below), and hang
the consequences.]
If we 'semiotify' H4-H6, we obtain the
following:
H4a: Something is bigger than
whatever "nothing" signifies.
H5a: Whatever "nothing"
signifies is bigger than
the universe.
H6a: Therefore, something is
bigger than the universe.
This shows why "nothing" can't "signify"
anything at all (in the semiotic sense) -- otherwise this would be a valid inference, too.
The same 'difficulties' occur if we try to
treat "nobody" and "somebody" as names, or even as
"signifiers". Compare the following legitimate
inference using names, with the one following it, which uses "nobody" as
if it were a name:
H7: NN lives on Tenth Avenue.
H8: NM lives with NN.
H9: Therefore, NM lives on
Tenth Avenue.
Compare the
above with the following:
H7a: NN lives on Tenth Avenue.
H8a: Nobody lives with NN.
H9a: Therefore, Nobody lives on
Tenth Avenue.
And with:
H10: Nobody lives on the
Moon.
H11: Somebody lives with
nobody.
H12: Therefore, somebody
lives on the moon.
Also compare
it with its 'semiotified'
version:
H10a: Whatever "nobody"
signifies lives on the
Moon.
H11a: Somebody lives with
whatever "nobody" signifies.
H12a: Therefore, somebody
lives on the moon.
H7-H9
is a safe inference; the same can't be said of H10-H12 and H10a-H12a.
And neither is the following:
H13: Nothing travels faster
than light.
H14: Light travels faster
than a tortoise.
H15: Therefore, nothing
travels faster than a tortoise.
If words were signs (or if they "signified"
their "signifiers") then the mode of
signification of names (and other singular terms) would be the same as that of
quantifiers like "nothing" and "nobody" (that is,
they would both relate to whatever it was they "signified" in the same way -- there
would thus be no difference in this respect between names and quantifiers), and the above
inferences would all be safe.
Nothing can't express a lack or a privation,
either, otherwise it would be operating as a name of..., er, well, nothing,
and we have just seen it can't even do that!
H16:
Tony Cliff
left nothing out of his speech on the former USSR.
H17: The accused removed nothing
from his bag, which he then left in a locker.
H18: Your solution to that
mathematical puzzle was nothing short of amazing.
H19: The comrade from the
floor added nothing to the opening speech.
Are any of these privations? If any of
the above "signified" a lack or a privation, then they would be no different
from names.
Of course,
jokes about "nothing" and "nobody"
have been around for thousands of years -- in fact, at least since
Homer's day.
Indeed, what is apparently the first recorded joke in history runs along the
lines illustrated in the next quoted passage. The latter is part of a dialogue
found in
Homer's
Odyssey, where
Odysseus
tricks one of the
Cyclopes (Polyphemus)
into thinking his (i.e., Odysseus's) name is "Nobody". Here is Odysseus speaking
to the giant:
"'Nobody -- that's my
name. Nobody -- so my mother and father call me….' But he boomed back at me from
his ruthless heart, 'Nobody? I'll eat Nobody last of all his friends….'"
[Homer
(2003) Book 9.410-9.413, p.223. The Internet
version has "Noman" instead of "Nobody". The translation I have used is also
markedly different from the on-line version.]
Later, Polyphemus falls asleep, whereupon he
is attacked and blinded by Odysseus. He cries out to his fellow giants for help
and they respond by asking who is attacking him:
"'What, Polyphemus, what in
the world's the trouble…? Surely no one's trying to kill you by fraud or force!'
"'If you are alone,'
his friends boomed back at once, 'and nobody's trying to overpower you…it must
be a plague sent here by mighty
Zeus….'
"They lumbered off, but
laughter filled my heart to think that nobody's name -- my great cunning stroke
-- had duped them one and all." [Ibid.,
Book
9.450-9.463, p.224.]
It seems, therefore, that metaphysicians
(like Hegel) are as easily fooled as one-eyed giants.
This shows that Homer was aware of the
distinction between names and quantifiers nearly three thousand years ago! He
might not have been able to say precisely what that distinction was, or even put
this idea into a clear and coherent form. But, his joke tells us he was aware
that names and quantifiers don't work in the same way. As Wittgenstein noted:
"The problems arising through
a misinterpretation of our forms of language have the character of depth.
They are deep disquietudes; their roots are as deep in us as are the forms of
our language and their significance is as great as the importance of our
language -- Let us ask ourselves: why do we feel a grammatical joke to be
deep? (And that is what the depth of philosophy is.)" [Wittgenstein (1958),
§111, p.47e. Italic emphasis in the original.]
Wittgenstein is also reported to have said that a
serious work on philosophy could be written that consisted entirely of jokes.
[On this, see Pitcher (1978), p.317. In fact, Pitcher's article is an attempt to
do just that.]
More modern versions of the same sort of joke
include the following:
J3: Union militant:
"Management are ruthless. They'll stop at nothing."
J4: Union Bureaucrat: "In that case, we'll do nothing!"
If J4 were to be read in a
dialectical sort of way, it might provide Trade Union Leaders with an
excellent extra argument for official inertia, which would only work
on the rest of us if we were naïve enough to believe "nothing" is indeed a name.
African
Stereotype: "I
suggest we do nothing.... If we do nothing, perhaps he'll stop at it."
Update 20/01/2018: Since the above was first written a video of this
'comedy' has been posted on DailyMotion. This clearly isn't one of
John Cleese's
finest moments; it is in places definitely guilty of crude racial
stereotyping and the sort of 'soft racism' that was deemed 'acceptable' in the
1970s. The comedy itself is rather poor, too:
Video Seven: The Strange
Death Of..., Nothing
A few
years ago,
Mark Steel
and Pat Stack both published examples of this age-old gag:
"Most economists agree that
loans should only be advanced if Russia accepts 'austerity measures'. But
thousands of workers have gone six months without being paid. How do you get
more austere than that? Will
Yeltsin
address the nation thus: 'There are some miners earning as much as nothing. To
you I say this, you're pricing yourselves out of a job. For too long you have
insisted on getting nothing for something. The country can no longer afford
it.'" [Mark Steel in The Guardian, 02/09/1998.]
"'Four thousand
two hundred and seven, that's the exact number,' the King said, referring to his
book. 'I couldn't send all the horses, you know, because two of them are wanted
in the game. And I haven't sent the two Messengers, either. They're both gone to
the town. Just look along the road, and tell me if you can see either of them.'
"'I see nobody
on the road,' said Alice.
"'I only wish I
had such eyes,' the King remarked in a fretful tone. 'To be able to see Nobody!
And at that distance, too! Why, it's as much as I can do to see real people, by
this light!'....
"At this moment
the Messenger arrived: he was far too much out of breath to say a word, and
could only wave his hands about, and make the most fearful faces at the poor
King....
"'Who did you
pass on the road?' the King went on, holding out his hand to the Messenger for
some more hay.
"'Nobody,' said
the Messenger.
"'Quite right,'
said the King: 'this young lady saw him too. So of course Nobody walks slower
than you.'
"'I do my best,'
the Messenger said in a sulky tone. 'I'm sure nobody walks much faster than I
do!'
"'He can't do
that,' said the King, 'or else he'd have been here first. However, now you've
got your breath, you may tell us what's happened in the town.'" [Alice Through
the Looking Glass, quoted from
here.]
[There are several more examples like this in Pitcher
(1978). See also,
here.]
Again, such jokes only
work if "nothing" and "nobody" are viewed as names, or as parts of simple
predicate expressions. However, the
logical
role of words like "nothing" and "something" is far more complex.
[On
quantifiers,
see Dummett (1981a), pp.8-33, Haack (1978), pp.39-55, Westerathåhl (2001),
and Peters and Westerathåhl (2008).
More on this in Essay Twelve Part Five (when it is published).]
Nevertheless, it is unclear in which sense
Parrington and Holborow intended to use the word "sign" -- in an ordinary or in
a semiotic sense. Indeed, Voloshinov's own attempt to define his use of this
word leaves much to be desired:
"Any ideological product is
not only itself a part of reality (natural or social), just as is any physical
body, any instrument of production, or any product of consumption, it also, in
contradistinction to these other phenomena, reflects and refracts another
reality outside itself. Everything ideological possesses meaning: it
represents, depicts or stands for something lying outside itself. In other words
it is a sign. Without signs there is no ideology. A physical body
equals itself, so to speak; it does not signify anything but wholly coincides
with its particular, given nature. In this case there is no question of
ideology.
"However, any physical body
may be perceived as an image; for instance, the image of natural inertia and
necessity embodied in that particular thing. Any artistic-symbolic image to
which a particular physical object gives rise is already an ideological product.
The physical object is converted into a sign. Without ceasing to be a part of
material reality, such an object, to some degree, reflects and refracts another
reality." [Voloshinov (1973), p.9. Italic emphases in the original.]
Unfortunately, Voloshinov neglected to say what "image" we have of nothing,
or how the word or 'sign', "nothing", can "reflect" or "refract"
anything at all. Indeed, if it reflects nothing, it isn't doing its job properly, since signs are supposed
to reflect something, not nothing. On the other hand, if it reflects something,
then it must have reflected the wrong target: something rather than nothing!
[Astute
readers will no doubt have noticed that this piece of sophistry only works if we
again treat "nothing" as a name of, or as a 'sign' for, something. Imagine trying to
clear up this particular confusion with yet more signs! But, with words it is relatively
easy.]
Voloshinov then gives several other examples of objects that can and have been turned
into 'signs', ranging from religious artefacts to political emblems and
consumables. He adds:
"Thus, side by side with the
natural phenomena, with the equipment of technology, and with articles for
consumption, there exists a special world -- the world of signs.
"Signs are particular,
material things; and, as we have seen, any item of nature, technology, or
consumption can become a sign, acquiring in the process a meaning that goes
beyond its particularity. A sign does not simply exist
as a part of reality -- it reflects and refracts another reality. Therefore it
may distort that reality or be true to it, or it may perceive it from a special
point of view. Every sign is subject to the criteria of ideological evaluation
(i.e., whether it is true, false, correct, fair, good, etc.). The domain of
ideology coincides with the domain of signs. They equate with one another.
Wherever a sign is present, ideology is present, too. Everything ideological
possesses semiotic value." [Ibid., p.10. Italic emphases in the
original.]
So, it looks like Voloshinov is employing a
semiotic interpretation of 'signs'
after all (maybe
one he derived from
C S Pierce).
[Merrell (2001).] So, for him a 'sign' can be anything that is used to represent
something else -- even though Voloshinov adds that 'signs' can
"reflect"/"refract" [or R/R] reality, too. [More will be said about this in Essay Twelve and in
a later re-write of this Essay where I will consider Lecercle's commentary
on Voloshinov (i.e., in Lecercle (2006)).]
However, as noted above, it is also apparent that the
'semiotic' sense of this word is itself based on the 'Augustinian
Fallacy', exposed by Wittgenstein
-- that is, the idea that all words 'stand for' something, and operate as Proper
Names or other singular terms. [On this, see
Glock (1996), pp.41-45, Baker and
Hacker (2005a), pp.1-28, Baker and Hacker (2005b), pp.43-93, and
Ryle (1949b)
(this links to a PDF).
However, these should be read in conjunction with Carruthers (1984). See also
these comments, below.]
Now,
there
seems to be nothing inherently wrong with labelling words 'signs', but the temptation
is always there to regard them as signs in the ordinary (but now
fetishised) sense that,
(i) Their meaning lies on their face (since they now tell
us what they mean -- implying they are the agents while we are the patients);
and,
(ii) Their meaning is
what they stand for, transforming them into Proper Names or
other singular terms.
From what we have seen,
anddespite what they might say, Voloshinov and those that take their lead
from him
appear to have fallen into both traps.
In
what follows, I focus on the use of "sign" when it is used semiotically
and when
it is used more colloquially to refer to road signs, shop signs,
etc.
Here are several of the reasons why words aren't
signsper se
-- in the sense just indicated:
(1)
Clauses can be
negated, signs can't, nor can a collection of signs.
[This idea is connected
with an important insight in Wittgenstein's Tractatus [i.e., Wittgenstein
(1972)] that logical words can't stand for anything (they don't picture
anything), they simply enable the articulation of expressions that can picture
a possible state of affairs. More on that in White (2006), Palmer (2011) and
McGuinness (1974). That insight motivates much of the
discussion here,
here and
here. The reader is directed
there for more details.]
To be sure, signs can be
crossed out, torn up, ignored, flouted, painted over, destroyed, obliterated,
even disobeyed -- but not negated. But, can't someone do the opposite of what a
sign says, thus negating it? For example, couldn't someone move their car into a
space in a road that has "Keep Clear" painted on it? They certainly can, but the
sign still says "Keep Clear"; the sign itself hasn't been negated. [Signs
that have the negative particle emblazoned on them will be considered
presently.]
A sign can be employed as a negative
particle (as can be seen from sign languages), but here it is the use of a
sign that serves in this role. The content of a series of signs can, of course,
be negated -- but only if it has been put in a propositional or other sentential
context, etc.; but a sign itself can't. Once more, a sign can represent 'negatives' (as in a
"No parking", "No Entry" or "No U-turn" sign, if these are
conventional wordless signs, not those that might also be emblazoned with
their own translation into language), but that is done my means of yet
another sign, not the negation of an earlier one. [On that, see
here.]
However, even here, the negative particle is part of language, and so depends on
language. In that case, the use of language will explain the sign, not the other
way round. [I return to this below,
where I examine examples of falsehoods supposedly represented by signs.]
Admittedly, Highway Agencies employ special "No Entry" and "No U-turn" signs,
which can have something written or printed on them that has been crossed out. But, short of that being
part of a special code (or even a
rebus), crossing out isn't negation.
Negation typically turns truths into falsehoods, or vice versa. Crossing a sentence out doesn't do this, it merely
removes it from the field. [Naturally, there could be a rule that stipulates
that crossing an indicative sentence out negates that sentence, but in such
circumstances, it is the rule, not the crossing out that achieves this, and
rules aren't the sort of things that can be true or false.] Crossing a sign out might
very well turn it into another sign;
it wouldn't automatically remove it from play. Hence, a crossed out
picture of a cigarette is still a sign that smoking is banned in the vicinity,
not that another sign has been negated, banned or decommissioned.
Admittedly, signs can be crossed out
if they are no longer in use, but that would be apparent from the make-shift
nature of the crossing out; in which case, that crossing out won't be part of the sign in the
sense that this is:
Figure Eleven: No Smoking
Sign
The impromptu sort of crossing
out mentioned above will in fact amount to
another sign being imposed on the original sign, removing it from the field, as it
were.
Which
is why this was asserted above:
Crossing a sign out might
turn it into another sign; it wouldn't automatically remove it from play. Hence, a crossed out
picture of a cigarette is still a sign that smoking is banned in the vicinity,
not that another sign has been negated, banned or decommissioned.
Note the use of "automatically".
This is a different sort of crossing out.
Furthermore, the double negation of a
sentence reinstates its original status -- as in "It's not not raining". The
double tearing up or crossing out of a sign certainly does not restore the
original; it just creates smaller pieces, more confusion, or perhaps a greater mess.
Moreover,
quantified clauses and sentences can be internally or externally negated. For
example, "It is not the case that everyone is a member of the
CWI" does not mean the same as "Everyone is not a member of the CWI". The
first simply means that there are some (perhaps many) who do not belong to the
CWI; the second means that no one does. This flexibility doesn't exist with signs (unless,
once more, they are part of a code of some sort and can be translated into, and are thus parasitic
upon, certain already fixed/settled words or sentences). Or, to put this another way, signs can't represent 'negative situations': e.g., that it is not Friday. We
may infer such things from signs, but that is what we, as language users, bring to
them. As Jerry Fodor notes (if we
ignore for the moment his reference to 'the mind', and note his use of "symbol"
in place of "sign"):
"[I]f
the mind is in the inference-drawing line of work, there must be symbols in
which it formulates its premises and conclusions; there are no inferences
without a medium (or media) in which to couch them. That matters because you
can't say just anything you like in whatever kind of symbols you choose.
Pictures can't express negative or contingent propositions -- [such as] it's not
raining, or if it's raining that will spoil the picnic. But negative and
conditional thoughts play a central role in the kinds of inference that minds
routinely carry out. ([For example] it's certainly not Queen Victoria; if it's
certainly not Queen Victoria, then perhaps it's Dr Livingstone. So perhaps it's
Dr Livingstone.) Such considerations suggest, at a minimum, that the mind
doesn't do all its thinking in pictures." [Fodor (2003), p.16, paragraph 6.]
(2) Word order affects meaning: "Red flag"
means one thing, but "Flag red" means something else. The change of word order
here turns a noun into a verb. You can swap signs around all day long and this
won't happen -- unless, once more, they are part of a pre-arranged code,
picture
puzzle or
picture code (all of which, of course, depend on language).
Hence, language use is largely governed by
grammatical (i.e., socially-sanctioned) rules that enable the articulation of
words into clauses or sentences, any change to which will alter the sense of what
is, has been, or could be said by means of them. So (to use Chomsky's example), the phrase "Pretty little girl's school"
can be read in many different ways, depending on how it is parsed. But, signs
are not parsed into grammatical categories (except, again, as part of a code,
and, once more, codes depend on language, not the other way round), nor are there
similar rules for their employment (except in restricted areas of use, like those dictated by
Highway Authorities, for instance).
It could be argued that a sign for red
followed by a sign for a flag will convey a different message from an
arrangement where they are reversed, so that a flag sign is followed by a red sign. But, this difference is
based on translating those signs into words. It won't work the other way round.
For instance, without the use of any words to interpret it, how would a flag
sign indicate that it was a verb when it appears first, as opposed to a noun
when it is used second? Words typically function differently when they are nouns
and when they are verbs, but if they were merely signs, that wouldn't happen (and
for the above reason).
[If we absolutely have to call words
anything (but why isn't the word "word" good enough?) we could call written
words inscriptions (indeed, as I have done so in many places at this
site), or call spoken words vocalisations. This would at
least obviate many of the problems associated with "sign" outlined in this section.]
The
articulation of words permits, for example, the
formation of indicative sentences; this enables the latter to be
understood before their truth-values
are known. In this way, speakers can converse about falsehoods,
possibilities, events and objectsthat aren't in
their immediate vicinity, or which might even have ceased to exist -- or,
maybe,
which might one day start to exist. These could even involve remote regions
of space and time, non-existent objects, or fictional characters and situations.
Signs can't do this (unless, once more, they are part of a code, and hence
depend on language, again). A road sign (in the UK) with a black diagonal band on
a white background (conventionally) signals only what is on its face (or,
rather, we take it to signal this), which
might be something like "Maximum speed limit 60/70 mph", in the immediate
surroundings, or between that sign and other speed signs.
Figure
Twelve:
UK 'Max Speed 60 Mph' (On Most Roads) Sign
On a waste tip,
hospital corridor, or in a living room,
the same sign would be useless. Such signs are therefore location-sensitive;
not so with language,
per se. [I have dealt with context-dependencyelsewhere in this Essay.]
Moreover, without the use of linguistic
expressions (in or on a sign), falsehoods can't be depicted, as, indeed, Fodor noted.
It could be objected to this that certain
paintings manage to represent falsehoods: think of
Surrealist
and
Dadaist works
(like those pained by
Magritte,
for example).
First of all, only indicative sentences/clauses
are capable of being true or false.
[Note the use of "only" here as opposed to "all".] Second, quite apart from the fact that paintings aren't signs
either, the interpretation of a work of art is no more a given than is anything
else. Even Magritte's famous painting of a pipe relies on a use of the French
words "Ceci n'est pas une pipe" ["This isn't a pipe."] to state what many take to be a
falsehood.
Figure Thirteen: Magritte's
Play On Words
However, this could of course be seen as a
truth since a picture of a pipe is manifestly not a pipe! [Even that
assumes that these words refer to the pipe and not to the words themselves.]
Nevertheless, this
picture still has to use words arranged in an indicative sentence to achieve its aim.
If a painting is to be taken to represent a falsehood,
that falsehood would have to be expressed by an indicative sentence which
gives voice to an appropriate interpretation (of that picture) with a negatable
content, or it wouldn't be a falsehood (given what that word
means). One could not, for example, imagine another
sign (as opposed to another sentence) interpreting the meaning
of a work of art. But, that would be possible if words were signs.
But,
what about a picture of, say,
Sherlock
Holmes, or a sign that featured this fictional character?
Figure Fourteen: Holmes On
The Case?
Once more, as with pictures, such signs
have to be interpreted. So, a sign with Holmes on it could mean anything from "Conan
Doyle lived in this building" to "Shop here for detective novels." If
such a picture or sign
is to mean anything in particular, then it will have to be
conventionally associated with
whatever that sentence or object happened to be. [I am using "mean" itself
rather loosely here to mean(!) "allude to", "express", or even "refer to".] In which case, the sign would stand for that
specific object or sentence, and
no other. So, and once more, such a sign would only work because of a
pre-existing language in which it could be interpreted or decoded -- as the case may be
--
but not the other way
round. If so, such
signs and pictures can't account for language, they are parasitic upon
it.
It
could be argued that words also have to be interpreted. Maybe so, but this can't be so with all words, or the word "interpreted" would suffer a
rather unenviable fate (i.e., it would be locked into a vicious circle). Anyway, even
assuming
this were the case, any further interpretation of a word or set of words would
have to be expressed by the use of
yet more words. So, once again, a pre-existing language is indispensable, even here.
Again, it could be argued that falsehoods can in fact be depicted in or by
signs. Hence, one could imagine a road sign pointing in the wrong direction, a
"Closed" sign hung on the door of a shop that is still open, an "Out-of-order"
sign attached to the door of a working lift, and so on. But, with
respect to the last two examples, the falsehoods (if such they be) will be expressed
by certain words written on each sign. Without these the signs wouldn't
be able to state any falsehoods. Clearly, in the second case, "Out-of-work" would be
elliptical
for "This lift is out of order"; similarly with the "Closed" sign,
which stands for "This shop is closed/isn't open". [Once more, we see that indicative sentences
are implied here, too.]
As far as the road
sign is concerned, its status would depend on whether or not there were any words
written or printed on it. Hence, if a sign saying "Exeter 20 miles" was pointing in the wrong
direction, or if Exeter were really 25 miles away, this falsehood would plainly
be language dependent, too. In that case, we would be interpreting these words as
"Exeter is 20 miles in this direction" (etc.) -- again, using yet another indicative sentence. Of course, such signs are
indexical, that is they contain an implicitly indexical term. For
example, "Exeter 20 miles" is in effect short for "Exeter is 20 miles from here".
In that case, they resemble type sentences, such as "Cash today, cheques
tomorrow" (which is an example that
Peter Geach
used to employ). In that case, their import once more depends of a sophisticated use of
language, which can't be represented by further signs that aren't already
symbols.
[Recall that the point being made here isn't
that signs can't 'say' things but that whatever they do express will be
dependent on a pre-existing language or on linguistic conventions, not the other way round.]
It could be objected that if a "Max
Speed 60 mph" sign (in the UK, this is a circular sign with a black diagonal band on a
white background when placed on single carriageway roads, "70 mph" on dual carriageway roads
--
see
Figure
Four,
above) was either accidentally or intentionally placed outside a village school (where the sign should have
been "20 mph"),
then, even though it had no words on
it, that sign would be false.
But, such a sign expresses a command or
instruction so it can't be false, since, plainly, no command or
instruction is capable of being either true or false, only practical or impractical,
useful or useless, obeyed
or disobeyed, etc.Anyway, a sign like this outside a school, say, would
cause consternation since it indicates that the speed limit is 70 mph (on a
dual carriageway, 60 mph otherwise). If this were a mistake, the sign wouldn't be
false (let alone would it have been negated), just incorrectly located.
And we could say that with some confidence because the sign still indicates that
the speed limit is 60/70 mph, which is how anyone who knew road traffic law
would be able to tell it had been incorrectly located.
Finally, if such a sign
could be false, then it could also be true.
But, given UK law, it can't be true.
Figure Fifteen: UK 'No Entry' Sign
Of course, there is nothing in the above
to indicate that the present author thinks that signs can't be
incorrect (think of a "No Entry" sign (in the UK this is a white oblong on a red
background; see
Figure Twelve) placed at the entry
to a road -- but, even then, as with the speed sign we met earlier, it might
simply have been mis-located), but these would only state falsehoods when
translated into indicative sentences, or, because they can be so translated. In such
cases, as already noted, signs like this are a conventional shorthand for a
particular sentence or phrase (and they are such because of the standardised
interpretation we have been instructed, or socialised, to apply to them or read
from them -- perhaps as depicted in a
Road Manual or Highway Code). Hence, and once again,
such signs would only work because of a pre-existing language, and so
can't
account for language.
The idea that words are signs is perhaps
further motivated by (a) the fact that signs sometimes contain words, phrases or clauses
(which, of course,
can be negated), or (b) the fact that they can have negated sentences, or
sentence fragments painted, printed or written on them -- as in "No to student loans!", or "Don't Attack Iraq!"
Figure
Sixteen: A Sign Bush And Blair
Recklessly Ignored
But, not all signs contain words
-- and here the distinction between words and signs is clear.
Indeed, if words were signs (in the ordinary
sense) they would
fail to work, except they had other words printed on them --, unless, once
more, they expressed a conventionalised shorthand for a particular sentence.
Of course, a sign with nothing on it could be
a sign of something quite incidental (for example, that the printer had run out
of ink, or the designer of ideas), but words with nothing written on them still
manage to communicate -- but what they communicate isn't something over and above what they mean (saving
complications introduced by conversational or representational
implicatures
and prosody, noted earlier, and rehearsed again, below).
For example, sentences like the following:
S1: My other car is a
Porsche,
usually mean (i.e.,
linguistic meaning) what
they say, but they could just as well mean (i.e., speaker's meaning, or by
conversational or coded implication) that the owner of a
particular car is actually ashamed of the car he/she owns and is lamenting that fact.
Figure
Seventeen: Poseur -- Or Porsche
Owner?
Alternatively, it could be a sign that the
driver is, say, a spy (if S1 were code for another sentence/message), or, and
far more likely, that
he/she is just a poseur (etc.). Even so, the content of whatever is
communicated over and above what the words themselves actually say will still be
linguistically motivated, and that can't be occasion-sensitive or meaning
would be forever indeterminate --
as we have seen. Indeed, if that were so, we wouldn't
be able to grasp the content of any sentence that asserted even that
hypothetical shortcoming! [On
that, see Note 15.]
Signs are objects in their own right. Hence,
it is worth asking: If words are signs, how is it possible for one
object to say anything about another object? Voloshinov seems to think
signs (which are objects of a certain sort --, or any sort, for him) can easily do this, as we have seen (but, surely, that would be to
fetishise them). Of course, signs only depict whatever they do because
of conventions we set up for them, in use. [But, this has already been covered,
above and in
Note 23.]
It could be objected that signs
differ from naturally occurring objects in that they are human inventions, or artefacts, that
work in a different way -- perhaps, to reflect reality and to "depict or stand
for something outside of itself". But, even then it would be the use
of a sign that gives it its meaning. [Again, on this, see Note 23.]
However, even supposing Voloshinov meant by
"sign" something semiotic, as opposed to its more ordinary sense, that wouldn't help, since no sign (natural, semiotic, man/woman-made, picture, artefact,
etc.) can determine a rule. It is our rule-governed use of certain symbols that brings life to
language. [Note 23, once more!] To
suppose otherwise would be to fetishise signs (again, as noted
earlier).
Even more
importantly, a collection of signs (semiotic, natural or conventional), no
matter how they were arranged -- if they all 'signify' the 'signified' -- would simply
become a list, and lists say nothing. [On that, see
here.]
86.
Unfortunately, Holborow, Parrington and Doherty appear not to have consulted
any work in the Philosophy of Language written in the last 100 years -- or
perhaps
ever! Their references and bibliographies suggest that they haven't checked the work of a single
Philosopher of Language (that is, save the work of Linguists and Psychologists
who have dabbled in a little amateur 'philosophising'), or at least not one that was
written this century (other than Voloshinov, etc.). This is a pity since it
means they have repeated many of the errors committed by previous generations of
theorists. It seems that, like those ignorant of history, those ignorant of the
History of the Philosophy of Language are, alas, condemned to repeat past
indiscretions. [Indeed, condemned to resurrect theories of language little
different from touted by those classic bourgeois individualists, Hobbes,
Locke, Descartes and Hume.]
The same could be said of Collins, but
maybe less so of McNally. Both confine their attention largely to work written by other Marxist authors
or fellow travellers -- Vygotsky, Voloshinov, the Bakhtin circle, and
Gramsci --, but neither of them quote or reference any work done by Analytic
Philosophers over the last hundred and thirty odd years.
As
noted, this is less true of McNally; he gave an
intelligent talk at Marxism 1990 (an annual gathering of the UK-SWP) about Analytic Philosophy. Unfortunately,
McNally largely ignored the Frege-Wittgenstein tradition in the Philosophy of
Language, something a supporter of this site had to point out to him at the time. Although he acknowledged this omission in his reply, he seems not to have
incorporated any of its insights (concerning the nature of language and
thought) in his subsequent work -- McNally (1995, 2001, 2004). In fact, just
like the comrades mentioned above, the many confusions that litter McNally's
thinking (in this area) are perhaps a direct result of this omission.
[I will say more
about the above in a later re-write of this Essay.]
In contradistinction to the above, Lecercle
shows he is at least aware of
Austin's
work, as well as that of
Tarski,
Grice,
Searle,
and Wittgenstein, along with several other Analytic Philosophers. Having said
that, his book shows precious little sign he has learnt much from them, or is in
any way ready, or qualified, to engage with them. However,
and as if to compound the problem, he appears to have ignored totally the work of
Frege,
Russell,
Quine,
Carnap,
Davidson,
Ryle,
Strawson,
Geach,
Anscombe,
Kripke,
Lewis,
Donellan,
and
Dummett, to name but a few Analytic Philosophers of Language. Indeed, one of the main problems with
Lecercle's work is that he devotes far too much space to Stalin's ideas
-- yes, that expert, Stalin --
about language. In fact,
one paragraph devoted to Stalin would have been too much. As if to
compound this further, Lecercle is largely uncritical of Voloshinov's vague and confused ideas.
[I will say more about this,
too, in a future re-write of this
Essay.]
It could be objected that Marxist authors are
quite right to concentrate solely on what other Marxists have said about
language, ignoring the work of assorted bourgeois theorists.
In response, it is
worth making the following four points:
(1) Not one of the above Marxists
would argue this way about the thoroughly bourgeois Philosopher, Hegel. [The
usual reply -- that Hegel was writing at a time when the bourgeoisie were the
revolutionary class -- has been neutralised,
here.]
Nor, indeed, about assorted Continental Philosophers, like those mentioned
here.
(2)
As pointed out in Essay Twelve Part One, and as underlined throughout this
Essay, while Marxist theorists have acknowledged the social nature of language
and thought, all they have succeeded in doing (in their work) is undermine this fundamental Marxist insight, just
as they have ignored Marx's own abandonment of Philosophy (since, among other
things, he thought it based on a distortion of ordinary language, and amounted
to little more than religion rendered into theory ). [On that, see
here.]
(3) The Frege-Wittgenstein tradition in
Analytic Philosophy manages to do the opposite -- it underlines and defends
a
commitment to the social nature of language and thought. That alone should
recommend it to Marxists -- even though it won't, since its seems that the majority
are hide-bound
by their traditional
approach to Philosophy.
(4) Most of the above comrades underline the
links between "theory", language, practice, working class struggle and working class culture
(howsoever the latter is characterised), but not oneof them considers how
working people
actually use language, or how it is used in everyday life. If they were to do that, every single
doctrine they have inherited from Traditional Thought (and more specifically from
Hegel) would be rejected outright as
incoherent non-sense. That can't be said of the Wittgensteinian method employed in
these Essays.
Despite countless knee-jerk and -- one
is almost tempted to say -- disingenuous references to "practice" regularly intoned
by DM-aficionados (particularly those belonging to the
HCD-tendency),
these individuals regularly adopt theories that they would never
be able to sell to workers in practice, theories they themselves would find
impossible to translate into action, and which would spell disaster if
anyone even attempted to apply them in the class war. [Not that this comment,
or even those untoward prospects, will slow them down by so much as one
millimetre per second. As my old Professor of Logic used to say, their
heads are too "full of noise".] It is precisely here, in
practice, that
nostrums like these become all the more obviously crazy -- that is, should anyone be
foolish enough to try them out on workers, or test them in everyday life.
~~~~~~oOo~~~~~~
On the distinction between meaning in
language and 'speaker's meaning', and the difference between signs and
words (both of which Parrington, Holborow and McNally -- following Lenin -- equate
without any attempt at justification), see Harrison (1972, 1979), Lycan (2008),
and Platts (1997). See also Bickerton (2009).
In addition to the above,
there are now many general introductory
philosophical accounts of language and edited collections of papers (modern
classics and recent reviews), which manage to avoid the mistakes that
Voloshinov and other comrades commit, among which are the following: Black (1968),
Blackburn (1984),
Garcia-Carpíntero and Kölbel (2014),
Hacking (1975), Hale and Wright (1999), Lepore and Smith
(2008), Linsky (1977), Mackenzie (1997), Martin (1987), Miller (1998), Morris (2007),
Russell and Fara (2012), Stainton
(1996), Taylor (1998). More sophisticated accounts
can be found in Grice (1989), Hanna and
Harrison (2004), Horn (1989/2001), and Horwich (1998, 2005). Of course, these days there
is a surfeit of on-line material of varying quality -- see, for example,
here,
here,
here
(the latter links to a PDF), and especially
here.
[See also these
links.]
Easily the best on-line resource is the Stanford
Encyclopedia of Philosophy.
Stephen Schiffer,
a former 'high priest' of the idea that meaning depends on, or revolves around, intentions (etc.)
has recently recanted and authored a book explaining why that approach to
language can't work; cf.,
Schiffer (1987). [Although he has now backtracked somewhat in Schiffer (2003).
It would be good to know when he intends to make his mind up.]
There is a useful bibliography of modern work in this area in Preston (1997),
pp.237-43. Background details can be found in Martinich and Sosa (2005), and
Passmore (1966).
Accounts of language written from a
Wittgensteinian, or quasi-Wittgensteinian perspective can be found in many of the books and articles
listed below.
To these may be added the following: Baker (1988), Baker and Hacker
(1984a, 1988, 2005a), Bloor (1997), Bogen (1972),
Candlish and Wrisley (2019), Canfield (1981, 1994, 1996), Canfield and
Shanker (1993), Cavell (1996), Charles and Child (2001), Cook (1980, 1982), Crary and Read
(2000), Diamond (1995, 1997), Egidi (1995), Fann (1978), Fogelin (1987), Glock (1996, 2001, 2009),
Hallett (1967, 1977, 1984, 1988, 1991, 2008), Hanfling (1989, 2000, 2002a),
Hanna and Harrison (2004), Kennick (1972), Kenny (1998, 2006a), Kripke (1982),
Kusch (2006), Malcolm (1995a), Niles (1992), Palmer (1988), Pitcher (1968),
Rhees (1970b, 1999), Savickey (1999), Shanker (1986a, 1986b), Sluga and Stern
(1996), Stroud (2000a), Vesey (1974), Waismann (1997), Wittgenstein (1958/2009, 1969,
1972, 1974a, 1975, 1979, 1980d, 1981), and White (1974, 2006), as well as
several articles in Kuusela and McGinn (2011).
I do not propose to enter into a discussion
here of Wittgenstein's controversial claim that Psychology is "conceptually
confused" (but, see below), nor with the view that much of what passes for
contemporary Philosophical Psychology and 'Philosophy of Mind', is almost as
muddled as Locke and Descartes's writings; on this, see Fischer (2011a,
2011b). [Nevertheless, two relatively recent popular accounts of certain aspects of
contemporary work in this area inadvertently confirm that his negative
assessment isn't all that
wide of the mark: cf., Horgan (1996, 1999); many more examples can be found in
Bennett and Hacker (2008, 2021).] On this in general, see also Bennett, et al (2007),
Candlish (2002), and Fischer (2011a, 2011b). [This links to a PDF.]
A history of the
traditional approach to
'Mind' and 'Consciousness' (an ideology appropriated by
dialecticians, too) can be found in Bennett and Hacker (2008, 2021), Hacker (2003a, 2007,
2012, 2013a, 2013b), Chapter
Eight of Robinson (2008),
Rorty (1980), pp.17-69, and
Ryle (1949a),
pp.13-25. [These link to
PDFs. See
also, here.]
"The confusion and barrenness
of psychology is not to be explained by calling it a 'young science'; its state
is not comparable with that of physics, for instance, in its beginnings. (Rather
with that of certain branches of mathematics.
Set theory.) For in psychology
there are experimental methods and conceptual confusion. (As in the other
case conceptual confusion and methods of proof.)" [Wittgenstein (1958), §xiv,
p.232e.]
Naturally, that is a
controversial enough pronouncement on its own (on this, see
Hacker (2007c)). However, and in addition to the works listed above, extensive critical examination
of the perennial confusions found in both Psychology and the 'Philosophy of
Mind' (which includes much of what Voloshinov, Parrington, Holborow, Collins and
McNally have written on this topic) appear in the following: Anscombe
(2000), Arrington (2001), Baker and Hacker (1983a, 1983b, 1984, 2005a), Bennett
and Hacker (2008, 2021), Budd (1989), Button, et al (1995),
Canfield (1994), Coulter (1983, 1989, 1993, 1997), Coulter and Sharrock (2007),
Dreyfus (1992), Erneling (1993), Glock (1996), Goldberg (1968, 1991), Goldstein
(1999), Hacker (1987, 1991, 1993a, 1993b, 1993c, 1993d, 1993e, 1993f, 1993g,
1996, 1997, 2000b, 2000c, 2001b, 2002, 2007a, 2007c,
2012, 2013a, 2013b), Hanfling
(2001a, 2001b, 2002b), Hark (1990, 1995), Heil
(1981), Hilmy (1987), Hursthouse (2000), Hutto (1995), Hyman (1989, 1991, 2006), Kenny (1973a, 1975,
1984b, 1984c, 1992, 1993, 2003, 2006a, 2009), Malcolm (1968, 1977a, 1977b, 1980, 1984,
1986b, 1986c, 1995b, 1995c), Proudfoot (1997), Ryle (1949a, 1960, 1971a, 1971b, 1982a)
(this links to a PDF), Schroeder
(2001a), Schulte (1993), Shanker (1986c, 1987a, 1987b, 1987c, 1988, 1995, 1996b,
1997, 1998), Stern (1995), Stroud (1990, 1991, 1996), Suter (1989), Taylor
(1964), Vesey
(1994), Williams (1999), and
Wittgenstein (1958/2009, 1969, 1980a, 1980b, 1981, 1982, 1989, 1992, 1993, 2013).
However, having argued above that Voloshinov
conflates the meaning of a word or sentence with what a speaker might intend or
might seek to accomplish by means of it -- that is, with how a sentence might
be, or is actually, used -- it has to be admitted that
Wittgenstein himself wasn't above making the same mistake! In parts
of his later work (but mostly in notebooks), he began to experiment with the idea
that the use of a sentence could be a guide to the meaning of the words
it contained, or, indeed, the meaning of what was intended by whoever uttered it, so that the
point of saying
something, or who had said it, could affect its sense (or its meaning; Wittgenstein grew a little
vague over this distinction, it has to be said). These comments typically arose in
connection with what have come to be known as 'Avowals' -- that is, first person
reports of pain, etc. [On this, see Hacker (1993a), pp.83-96.]
Unfortunately, if true, this would imply that
sentences and words function in the same way, but beyond certain
sorts of clichéd and stereotypical sentences, rules of grammar, so-called
'occasion sentences', 'one word sentences' and some figures of speech, this
idea has little to recommend it -- for reasons outlined in the
main body
of this Essay, and again below.
Returning to Wittgenstein, it is worth
pointing out that much of his posthumously published writings (where this change
of emphasis began to appear) wasn't intended for publication -- he
actually wanted most of it destroyed. In these late writings, Wittgenstein was
experimenting with new ideas, many of which were highly tentative.
Anyway, the notion that sentenceuse could determine meaning in
general (or that it could form part of a general account/theory of language)
would have been totally alien to his entire philosophical approach, which was to
emphasise how diverse language use is, and how any attempt to find the
'essence' of meaning, or of use, was completely misguided. Hence, an account that linked
word meaning to sentence use could, at best, only ever have assumed a
peripheral role in his account of language.
Unfortunately,
many of Wittgenstein's
half-formed thoughts have been ossified into dogma by his epigones and have thus
been turned into
eternally true statements that supposedly represent his 'official position',
even though the 'proof texts' on offer often come from such private
notebooks not intended for publication! Indeed, his last major work
(Wittgenstein (1958/2009)) was under constant revision right up until his death, and
thus remained incomplete. It was 'completed' by his literary executors on what now
appear to be unsound lines. [On this, see Stern (1995, 1996). On the difficulties
of interpreting Wittgenstein, see Cavell (1971b, 1996), and Heal (1995).]
Many of the changes
Wittgenstein intended to make are now reflected in the Fourth Edition of the
Philosophical Investigations, i.e.,
Wittgenstein (2009). [For details, see
the Editorial Preface -- pp.viii-xxiii.
(These link to a PDF.)]
As noted above, powerful
support for an Occasionalist
interpretation of Wittgenstein can be found in Travis (1989, 2000, 2006). [See
also Hilary Putnam's review of Travis (2000) in Putnam (2002).] Clearly, it
would be out of place to discuss Travis's work in detail here; however, since
this Essay sets its face against the general application of
Occasionalist theses in the Philosophy of Language, a few comments seem to be
in order.
Travis's work shares much with
many other
'American' attempts to appropriate Wittgenstein's method: it misses the
point entirely. In an analogous fashion, it is almost as if someone thought
that they could further the cause of workers' power (in a 'western democracy')
by forming a clandestine conspiratorial,
Blanquist, group. No matter what finally emerged as
a result (and howsoever noble or well-meaning the intentions), workers' interests wouldn't be advanced one millimetre by such tactics.
Analogously the same applies to those
commentators who fail to notice (or who miss the significance of) Wittgenstein's
implacably anti-theoretical stance (in Philosophy). [On this, see Kuusela
(2006, 2008).]
Travis's work is blessed with commendable
attention to detail, and if correct, much of modern semantic and cognitive
theory would surely go out of the window. However, whatever one thinks of that
particular point, the methods Travis will have used to attain it (assuming
that this was his goal, of course) are as un-Wittgensteinian as one could imagine.
Naturally, that in itself isn't sufficient to condemn Travis's approach (which
clearly must stand or fall on its own merits), but as was the case with the
political example mentioned above, if Travis were correct, much of
Wittgenstein's work would end up in the trashcan, too. And that most certainly
wasn't Travis's aim.
The main problem with Travis's work, as far
as I can see, is that he completely ignores those features of language we
already have that allow us to speak of the non-occasional meaning of the words
we use. To be more precise, we have in our repertoire linguistic expressions
whose meaning would become problematic if they were 'occasion-sensitive' in the
way Travis envisages. Indeed, if he were correct, we wouldn't be able to make
much sense of Travis's own work, save we knew the occasion on which (or for which) he
wrote it. Since this topic was
discussedearlier, no more will be said about it
here. [As noted above, on this see Bridges (2010).]
87.
Voloshinov's book contains little that is recognisably scientific. It
does, however, contain much that is dogmatic. In that respect at least
it resembles 'Continental
Philosophy' sure enough. But, little comfort can be garnered from that fact -- since
it means his work resembles science even less.
[McNally (2001, 2004) added a page or two of comments about dogmatism. I will examine what he
has to say in a later re-write of this Essay.]
87a.
This isn't entirely fair; as noted
above, there are rare flashes of
good sense in Voloshinov's book.
88.
Admittedly, Parrington and Holborow share this idea with Marx and Engels, who
asserted as much in The German Ideology. Cf., Marx and Engels (1976),
p.44 -- i.e., Marx and Engels (1970), p.51.
89.
Naturally, this list doesn't exhaust the possible uses of this word.
89a.
It could be argued that this isn't so since this "need to communicate"
must have arisen
directly on the basis of, or out of, rudimentary social development and co-operative labour. While that might indeed be
the case, this "need" can't have been a felt need.
To conclude otherwise would be tantamount to holding the view that this need was
represented or expressed in language before language was invented!
If, on the other hand, it is argued that this "need to communicate" wasn't based on language, but on
"thought", that would collapse this alternative back into representationalism
and the belief that human beings initially represent the world to
themselves in some form or other, which would then prompt one or more of them to form the 'languageless thought'
(whatever that means!) that communication would be a good idea, etc. But,
for that to work, several individuals will have had to have had the 'same
thought' at the same time. How such thoughts could be called the "same" when
there is no way that they could be the same (except miraculously, or coincidentally), is,
of course, something of a mystery.
Naturally, this enquiry itself
has already been heavily
biased (if not compromised) since it is we who are attributing the thought "We need
to communicate" to those who had as yet no language, and no concept of communication, let
alone any medium by means of which either could be expressed.
On
the other hand, if each individual formed their own language first, so that they
could form the thought "We
need to communicate", expressed in their own
idiolect, or
proto-language -- even if that were
possible -- it would once again imply
that language is primarily
representational and individualistic, not social and collective. [On this, see
Note 61.]
Indeed, Marx had already argued that the
invention of language by a lone individual isn't credible:
"Production by an
isolated individual outside society -- a rare exception which may well occur
when a civilized person in whom the social forces are already dynamically
present is cast by accident into the wilderness -- is as much of an absurdity as
is the development of language without individuals living together and
talking to each other. There is no point in dwelling on this any longer. The
point could go entirely unmentioned if this twaddle, which had sense and reason
for the eighteenth-century characters, had not been earnestly pulled back into
the centre of the most modern economics by
Bastiat,
Carey,
Proudhon etc...." [Marx
(1973), pp.84-85. Bold emphasis alone added.]
"...this relation to land and soil,
to the earth, as the property of the labouring individual -- who thus appears
from the outset not merely as labouring individual, in this abstraction, but who
has an objective mode of existence in his ownership of the land, an
existence presupposed to his activity, and not merely as a result of
it, a presupposition of his activity just like his skin, his sense organs, which
of course he also reproduces and develops etc. in the life process, but which
are nevertheless presuppositions of this process of his reproduction -- is
instantly mediated by the naturally arisen, spontaneous, more or less
historically developed and modified presence of the individual as member of
a commune -- his naturally arisen presence as member of a tribe etc. An
isolated individual could no more have property in land and soil than he could
speak. He could, of course, live off it as substance, as do the animals." [Ibid.,
p.485. Bold emphasis alone added.]
Those who have read Lenin's
MEC might be tempted
to reply that we do indeed represent the world to ourselves in "images", and
this must have been the case before language developed. However, that theory has
been subjected to devastating and sustained criticism in Essay Thirteen
Part One. In addition, this line-of-thought
also undermines the social nature of language and thought, as has
been made plain above several times.
89b.
It is worth pointing out at this stage that there is much in Harman's article
with which I agree. Even though
Anthropology
(etc.) has moved on since his article was written, as have all the sciences, I think Harman gets most
things right -- from a Marxist
perspective, that is.
89c. This is despite the fact that Harman says the following:
"The sparsity
of reliable information makes it very easy for people to make elaborate,
unsubstantiated conjectures about what might have happened, with no facts to
confirm or deny them -- the modern version of the 'Just
So' stories
Rudyard
Kipling
wrote for children nearly a century ago. All sorts of writers on
human evolution make hypotheses of the form, 'And, so, perhaps, we can explain
the descent of certain apes from the trees by their need to do X'. Within a
couple of paragraphs, the 'perhaps' has gone, and X becomes the origin of
humanity.
"This method is
the special hallmark of sociobiologists, but there are also some very good
theorists who fall into it on occasions. It is a method Marxists have to reject.
We are not interested in story telling for the sake of story telling. So I
will try to concentrate on what we know for certain. [Harman (1994),
p.87. Bold emphasis added.]
If Harman ever gets round to writing an
updated version of
this article, it is to be hoped he tries even harder to rely on what we know for
certain -- and thus drops any and all references to Engels's shaky first 'Law', which,
as we have seen, is no less dubious
than many of the 'Just So' stories Sociobiologists try to sell their readers.
[These
comments were, of course, written before
Chris's untimely death.]
90.
Harman appears to be somewhat ambivalent about what is and what isn't genetically
'determined'. On the one hand he says:
"So there has to be a
recognition of how quantity turns into quality, of how through successive
changes animal life gave birth to that new form of life we call 'human', which
has a dynamic of its own, shaped by its labour and its culture not by its
genes…." [Harman
(1994), p.102. Bold emphasis added.]
On
the other hand, he also endorses Chomsky's genetic/nativist theory of the origin of language.
On that, see Note 109, below.
Indeed, in the previous couple of paragraphs (to the one quoted above), he had this to
say:
"It is this which also
explains the development of those most peculiarly human attributes, language and
consciousness. The distinctive feature about human language, as opposed to the
sounds and gestures made by other animals, is that we use words to refer to
things and situations that are not actually present in front of us. We use them
to abstract from the reality that confronts us and to describe other realities.
And once we can do this to others, we can also do it to ourselves, using the
'inner speech' that goes on inside our heads to envisage new situations and new
goals. The ability to do these things cannot have arisen at one go. It must have
grown up over many generations as our remote ancestors learnt in practice,
through labour, to abstract from and to change immediate reality -- as they began
to use sounds and gestures not merely to indicate what was immediately in front
of them or what they immediately desired (which is what some animals do) but to
indicate how they wanted to change something and how they wanted others to help
them. In tool use we know there was a significant change from the ape to the
early humans: the ape picks up a stick or stone to use as a tool; the early
humans of 2 million years ago were already not only shaping the stick or stone,
but using other stones to do the shaping, and, undoubtedly, learning from each
other how to do this. This implies not merely conceptions about immediate things
(food stuffs), but about things once removed from immediacy (the tool that can
get the food stuff) and twice removed from immediate reality (the tool that can
shape the tool that gets the food stuff). And it also implies communication,
whether by gesture or sound, about things two stages removed from immediate
conditions -- in effect, the first use of abstract nouns, adjectives and verbs.
The development of labour and the development of communication thus,
necessarily, go hand in hand. And as they both develop, they both encourage the
selection of those new genes which made people more adept at both: the more
agile hand, the larger brain, the larynx that made a wider range of sounds.
"Such developments do not
involve just quantitative changes. As the growth of labour, the growth of
sociability and the growth of language reinforced each other, encouraging the
selection of a whole range of new genes, new networks of nerve cells would
emerge in the brain, making possible whole new ranges of interaction between
people and the world around them. This may well explain why suddenly new
species of humans developed that lived alongside and then superseded those that
went before, as with the successive emergence of
homo habilis, of
homo
erectus, of the various sorts of archaic human. Thus, it may well be the
case that modern humans eventually replaced the
Neanderthals
because they were
able to communicate more quickly and clearly with each other (although we will
probably never know for certain if this was so)." [Ibid.,
pp.101-02. Bold emphasis added.]
But,
as we will see, none of this can be made consistent with Chomsky's theory. [On
that, see here.]
Even John Parrington spotted this!
"Both Chomsky
and
Pinker
are similarly incapable of explaining the influence of social change on language
and thus of the development of language over time. Both these flaws are linked
to a much deeper and more fundamental one -- the separation between individual
thought and social language that we found in the earlier theories of language.
Voloshinov's work is as relevant today as it was in the 1920s because it bridges
the gap between individual thought and social language." [Parrington (1997),
p.125.]
"And it also implies
communication, whether by gesture or sound, about things two stages removed from
immediate conditions -- in effect, the first use of abstract nouns, adjectives
and verbs. The development of labour and the development of communication thus,
necessarily, go hand in hand." [Harman
(1994), p.102.]
Abstract nouns (I'm not too sure what an "abstract verb" is) do not seem to be related in any obvious way to our ability to
talk about remote objects and events. For example, suppose one of our ancestors
had said either of the following (in their own language, of course):
Which
of these involves the use of abstract nouns?
Of
course, it could be argued (along lines pursued by John Rees) that "yesterday",
for example, is abstract -- perhaps on the following grounds:
"Indeed, the very
concept of 'fact' is itself an abstraction, because no one has ever eaten,
tasted, smelt, seen or heard a 'fact,' which is a mental generalization that
distinguishes actually existing phenomena from imaginary conceptions." [Rees
(1998a), p.131.]
That is, a noun is abstract if it refers to, or picks out, something that can't
be "eaten..., etc." Except, Harman seems to want to add the extra condition that
something need only be remote for it to require an abstract term to pick it out.
In that case, the Sun would be an abstraction (or "Sun" would be an abstract
noun), to say nothing of the stars! One might now wonder how such an abstraction is able to supply the earth with abundant energy.
However, as we will see in Essay Twelve Part Four, Rees's crude test is highly
questionable,
to say
the least.
Again, it could be replied that "sabre tooth tiger" and "deer" are general, and
are therefore abstract terms. That conclusion is controversial; see Essay Three Parts
One and
Two on this, where we saw that
traditional theories of abstraction in fact turn general words into the Proper Names of
'abstract particulars', destroying generality
--, and
with that would go the capacity of
predicative propositions to say anything at all.
Even so, "sabre tooth
tiger" and "deer" both pass 'the Rees test', above.
However, those that still think such words are abstract need only consider the
following (schematic) sentence:
T3: NN is in XX.
Where
"NN" is the Proper Name of one of the above ancestors and "XX" is the name of the
village/community/cave he/she comes from. We only have to imagine that when T3, for
instance, was
uttered, both hearer and speaker were situated several miles away from XX.
In this not unlikely scenario, there are no abstract terms. [Unless we suppose
that NN lives in an abstraction!]
Independently of this, Andrew Sayer has shown that if one accepts traditional
theories of abstraction (and I must confess I don't), many abstractions are
just as "real" as their concrete cousins. [Sayer (1992), pp.45-84.
This is now Sayer (2010), pp.31-57.]
91.
Harman's
imprecise -- if not sloppy -- use of language here is not unconnected with
several other aspects of DM examined in earlier Essays. In this particular case: the
projection and imposition of human categories onto 'reality' subsequently re-'emerge' as the very thing that had been inserted at the beginning!
Hence, if nature is
anthropomorphised and fetishised from the get-go -- by means of the use of concepts appropriated from
Traditional Philosophy, but more specifically, from
Hegelian Hermeticism --, then it
is no surprise to see 'intelligence' miraculously popping out at the other end.
This then fools dialecticians into thinking that they have made an important
contribution to scientific knowledge when, of course, all they have done is work
out the logical consequences of that earlier, unwise projection/imposition/insertion. If human categories are
reified
and foisted on nature, they will, sooner-or-later, re-emerge in a theory
that pictures objects and processes acting as if they are quasi-agents,
with aims and intentions (i.e., in particular, the use of concepts and jargon that motivate
a belief in 'determinism'), engaging in conflicts -- all those
'dialectical contradictions' --,
possessing goals
-- via the operation of the
NON --, and having aims and interests of their own -- e.g., "natural selection" will "bring about"/"encourage" X, Y or Z....
This helps account for the widespread use by dialecticians of
teleological
language in their attempt to explain natural processes. If nature
has been animated from the start by borrowing the mystical jargon invented by
German Idealists
and other
assorted mystics
-- a fatal blunder that was further compounded by the misconstrual and
fetishisation of language along the
way --, it is hardly surprising that teleological properties
suddenly "emerge" at the end. In this instance dialectical
word-juggling has only succeeded in giving birth to a novel form of teleologicalSuperscience, which then took the
place of ancient
forms of Christian Mysticism -- an approach to theory, incidentally, that has
had much the same impact on Traditional Thought over the
last two thousand years or so as it has had on DM.
[NON = Negation of the
Negation.]
[On this, see Essay Twelve Part One
and the rest of Essay Twelve when it is finally published (summary
here).]
With
respect to DM, this means, for example, that a particular word -- which should
normally only feature in the depiction or description of a conversation or an
argument between two human beings (i.e., "contradiction") -- has now re-surfaced
in one of
the 'laws of dialectics'. As a result, we not only have the NON governing
the development
of nature and society, there appear to be 'contradictions' and 'abstractions' at every
turn.
As Hegel argued (of course, without bothering to supply any proof, or
even a single scrap of supporting evidence -- and the DM-nodding-dogs dutifully
concurred):
"Contradiction is the very
moving principle of the world: and it is ridiculous to say that contradiction is
unthinkable. The only thing correct in that statement is that contradiction
is not the end of the matter, but cancels itself. But contradiction, when
cancelled, does not leave abstract identity; for that is itself only one side of
the contrariety. The proximate result of opposition (when realised as
contradiction) is the Ground, which contains identity as well as difference
superseded and deposited to elements in the completer notion." [Hegel
(1975), p.174;
Essence as Ground of Existence, §119.
Bold emphasis added.]
"But
it is one of the fundamental prejudices of logic as hitherto understood and of
ordinary thinking that contradiction is not so characteristically essential and
immanent a determination as identity; but in fact, if it were a question of
grading the two determinations and they had to be kept separate, then
contradiction would have to be taken as the profounder determination and more
characteristic of essence. For as against contradiction, identity is merely the
determination of the simple immediate, of dead being; but contradiction is the
root of all movement and vitality;
it is only in so far as something has a contradiction within it that it moves,
has an urge and activity." [Hegel (1999),
p.439, §956. Bold emphasis added.]
Hegel
did, however, attempt to cobble-together several 'arguments' of
dubious merit in support. On that, see
here and
here. I
will be examining these 'arguments' in more detail in Essay Twelve Parts Five
and Six.]
Whether or not it is true that Darwin
banished teleology from nature, one thing is for sure, DM only succeeded in
re-introducing
it. Indeed, as I argued in Essay Four, dialecticians have
re-enchanted nature.
92. Elsewhere (here
and here), I
have likened this tactic to the 'God-of-the-gaps'
dodge pulled by the Creationists. The way the latter works is as follows: if there is
(at present) no
scientific explanation for some phenomenon or other, then 'God', a 'miracle' or some
other supernatural 'cause' is introduced to account for it.
In like manner,
DM-theorists use similar verbal tricks to plug the gaps in their theory. For example,
they employ the ubiquitous word "dialectical" (a) To account for
any processes which they otherwise can't explain, and
(b) To neutralise the contradictions apparent in their own theory. [On that, see
here.] Another
word they also like to overuse in this way is "emerge". New properties simply
and miraculously seem to "emerge" from unsuspecting configurations of matter, which "can't be
reduced to" underlying processes. No further explanation is required -- so don't ask. [An excellent example of
this approach can be found in
Note 94, below.]
I will,
however, destructively analyse this
verbal con-trick in Essay Three Part Three (when it is published). [Until then, see
here, where I have called
this dodge "The Nixon Gambit", after the
37th
President of the United States and
the things he said about the
Vietnam War in the run-up to his election in November 1968.]
92a. One suspects, however, that
there might have been a misprint here, or some other typographical error. Either that, or
Harman has conflated the meanings of "contingent" and "unlikely"
-- or
perhaps even "neither likely nor unlikely".
Be this as it may, it is clear that Paul
McGarr has a slightly firmer grasp of the issues involved, even though he too proceeds to make the usual mistakes,
in
McGarr (2003), pp.102-09. I will say more about that in Essay Three Part
Five.
93. On this, see Note 91 above. Of course, it could
be argued that Harman is merely highlighting the uncontroversial fact that certain events are more
likely to happen given previous stages in evolution and in cultural
development. Hence, it could be maintained that while he rejects
contingency here, he is still not embracing teleological necessity. But, in that case,
Harman must be using "contingent" in a new and as-yet-unexplained way.
Indeed, as we will soon see, several other things he says don't harmonise at all well with this
hypothetical defence.
Further proof that Harman is a necessitarian-still-in-the-closet can to be
found in the following passage from Engels, which Harman quotes approvingly:
"'Men-in-the-making arrived
at the point where they had something to say to each other. Necessity
created the organ; the undeveloped larynx of the ape was slowly but surely
transformed by modulation to produce constantly more developed modulation, and
the organs of the mouth gradually learned to pronounce one articulate sound
after another.'
"Parallel with
this there was a necessary development of the brain:
'The reaction
of labour and speech on the development of the brain and its attendant senses,
of the increasing clarity of consciousness, power of abstraction and of
conclusion, gave both labour and speech an ever renewed impulse to further
development.'" [Harman (1994),
p.85; quoting Engels (1876),
p.356. Bold emphases added.]
This
seems pretty clear.
Concerning the
teleological nature of Darwinism (in
its classical and
more modern forms), see Stove (2006), pp.205-306, and Fuller (2006).
[Readers unfamiliar with David Stove's work should, however, consult
this warning.]
As
the late
Jerry Fodor noted (in relation to Daniel Dennett's
attempt to explain 'design', 'selection for' and 'reverse
engineering' in naturalistic terms:
"Suppose, however, that
adaptationism
is true; is it able to ground a notion of natural teleology?...
Is it then reasonable to speak of [a certain property] P as a property that
[organism] O was 'designed' to have? Or as a 'solution' to an 'engineering
problem' that O's ecology posed?...
"The subtext is the thing to
keep your eye on here. It is, no doubt, an interesting question in its own right
whether adaptationism licences teleological notions like SELECTION FOR. But what
makes that question interesting in the present metaphysical context is that
SELECTION FOR is presumably
intensional.... If so, then maybe a
naturalistic teleology is indeed a first step toward a naturalistic theory of
mind.
"But, promising though it may
seem, I'm afraid this line is hopeless, and for familiar reasons. Design (as
opposed to mere order) requires a designer. Not theologically or
metaphysically..., but just conceptually. You can't explain intentionality by
appealing to the notion of design because the notion of design presupposes
intentionality....
"Patently, not every
effect that a process has is
ipso facto
an effect that it designs; short of theology, at least some effects of every
process are merely adventitious. This must hold of the process of natural
selection
inter alia. So, in evolutionary theory as elsewhere, if you wish to deploy
the idiom of posed problems and designed solutions [as Dennett does -- RL], you
must say something about what designing requires over and above mere causing.
Lacking this distinction, everything a process causes is (vacuously) one of its
designed effects, and every one of its effects is (vacuously) the solution to
the problem of causing one of those.
"To be sure, if solutions
aren't distinguished from mere effects, it does come out -- as Dennett would
want it to -- that the giraffe's long neck solved the problem of reaching to the
top of things, and did so under precisely the ecological conditions that
giraffes evolved in. But equally...the
Rockies
solve the problem of how to make mountains just like the Rockies out of
just the materials that the Rockies are made of and under just the conditions of
upthrust and erosion in which they formed; and the Pacific Ocean solves the
problem of how to make a hole of just that size and just that shape that is
filled with just that much salt water; and the tree in my garden solves the
problem of how to cast a shadow just that long at just this time of the day.
This, however, is no metaphysical breakthrough. It's just a rather pointless way
of talking; neither [the Rockies] nor the Pacific get any kudos for being
solutions in this attenuated sense. That's because problems like this are like
headaches; they don't float free of people having them. The Pacific [and the
Rockies] didn't really solve anything because nobody had the
problems that [these] would have been the solutions to....
"Serious talk about problems
and solutions requires a serious account of the difference between designing and
merely causing. Notice, moreover, that if your goal is a reductive theory of
intentionality, then your account of this difference cannot itself invoke
intentional idiom in any essential way. This really does make things hard for
Dennett. In the usual case, we distinguish designing from mere causing by
reference to the effects that the designer did or didn't intend. For
example: The flowers that Sam gave Mary made her wheeze and did not please her.
They were, nonetheless, a failed solution to the please-Mary problem, not a
successful solution to the wheeze-Mary problem. That's because Sam intended that
receiving the flowers should please her and did not intend that they should
excite her asthma. Suppose, by contrast, that Mary merely came across the
flowers, and they both pleased her and made her wheeze. Then the flowers didn't
solve, or fail to solve, anything;
they just had whatever effects they did.... It certainly looks [from this that]
the concept of design presupposes, and hence cannot be invoked to explain,
the accessibility of intentional idiom.
"If you found a watch on a
desert island, you'd have a couple of options. You could argue that since it was
clearly designed, there has to have been a designer; or you could argue that
since there certainly was no designer, the watch can't have been designed. What
is not, however, available is the course that Dennett appears to be
embarked upon: there was no designer, but the watch was designed all the same.
That just makes no sense." [Fodor (1998c), pp.176-78. Capitals and
italics in the original; Quotation marks altered to conform with the conventions
adopted at this
site.]
Incidentally, the above (in a way) is just a
modern (and perhaps more sophisticated) version of the argument between
Rationalist and Empiricist Philosophers over the nature of causation two or
three hundred years ago. I have summarised the main points relevant to Hegel's
reply to Hume,
here.
[In this, Dennett is Hegel (anthropomorphising
nature in order to make his ideas even seem to work), and Fodor is a
'sort of Hume', minus the empiricism and
associationist psychology, among other things.]
Naturally, the determined DM-supporter (no pun intended) might want to argue
that these two cases (that is, those involving water and language) are not at
all comparable. The obvious differences between these two mean that the origin
of language will require its own, but more complex and sophisticated, explanation.
[I
will discuss several standard accounts of the
origin of language in more detail in a later re-write of this Essay -- those found,
for example, in the following works (among others): Bickerton
(2009), Boyd and Richerson (2005), Burling (2005),
Carruthers and Boucher (1998), Carruthers, Laurence, and Stich (2005, 2006,
2007), Carstairs-McCarthy
(1999), Christiansen and Kirby (2005), Crozier (2008), Deacon (1997), Dunbar, Knight, and Power (1999),
Erneling and Johnson (2005), Hurford (2007), Knight, Studdert-Kennedy, and
Hurford (2000), MacNeilage
(2008), Richerson and Boyd
(2005), Sterelny (2003), Stich and Warfield (1994), and Wray (2002). I will also look at
Ruth
Garrett Millikan's interesting work. E.g., Millikan (1984, 2005a).]
However, this issue now introduces the well-worn --
but misleadingly entitled -- topic: "Determinism", a subject I will be
discussing in more detail in Essay Three Part Five. [Among other
things, that Essay will contain a response to
Molyneux (1995).] However, in view of the above, a few comments are clearly called for. [I have also said more about 'Determinism' in the
Appendix, where I have
reproduced an edited version of an article I posted to Marxmail a few years ago.]
The critical problems facing any DM-account
of nature aren't unconnected with the severely limited logical, linguistic and
conceptual resources available to its adepts, and this in turn is largely because
dialecticians
have based their thinking on the sub-Aristotelian Logic
and mystical
confusions they inherited from Hegel (upside down or 'the right
way up').
According to Engels [for example,
Engels (1976), p.180 and
(1954), p.211], dialectical concepts provide
its adherents with an explanation of how the entire history of the universe has
proceeded up to the present (of course, most dialecticians also advance
similar hyper-bold claims). In practice, however, these concepts have been applied in a
consistently
aprioristic manner far in excess of the embarrassingly weak and
highly equivocal evidence offered in support,
and in many cases in defiance of
the
facts --
an approach further compounded by a reckless disavowal of the normal
cannons of reasoning. [On that, see
Essay Two, and Essays Four through Eleven Part Two.] If the ubiquitous
fetishisation of language perpetuated
by DM-theorists is now taken into account, the terminology dialecticians employ
-- strike that: the jargon they have imported from Hegel -- is inevitably going to make their theories of development appear far more
deterministic than they might otherwise have been.
[DM = Dialectical
Materialism.]
In
fact, on this issue, as elsewhere, DM-theorists share with Traditional
Philosophers an age-old misuse of the word
"determine". This term, alongside its cognates, normally relates to the application of
intelligence, purpose and direction (a point outlined in more detail Essay Three
Part Five; cf., Gallop (1962), and Russell (1917a)). Given these factors, it is no mystery
at all why DM is not only deterministic, it is covertly
animistic and
overtly teleological into the bargain
-- indeed, just as Traditional Philosophy is heavily slanted
in the same direction. [On the latter, see Note 91, above.]
For DM-theorists, as noted in Essay Eight Parts
One and
Two, the
ultimate link between events appears to be 'conceptual' (i.e.,
these links -- those "mediacies" --, are governed by what Lenin called
a "law of cognition" --
Lenin (1961), p.357) -- or, at least, it seems that DM-theorists alone are able to apprehend them
in this way.
[But, even then, if our thoughts about 'reality' are true (or are meant to be
increasing true), then nature itself must somehow be governed by an
analogous set of 'laws'. In Essay Twelve Part Four, I aim to show that this is
indeed the implication: if, per impossible, DM were true, Nature would be 'Mind'.] This can be seen from the
way that these connections were depicted. Human cognition of nature is supposedly based on a process
of active "reflection" and "abstraction"; the resulting 'abstractions' are then
supposed to "reflect" reality increasingly accurately (i.e., when tested and refined in practice).
Despite claims to the contrary, and as noted
above, this
can only mean that nature is ultimately 'mind-like', governed by the products of
some intellect or other; to be more precise, it must be governed by these
'abstractions', or whatever it was that first created such 'abstractions'
-- or even by what they allegedly 'reflect' -- before humanity became aware of them. After all, if we take away the human
intellect, what could these 'abstractions' actually 'reflect' in reality? If there is
nothing for them to 'reflect', they
can't be part of the material universe. On the other hand, if they do 'reflect'
something in nature, then the universe
must be the product of 'mind' -- or, it must be 'mind-like'. Once more, this is hardly
surprising given the origin of
DM in Mystical
Christianity and 'Pagan'
Hermeticism.
[This argument is set out in
more detail in Essay
Three
Part One,
where we saw that dialecticians are rather unclear whether their 'abstractions'
are (i) merely 'mental' constructs that help us understand and control nature, or
whether they (ii) 'reflect' objects and processes in 'external reality' -- or, indeed,
whether (iii) they are those objects and processes themselves. Once again, given that this
theory grew out of Mystical Christianity and
Cabbalistic Hermeticism
-- a set of theories that in the end can't distinguish fact from fantasy -- that is
no big surprise.]
On the other hand, DM-theorists also believe
that there are in nature what can only be described as mysterious 'pockets of freedom',
because certain processes are deemed to be free from 'determining causes' --,
or, at least there are processes in nature (for example, those connected with
human agency/action) which can't be "reduced" to them. The story seems to be
that 'freedom' just "emerges" from a sufficiently intricate material (or
social?) base as a result of the operation of Engels's First 'Law'. So, when
this 'Law' is applied to, or operates in/on systems that display the right level
of complexity, it seems capable explaining how such acts of will become 'free'
of their material base. Hence, while a given system might be 'determined' at one
level, at another it miraculously changes and qualitatively "leaps" into another
realm, one that remarkably, if not magically, displays 'freedom'. No
explanation is given as to how this is possible, or even how it actually
happens; it is
left a complete mystery. In the end, this is no explanation, it is merely a
re-description (in a 'favourable form') that shrouds the relevant phenomena in a dense fog
-- on a par
with religious 'explanations' of certain
events as 'acts of god'/'miracles'. Once again, this is hardly surprising
when we recall the mystical provenance of
this 'theory'.
But,
given other constraints operating on DM, this supposedly
supervenient property of matter (whereby 'free' events appear to float
unconstrained on a sea of 'determining' processes) can't itself be causally linked
to the base from which it has just "leaped", otherwise it, too, would be subject to
the usual 'determining' influences feeding up from that base --, thus cancelling,
or perhaps nullifying,
this alleged 'freedom'. So, given DM, there can be no physical, or even causal, link between matter
and human freedom! One wonders why human action is capricious.
That partly accounts for the hostility
DM-theorists have traditionally shown toward "reductionism". The latter
doctrine threatens to undermine their account of freedom, and hence human agency
in the class struggle, for example.
[At this point, perhaps only
Second International Marxists and Stalinists would want to hand-wave that particular
'concern' away.]
And yet, this means that there must now exist
a very real causal gap in the DM-account of natural systems. Given the truth
of DM, it seems that certain aspects of reality (i.e., those displaying
"freedom") simply can't
be given a causal explanation (or, at least, they can't be given a classically causal one), while
others can.
The DM-solution, if such it may be called, is
to (a) Brand any who object to this 'miraculous theory' of theirs, "mechanical/crude materialists",
or (b) Reject as "bourgeois" any demand that a causal account be given of
these miraculous 'pockets of freedom'.
As noted above, this tactic is different
in name only from the approach adopted by Christians, who appeal to the 'God-of-the-gaps',
'the soul', or other assorted 'miracles' to bale their 'theory' out.
On
the other hand, if these events and processes are fully and trulyfree, what could possibly explain their operation?
The question now naturally arises: Are such events completely
adventitious and
un-caused? If they are, then they will have no explanation, being
entirely capricious and totally unrelated to an agent's previous character and
social or individual circumstances.
[Of course, this depends on how "free will" is defined by DM-fans.
Good to anyone seeking a clear answer to that one!] How then are such acts any different
from those that are merely
coincidental? Or, from those that are completely random? Are these DM-'causes' noumenal?
Maybe these events and processes are
noumenal, too?
[On
coincidences, see Owens (1992). I must, however,
add that I distance myself from the latter's account of causation.]
Furthermore, if these events
and processes are indeed un-caused, nothing physical,
or recognisably physical, will actually connect them to their supposed material base. Such a 'connection' (if one
were to exist) can't therefore be material, but must be conceptual --
something which only a 'mind' can appreciate or even constitute -- as was pointed out
earlier. This helps explain why Lenin said
the following:
"Intelligent idealism is
closer to intelligent materialism than stupid materialism. Dialectical idealism instead
of intelligent; metaphysical, undeveloped, dead, crude, rigid instead of
stupid." [Lenin (1961), p.274.
Paragraphs merged.]
And yet, if there were such conceptual
links between a given material base and the aforementioned 'emergent' properties,
which preceded the evolution of the human species, then nature itself must once again be 'Mind', or
be governed by 'Mind'.
On the other hand, if there is
a
physical link between the material base and these mysterious 'emergent'
properties, then there will be an explanatory or (possibly) reductive law
(etc.) -- or a 'bridging law', at least -- that connects them to that base, and we
are back at square one. [Otherwise, the relevant objects and process will be
uncaused, once more, and thus merely capricious.]
[But,
as we will see, that
alternative is
no solution, either, for these 'laws' are themselves conceptual! Naturally, this means
that there can be no metaphysical solution to this classical 'problem', since
it, too, is
based on linguistic confusion, and nothing more.
That explains, of course, why this pseudo-problem has been around for so long
and resists all attempts at its resolution.]
In fact, in DM, 'free' events appear to be
completely autonomous -- they look for all the world as if they are non-material,
semi-ethereal
entities -- which, in mystical belief-systems, were attributed to the operation of the 'soul'.
Indeed, it isn't too
preposterous to suggest that spurious freedom like this, that just 'emerges'
from nowhere, is the DM-surrogate
for the 'soul', and which, in like manner, simply 'emerged' or emanated from 'God',
and which also enjoys its own
problematic links with matter.
That, of course, accounts for the allegations made throughout this
Essay (especially
here) that dialecticians have bought into the Traditional Platonic-Christian-Cartesian theory of
'the mind' as a non-material substance. Indeed, it also explains why Lenin
'defined' matter as that which exists "objectively outside the mind",
also
examined briefly above.
From that we can see that he was a closet dualist (despite his
pretensions, and that of his epigones, to the contrary), for how else can matter
be outside the 'mind'?
In addition, we saw earlier how Trotsky appealed to Engels's 'Law' to account for human
freedom (i.e., presumably, 'free will'):
"...[I]t
is precisely dialectical materialism
that prompts us to the idea that the psyche could not even be formed unless it
played an autonomous, that is, within certain limits, an independent role in the
life of the individual and the species. All the same, we approach
some sort of critical point, a break in the gradualness, a transition from
quantity into quality: the psyche arising from matter, is 'freed' from the
determinism of matter, so that it can independently --- by its own laws --
influence matter." [Trotsky (1986), p.106. Quotation marks altered to
conform with
the conventions adopted at this site. Paragraphs merged.]
And yet, if the "psyche" is bound by certain 'laws',
what are they, and how 'free' can the "psyche" be as a result?
How strange then that in trying to banish
mysticism from our understanding of nature and 'mind', DM-theorists invite this
Cartesian Ghost back in through the front
door, spread out a "welcome" mat, crack open a bottle of
Bollinger,
strike up a brass band and bestow on it the honorific title: "Yet Another Emergent Property"!
This also
partly accounts for the
fanfare that accompanies the many 'scientific' breakthroughs
DM-acolytes claim for their theory. DM-fans seem to be able to "grasp" a-contradiction-a-day
(or, in many cases, more than one),
which 'allows' them to claim, in this
instance, that they have solved the 'contradiction' between freedom and
necessity merely by saying that they have done so, and then refusing to give
any more details -- that is, other than telling doubters that they
just don't
'understand' dialectics.
[In a similar vein,
George W Bush
was, of
course, in favour of peace, freedom and democracy simply because he said so! Problem solved! In Essay Eight
Part One, this DM-ploy was called "The "Nixon Gambit".]
Worse still, some 'free' events seem capable
of operating in reverse gear, as it were. For example, acts
of 'free will' allow (supposedly) dialectically inspired agents to alter their own bodies and
affect the world
around them (aka "revolutionary practice"). While DM-theorists might want to
side-step
the minefield of reductionism, none of them, not
one, will deny that human beings can affect the material world by
certain 'acts of will'. But,
given their theory, this would seem to be impossible (or, at least, at present
it lacks any sort of
materialist explanation). If such 'free states' of mind (that accompany
these 'free' acts of will) are
autonomous and have no causal connection with underlying material micro-states,
how are they then capable of affecting matter on the return journey, so to speak?
Consider a simple example:
NN intentionally raises her arm. Here, we might be tempted to say that
she formed the intention,
acts upon it, and the arm was raised. But, what connects the material processes in
NN's CNS to
that action? Presumably, the intention. But, if that intention was 'free' then nothing
can physically, or causally, connect it with the CNS, and if that is so, nothing can
physically connect the CNS with the action itself. In which case, how is that action
attributable to the one supposedly 'intending' it?
On the other hand, if something physical does
connect the intention with the CNS, thus raising the arm, then it can't be free, but must be 'causally
determined'. Another appeal to 'emergent' properties here would be to no avail.
Even
if the said intention were an 'emergent' property (or based on one), it can't be connected by a causal law
to the CNS. That is because if it were connected by such a law, then that
intention can't be free, and neither can the act itself. And, if that is so,
neither can more complex human actions. [Plainly, this alternative collapses the
traditional DM-account right back into Second International Determinism.]
Neutral observers can now, perhaps, see why
dialecticians respond with accusations of 'crude materialism' at this point --
they have no other reply available to
them: name-calling.
[Incidentally, the above paragraphs shouldn't be
taken to mean that I accept this way of framing or conceptualising this topic! I am merely
rehearsing the traditional account, which DM-fans accept up to a point,
even if they then have to appeal to 'emergent properties' to save their theory
from accusations of extreme voluntarism on the one hand and reductionism or Second International Determinism
on the other.]
Others may be tempted to ask at this point: "Ok, Ms
Lichtenstein, what's your solution?" Such individuals are referred back to
this comment for
an answer. I reject the original 'problem' as both non-sensical andincoherent. [Impatient readers will
need to wait until Essay Three Part Five has been published for more details.
Until then, see Appendix A.]
However, the main lines of my reply have already been laid down above, and in
Essay Twelve Part One.
[In the
meantime, readers might like to see how I have responded to criticisms levelled
at me over at RevLeft,
here,
here,
here,
here, and
here (and follow the links there, too). Unfortunately, these links are now
dead!]
Alternatively, again, if these processes do
enjoy causal connections with their substrata (in the sense that they aren't entirely free-floating on
some sort of material substratum), how could spontaneity possibly have 'emerged'
in the first place? Are there 'holes' in the fabric of each brain that allow
'free will' to sneak in, and then back out again -- analogous to the Cartesian Soul
with its associated 'gateway to the material world', the
Pineal
Gland? If not, does 'free will' arise from something non-material? It seems
it must if it enjoys no causal links with its material base.
Naturally, conclusions like these are
unacceptable -- even to those DM-theorists who don't look to
probabilistically-driven quantum states to rescue spontaneity from the jaws
of inevitability -- for they would appear to be based on an Idealist solution to
the original 'problem'. That is because the DM-approach clearly divorces matter from 'mind', introducing
discontinuities in nature (i.e., the 'gaps' that Engels's First 'Law' is
supposed to have allowed the development of nature to "leap" across, and to
which
Trotsky and
John Parrington referred),
discontinuities that must be bridged by a DM-'law of cognition', along with its associated
conceptual links. Into such 'gaps' a
non-material mind can all too easily be smuggled. A scenario like this
clearly allows space for just such a ghostly human 'soul' to 'account' for
spontaneity and freedom, which prospect will, naturally, horrify DM-fans. This
also helps
explain the haste with which Alex Callinicos, for example, clutched at the
few meagre straws offered by
Daniel
Dennett's equally
defective explanation of
"consciousness", which 'explanation' only seems to work because it
anthropomorphises the brain with its appeal to diminutive Dennettian homunculi and 'dim algorithms', thereby
attributing our 'free acts' of will to the 'free acts' of will exercised by little men in our heads -- a
dishonest solution, at best. Others seek solace, if not refuge, in the impenetrable
gobbledygook scattered across Hegel's oeuvre -- and, perhaps worse, in much
that litters post-Heideggerian 'Continental
Philosophy', the metaphysical equivalent of
destroying a town
in order to save it.
[Anyone who doubts this hasn't read much of
what passes for 'Marxist Philosophy' these days and perhaps knows little of the intellectual gyrations and contortions
routinely performed by High Church
Dialecticians, for instance. Nor have they eye-balled the labyrinthine musings of
the likes of
Gerry Healy
-- samples
of which can be found
here and
here. (Health Warning: the latter can seriously damage your militancy!)
Sceptical readers should also consult Healy (1982, 1990) -- where, if they
follow this advice, they will deserve all
they get. Alternatively, try
this,
this and
this -- swiftly followed by a half bottle of single malt whiskey.]
Again, there appears to be no obvious way out
of this Dialectical Hole: either DM-theorists admit that everything is causally
related to (or "mediated" by) something/everything else, or
they concede that some things are uncaused (i.e., un-mediated, and hence are
non-material/immaterial),
thereby fragmenting their Totality.
[The details surrounding this dialectical
'difficulty' have been worked out extensively in Essay Thirteen
Part One.
Of course, dialecticians appear not to mean the same by "cause" as 'mechanical
materialists' appear to do, but then it isn't easy
to say what dialecticians
do mean by "cause", either. As noted
earlier, from the way they characterise objects and processes (with their 'internal relations'),
causes in effect appear to be little different from conceptual links
imagined to exist between certain objects and processes, which links are, alas, impossible to confirm by any known physical means.
(This quandary is, of course, a direct descendant of
Hegel's
attempt to respond the Hume's criticism of Rationalist Theories of Causation.) Hence, as is the
case with
traditional ruling-class thought, these 'links' are 'discoverable' by -- and
werein fact 'discovered' by -- the application of thought
alone, and relate to a non-material (Ideal) world anterior to experience, or,
indeed, 'appearances'. Again,
this is no surprise in view of the fact that these obscure notions were
appropriated from Hegel and Mystical
Neo-Platonism.]
The DM-solution (again, if such it may be
called) is to "grasp" this paradox and admit both halves of it!
This ploy isn't as harebrained
as might
at first sight seem: if your theory readily accepts the universal existence of 'contradictions',
then an additional paradox -- which has now arisen between the material base and human
"freedom" -- is just more grist to the DM-mill. However, as noted in Essay Eleven
Part One, the indulgence DM-theorists readily extend to the contradictions
inherent in their own theory is not only jealously guarded, it is strictly
rationed. In fact, such benign indulgence is consistently
withheld from
rivals and opponents and the contradictions their theories are said to exhibit -- including those
devised by theorists who display
only half as cavalier an attitude to logic as the average
dialectician. But, worse still: as also noted in Essay Seven
Part One, the welcome mat laid out for the many contradictions
inherent in DM would mean that -- according to dialecticians' own criteria for
scientific advance -- it must either (i) Fail to be a science, or, (ii) If it is
a science, it implies that science (and DM) can't advance. [On why that is so, see
here and
here.]
95.
Unfortunately, the only reference Harman gives in support of his allegations about
Sociobiology is a book called Blood Relations, by Chris Knight (i.e.,
Knight (1995) -- a work he had reviewed in an earlier issue of International
Socialism -- i.e., Harman (1992).
However, to be fair to Harman, the claim made in the main body of this Essay
doesn't represent the whole truth; in a
rare display of honesty about Engels, he admitted that Engels also
indulged in similar flights of fancy [Harman (1994),
p.184, n.12].
One wonders, then, what Harman made of Engels's rather odd belief in the
existence of a 'parrot language', or of Parrington and Holborow's reference to
the "need to communicate" as a revealing example of a Lamarckian explanation of the origin of
discourse. On this, see Note 96.]
96.
It shouldn't be concluded from this that the present author either accepts or
rejects Lamarckian explanations of the evolution of human society. The comment
in the main body of this Essay (at this point) relates to pre-linguistic
human, or proto-human, groups (which couldn't at that
stage be called communities); clearly, Lamarckian concepts can only be
applied to such groups if they possess linguistic, social and hence genuinely
teleological capacities -- or, alternatively, if such groups are
(metaphorically) regarded as organisms
in their own right.
97.
Although he wrote these words before Daniel Dennett's egregious book was
published [Dennett (1995)], if he re-reads Dennett's sophisticated
Neo-Darwinian/Sociobiological account of evolution he will see this point for
himself.
[That comment was, of course, written
before Chris's untimely death!] Dennett's work not only doesn't commit the intellectual crimes Hartman (rightly) attributes to
certain sociobiologists, Dennett actually attempts to answer Gould's criticism
that sociobiological
theories are little more than elaborate myths. [Cf., Dennett (1995), pp.238-51.]
Dennett's work, is however, susceptible to other, more searching criticisms; on that, see
here,
here and here. Anyway, as
noted above, Alex Callinicos certainly thinks highly of Dennett's
work [Callinicos (1996, 1997)]; apart from a few quibbles over relatively minor
issues, he downed the contents of that poisoned chalice with remarkable ease.
The serious problems that afflict
Neo-Darwinism won't be entered into here. On this in general, see Stove (2006).
[Readers unfamiliar with David Stove's
work should, however, check out this
warning.] Also see, Goodwin (1994, 2001), Kauffman (1993, 1995), Schwartz (1999), Webster
and Goodwin (1996), as well as the following interview given by
Brain
Goodwin:
"King:
How does your new model of biology incorporate genetics?
"Goodwin:
A major problem is that in contemporary Darwinism, organisms are actually
reduced to genes and their products. Darwinism has given us a very good theory
of inheritance in terms of a theory of the genes, but what it has done is to
sacrifice the whole organism, as a real entity, to this reductionism, genetic
reductionism.
"That means that organisms have disappeared as real entities from biology, and
that, I think, this (sic) is a fundamental scientific error. There's another aspect of
this problem which has to do with the way Darwinists explains (sic) embryonic
development. They say that there is a genetic program that determines the
development of an organism. An organism wants to become a newt, say, or a sea
urchin. Because it has particular genes, they say, it undergoes a particular
embryonic development and that is sufficient, in other words knowing the genes
is sufficient to understand the details of the embryonic development, and the
emergence of a species with its characteristic form and behaviour. That sounds,
on the face of it, plausible because we know that mutations actually cause
transformation of morphology.
Drosophila can have a mutation that transforms a
two winged fly to a four winged fly. Now that is a pretty major transformation,
and a single gene can do it. So you might say that's the sort of thing that is
involved in evolution. Well, you see, the burden of proof then is on the
neo-Darwinists to demonstrate exactly how the genes do this. They use the term
genetic programming, and it is a metaphor for what happens in a computer, but if
you ask them to use a genetic program to generate an organism, they can't do it,
and the reasons are very simple. You need to know more than gene products in
order to explain the emergence of shape and form in organisms. You actually need
a theory, a theory that involves physics, chemistry, forces and spatial
organization. You can have complete details about genes and you are not going to
be able to explain how development occurs. So I think that is the fundamental
test. When Darwinists say to me 'genes are enough', I say 'Show me.'...
"King:
How would the new science affect our social theories?
"Goodwin:
Well, another consequence of this new view of species and evolution is it does
shift the metaphors that are used to understand evolutionary processes. In
Darwinism, you know, the metaphors are of competition and conflict and survival,
and in Dawkins' writing it becomes embodied in the notion of selfish genes.
"Well, from the perspective of organisms as complex dynamic systems, with
natures and trying to understand the ecosystem from the (sic) point of view, what you
find is that organisms are interacting with each other in all kinds of different
ways. They are as co-operative as they are competitive, and a lot of the time
they are simply making a living. In other words, it's not this nature red in
tooth and claw, with fierce competition and the survivors coming away with the
spoils. In fact, species extinction seems to be as much to do with the lottery
which comes from the dynamics of complex systems, as from anything else. The
whole metaphor of evolution, instead of being one of competition, conflict and
survival, becomes one of creativity and transformation...." [Quoted from
here.]
The rise of
alternative
epigenetic theories of development and transmission has also begun to
threaten the intellectual hegemony enjoyed by Neo-Darwinism for the best part of
the last eighty years. [On this, see, for example, Carey (2011), Francis (2012)
and Ward (2018).]
The discussion between
Elliott
Sober and Jerry Fodor can be accessed
here. [See
also
Sober (2010); this links to a PDF.]
[Once more, it is important to note that even
though I am referencing the above works, I do not necessarily agree with everything
they contain. The above was written before
Jerry Fodor's recent death.]
97a.
Although, to be scrupulously accurate, Harman doesn't actually use the word "emergent" in
his article. However, what he does say about 'consciousness' and language
certainly implies he wouldn't disagree with this use of that term:
"It is this
which explains why our forebears were able, a million or so years ago, to move
out of their African ancestral home into the very different climatic conditions
of Eurasia, and why the
Neanderthals
were able to survive the harsh conditions of the European ice age for 100,000
years or more. However great or little their differences from us, they could not
have survived unless they had at least substantial rudiments of culture,
language and intelligence. After all, they were like us in one very
important respect: they had nothing else to protect them -- no body fur, no
great speed in flight, no tusks or claws, no ready ability to disappear into the
trees.
"It is this
which also explains the development of those most peculiarly human attributes,
language and consciousness. The distinctive feature about human language,
as opposed to the sounds and gestures made by other animals, is that we use
words to refer to things and situations that are not actually present in front
of us. We use them to abstract from the reality that confronts us and to
describe other realities. And once we can do this to others, we can also do
it to ourselves, using the 'inner speech' that goes on inside our heads to
envisage new situations and new goals. The ability to do these things
cannot have arisen at one go. It must have grown up over many generations as our
remote ancestors learnt in practice, through labour, to abstract from and to
change immediate reality -- as they began to use sounds and gestures not merely
to indicate what was immediately in front of them or what they immediately
desired (which is what some animals do) but to indicate how they wanted to
change something and how they wanted others to help them. In tool use we know
there was a significant change from the ape to the early humans: the ape picks
up a stick or stone to use as a tool; the early humans of 2 million years ago
were already not only shaping the stick or stone, but using other stones to do
the shaping, and, undoubtedly, learning from each other how to do this. This
implies not merely conceptions about immediate things (food stuffs), but about
things once removed from immediacy (the tool that can get the food stuff) and
twice removed from immediate reality (the tool that can shape the tool that gets
the food stuff). And it also implies communication, whether by gesture or sound,
about things two stages removed from immediate conditions -- in effect, the
first use of abstract nouns, adjectives and verbs. The development of labour
and the development of communication thus, necessarily, go hand in hand. And
as they both develop, they both encourage the selection of those new genes which
made people more adept at both: the more agile hand, the larger brain, the
larynx that made a wider range of sounds.
"Such
developments do not involve just quantitative changes. As the growth of labour,
the growth of sociability and the growth of language reinforced each other,
encouraging the selection of a whole range of new genes, new networks of nerve
cells would emerge in the brain, making possible whole new ranges of interaction
between people and the world around them. This may well explain why suddenly
new species of humans developed that lived alongside and then superseded those
that went before, as with the successive emergence of
homo habilis,
of
homo
erectus, of the various sorts of archaic human. Thus, it may well be the
case that modern humans eventually replaced the Neanderthals because they were
able to communicate more quickly and clearly with each other (although we will
probably never know for certain if this was so).
"So there has
to be a recognition of how quantity turns into quality, of how through
successive changes animal life gave birth to that new form of life we call
'human', which had a dynamic of its own, shaped by its labour and its culture
not by its genes. But this should not lead to a collapse into a new idealism
which sees culture and language (sic) as emerging from nowhere in the fairly recent
past. If such an approach is fashionable in some circles, it is not because it
can provide a scientific, materialist account of our origins, but because its
fits in with the much wider mood of the intelligentsia since the late 1970s. In
virtually every discipline there has been the attempt to separate off the
development of language and ideas from the development of material reality. As
in the days of Marx and Engels, the struggle for science is a struggle against
both idealism and mechanical materialism -- with idealism today taking the form
of 'post modernist' fashions, and mechanical materialism of sociobiology."
[Harman (1994),
pp.101-02. Bold emphases alone added.]
99.
A supporter of this site has outlined some of the weaknesses of Adaptationism -- along with references
to work written by prominent anti-Adaptationists, in response to Alex
Callinicos's eulogy of Daniel Dennett's book -- in two unpublished Essays. These
were submitted to International Socialism several years ago, but they weren't published as
part of that 'debate'. As things turned out, that was fortuitous
since those articles were rather poor (as their author now admits)! I might persuade him to post re-vamped, and
hopefully improved, versions of them at
this site at a later date.
Ultra-Darwinian Adaptationism can be found,
for example, in
Dawkins (1982, 1988, 1996, 2003, 2004, 2006, 2010), Dennett (1995), and Williams
(1966, 1996). It has been subjected to searching criticism by more than a handful
theorists; these include Eldredge (1996, 2004), Gasper (2004,
2005), Gould (1980b, 1983b, 1983c, 1990b, 1990c, 1990d, 1992, 1997a, 1997b, 1997c,
2000, 2001, 2002a, 2006b), Gould and Eldredge (1977), Gould and Lewontin (1979),
Lewontin (2000a, 2000b),
McGarr (2003), Rose (2005), Rose and Rose (2000, 2010, 2012), Sober (1993a, 1993b), and
Sober and Wilson (1998).
See also,
Dupré (2001, 2002, 2003a, 2003b, 2012),
Barnes and Dupré (2008), Gould
(1991b), Kirschner and Gerhart (2005), Kropotkin (1939), Moss (2003), Ryan
(2002) and Woolfson (2000). [Also see this
letter written by Engels.]
100.
Of course, this is to judge
these pre-/non-human 'proto-linguists' by our own standards, but, how else are we to apply
the words "understand" and "language"? If we don't intend to use these words as they are
ordinarily used, but in some specialised sense, then,
as noted above, we won't in fact be addressing understanding (or, indeed, language),
but 'understanding', a term still in want of explanation. In that case, we won't have explained
anything that is relevant to the issues at hand.
Update January 2010: It has to be
admitted that
this BBC programme throws (slightly) into doubt some of the things alleged
about dogs in this part of the Essay. Even so, it is quite clear that the
intimate and protracted relationship between dogs and human beings (probably going
back now over 20,000 years) has (artificially) selected certain behaviours which
has in turn resulted
in dogs resembling us in quite astonishing ways. [This contradicts somewhat what
I have to say in Note 101, but not in relation to the acquisition of
language.]
In addition, our use of "understand" and
related words allows for some flexibility.
[I
will say more about this in a later re-write of this Essay.]
101.
Brain size can't be the determining factor here, nor can the length of time
these animals have been in human company. As should seem obvious, cats and cows
have bigger brains than parrots, and have lived far closer to human beings for far longer
(as have rats and mice).
It
could be argued that infants don't engage in collective labour but they still
learn to speak, so why can't parrots and
other animals? But, as pointed out in the main body of this Essay, if mere
proximity to human beings could account for language, then we should expect
cats, cows, sheep, rats, mice, gerbils, fleas, bacteria and lice to be able to
communicate with us.
Clearly, being a human being is a necessary but not sufficient condition for
an individual to learn language; having certain physical capacities and
behavioural traits -- among other things, these include a
CNS, the capacity to
mimic, curiosity, and a propensity to respond as other humans do to social and
inter-personal cues and stimuli (skills and propensities which animals lack),
etc. -- are also necessary.
When these are present, a sufficient condition
for an infant to learn a language is for it to be raised and socialised in a speech community
that is based on collective labour. Parrots, dogs, and the other
organisms listed above plainly lack these necessary concomitants; moreover, they
can't engage in communal life. They may 'tag along', but that is all they
can do. This isn't so with human babies.
There is, of course, much more to learning a
language than the above suggests, but further consideration of this topic would
side-track this Essay. [On this, see
Erneling (1993), Greenspan and Shanker (2004),
and Williams (1999).]
101a. To be sure, an application Engels's first
'Law' might succeed in ruling out several of the organisms listed in the main body of this Essay -- for example, it might exclude the final four or five.
[But, parrots in fact have
smaller, less developed brains than rats or sheep.]
102.
I examine the background to much of this in detail in
Essay Seven
Part One.
Oddly enough,
Darwin also believed that dogs could understand certain words and sentences:
"The
habitual use of articulate language is, however, peculiar to man; but he uses,
in common with the lower animals, inarticulate cries to express his meaning,
aided by gestures and the movements of the muscles of the face. This especially
holds good with the more simple and vivid feelings, which are but little
connected with our higher intelligence. Our cries of pain, fear, surprise,
anger, together with their appropriate actions, and the murmur of a mother to
her beloved child are more expressive than any words. That which distinguishes
man from the lower animals is not the understanding of articulate sounds, for,
as every one knows, dogs understand many words and sentences. In this
respect they are at the same stage of development as infants, between the ages
of ten and twelve months, who understand many words and short sentences, but
cannot yet utter a single word. It is not the mere articulation which is our
distinguishing character, for parrots and other birds possess this power. Nor is
it the mere capacity of connecting definite sounds with definite ideas; for it
is certain that some parrots, which have been taught to speak, connect
unerringly words with things, and persons with events.... The lower animals
differ from man solely in his almost infinitely larger power of associating
together the most diversified sounds and ideas; and this obviously depends on
the high development of his mental powers."
[Darwin
(1871), p.54. I owe this reference to Aitchison (1996), p.79. Bold emphasis
added.]
Conditioned response is not, however, the
same as understanding.
In fact, anyone who reads this inferior work
of Darwin's will see that he held far more batty ideas than even Engels managed
to accumulate.
Apparently, a belief in parrot 'languages'
isn't just shared by some DM-fans; cf., the New Scientist,
08/11/2008, p.47, and Pepperberg (2008).
On this in general, see Bermúdez (2003), and
Griffin (2001).
[I will add several comments on these two books at a later date.]
Also
see the review of Bermudez's book in Fodor (2003). While I don't endorse much of
what Fodor says, he manages to show why representationalists can't themselves agree with many of
the conclusions drawn about 'animal thought' in Bermudez's book.
Now,
Harman seems to agree with Engels that the origin of 'thought' post-dated (or at least
accompanied) the development of language, arguing that the opposite view is in
fact a throw-back to Idealist
theories about the human mind:
"Darwin assumed
that the growth in brain size and intellect occurred before the
transition to two-legged walking and the use of hands to make tools. Engels
argued the sequence of events was the other way round. It was the freeing of the
hands that made co-operative labour possible on a scale unimaginable among apes,
and from this flowed the development of the brain. As the archaeologist
Bruce
Trigger tells:
'Darwin
was...constrained by reluctance to challenge the primacy which the idealistic
religious and philosophical thinking of his time accorded to rational thought as
a motor in bringing about cultural change. Hence in discussing human
evolution...it was the development of the brain that in turn resulted in tool
use.'
"By contrast:
'Engels
argued that an increasingly terrestrial life-style had encouraged...increasing
use of tools. This caused natural selection in favour of
bipedalism
and manual
dexterity as well as...a more complex division of labour: Tool making and the
development of a capacity for language the better to co-ordinate productive
activities led to the gradual transformation of the brain of an ape into the
that of a modern human being....'
"Darwin's view
of the sequence of stages dominated research on human origins for the best part
of a century, leading to the belief that any 'missing link' between apes and
humans had a large brain but an ape-like posture and throwing the whole study of
our evolution askew. It encouraged acceptance for some 50 years of one of the
great scientific frauds of all time -- the
Piltdown
affair, in which the skull of a man and the jaw of an ape were presented as
the remains of one of our earliest ancestors. And it led to the refusal for 30
years to take seriously a genuine find, the discovery in South Africa by
Raymond Dart
of the remains of an apelike creature which had adopted two legged walking. It
was not until the discovery by
Donald
Johanson in 1974 of a complete three and half million year old skeleton with
an ape sized brain and a erect posture that Darwin's sequence was finally
abandoned. Only then could archaeologists begin to explain the evolution of one
set of skeletons from another." [Harman (1994), pp.85-86. Italic
emphasis in the original. Formatting and Quotation marks altered to conform with the
conventions adopted at this site.
Links added. Harman is quoting Trigger (1992), p.275.
I haven't yet been able to check this source.]
This
suggests that only Idealists would want to argue that thought precedes language,
and thus that animals can think. In which case, it is perfectly understandable
that Darwin himself believed that animals could think, and that they possessed, or could grasp, some form of language, just as it is understandable that the majority of
contemporary theorists
appear to agree with him.
In
view of this, it isn't easy to explain why Engels believed that certain
animals are capable of understanding speech --
except that we already know he made damaging concessions to
Idealism by appropriating far too
many forms-of-thought
from Hegel (upside down or the 'right way up'), as well as helping concoct DM, in the
first place!
Once again, we see that ideas which properly belong to
HM have been
traduced by
those smuggled into Marxism by DM-fans.
103.
On this, see Dupré (1991, 2001, 2002, 2003a, 2003b), and Davidson (1975, 1982,
1997).
It is unfortunate that Mike Beaken's
analysis of this topic is a little too uncritical of several studies of animal 'communication'; cf., Beaken (1996), pp.44-59. In many
ways, his view of language and meaning are still stuck in a 17th
century time warp, which similarly holds much of Cognitive Science in its
grip, as has
already been pointed out.
There appear to be three competing views in
this area concerning language and thought: (i) That the difference between
primates and humans is simply one of degree; (ii) That human beings are
unique in their capacity to use language and their ability to think; and (iii) That it
makes sense to attribute to animals some capacity for thought.
I am here using the
rough classification found in Glock (2000), who defends option (iii), and
attributes it to Wittgenstein. I follow
Donald Davidson in this, who defends (ii) --, although it is also important
to add that I don't agree with
everything Davidson says. [Davidson (1982, 1997).]
I
would also want to defend (ii) partly because (i) and (iii) undermine the idea that
language and thought are the product of collective labour and communal life, and
partly for the reasons Davidson gives (even if I would
phrase the arguments differently). On this topic in general, see Radick (2007) and
Bickerton (2009).
[I hesitate to include Guldberg (2010), since
the latter contains much
that is inconsistent with the views defended in this Essay, including the unsustainable
idea that thought can be identified with "inner
speech" -- to say nothing of the offensive attitude it adopts toward animal
welfare (which is of a sort we have only come to expect from those associated with
LM magazine and Spiked).]
104. On this, see Broad and Wade (1985),
pp.110-12. The latter was written over twenty years ago, and, plainly, this area
of science has moved on significantly since then. But, many of the things Broad
and Wade had to say are still relevant. For a more recent assessment, see Pinker
(1994), Chapter 11, and Wallman (1992). Cf., also Guldberg (2010) (but read the
caveat on the latter book added to Note 103).
The survey article on 'animal
communication' at Wikipedia is also seriously compromised by the usual
confusion between signalling and linguistic communication. And, it seems, so are several of
the books and articles I have consulted on this topic -- for example, Noble (1998).
[Also
see Note 105 and Note 106,
below.]
Update
April 2010: Since writing the above, I have had the great pleasure of reading
Bickerton (2009), which subjects the idea that apes (and hence other animals) can
master language to some timely and well-aimed criticism -- pp.73-91. It is also worth noting that Bickerton's analysis of Animal Communication
Systems [ACS] is highly conducive to the approach adopted in this Essay, even if his
account of the origin of language isn't. [More about that later.]
105.
Even an otherwise sophisticated philosopher like
John Dupré (referenced above) failed to notice
the change of meaning that occurs when we use typographically identical words to refer to
animal 'communication'.
On the difference between a system of signs
and language itself, see Note 23
and Note 85, above. [Also see,
Bickerton (2009).]
106. "But, what about
dolphins?"
someone might ask. Certainly, dolphins can signal one another, as can
many other animals, but signalling isn't the same as linguistic communication. Animal signals are
highly specific stimulus response reactions. [Voloshinov half makes this point
himself; cf., Voloshinov (1973), p.101. On this, also see Bickerton (2009).]
To be sure, human beings can communicate by signalling, but
that is because they have a language into which these signals can be translated, coded and
then decoded. This isn't the
case with animals -- unless, of course, we suppose that animals have engaged in
collective labour at some point in their evolution. The same comment applies to
bees and their "dance"
--
although, interestingly here, we do have an example of collective labour!
However, according to an article in the New Scientist, recent
research has thrown into doubt much of what was once concluded about the "waggle
dance" of bees:
"When
Karl von Frisch
decoded the secret language of bees in 1946, even he
couldn't quite believe what he had found. Was it really possible for a creature
with a brain smaller than a pinhead to do something so clever? 'It is
conceivable that some people will not believe such a thing. Personally, I also
harboured doubts in the beginning,' he said in his
Nobel lecture in 1973.
"Countless experiments later, the bee's waggle dance has become an established
scientific fact. Even schoolchildren are taught that honeybees dance to tell
hive-mates about good food sources. Most researchers have long since stopped
asking whether bees communicate in this way and concentrated on working out how
the dance -- among the most sophisticated forms of animal communication outside
of primates -- evolved. In the waggle dance as described by von Frisch, a bee returning from a
plentiful food source heads for one of the hive's vertical honeycombs, where it
runs in a figure of eight. On the straight part of the run, the bee buzzes its
wings and vibrates its abdomen -- the 'waggle'.... Von Frisch's insight was
that this middle portion of the dance contains two crucial pieces of information
about the location of food.
"First of all, direction is given in relation to the sun's position. If the food
source can be found by flying directly towards the sun, the middle of the dance
is perfectly vertical. Any angle to the right or left of the sun is communicated
by running at the same angle to the vertical. Distance, meanwhile, is
communicated by the duration of the waggle. The longer the bee waggles, the
further away the food is: about 75 milliseconds is added to the waggle for every
100 metres. These two pieces of information are what von Frisch dubbed 'the
dance language'.
"The dance, which can go on for several minutes, attracts other bees, which
become increasingly excited as they watch the dancer. Once a follower has
observed five or six runs, it leaves the hive and flies directly to the food, as
if by satnav. This behaviour is seen as crucial to a hive's success. Or that's how the story goes. In recent years, some researchers have begun to
suggest that the waggle dance is too good to be true. While they accept that the
dance contains information about the location of food, they argue that its
importance has been massively overstated. A litany of recent evidence suggests
that while bees can follow the dance, they often fail to decode it properly, or
ignore it completely (Trends
in Ecology and Evolution, vol 24, p.242).
"'I think the atmosphere is changing,' says
Christoph Grüter at the
University of Sussex in Brighton, UK. 'People are much more open to the idea
that the dance language is not that important.' In one study, Grüter and his colleague Walter Farina of the University of
Buenos Aires in Argentina found that among bees that attend to a dance, 93 per
cent ignore the instructions and head to a food source they already know about (Proceedings
of the Royal Society B, vol 275, p 1321).
Similarly, bees often seem unable to follow the instructions. Some watch more
than 50 runs and make several sorties out of the hive but never find the food.
"The waggle dance also turns out to be much less important to foraging success
than has been suggested. Hives in which the honeycombs are laid horizontally,
preventing the bees from indicating direction properly, don't fare any worse
than others, except when natural food sources are severely depleted....
"Instead, Grüter and colleagues believe the waggle dance is just one component
of a more complex system for directing foraging. The dance doesn't just convey
spatial information, they say, it also passes on odour clues and generally
motivates other bees to go foraging. Bees also glean information by observing
their colleagues flying off to gather food." [Williams
(2009). Quotation marks altered to
conform with the conventions convention adopted at this site.
Several links added, and paragraphs merged.]
Of
course, the reference to "information" above must be metaphorical, since information is conveyed by indicative sentences which bees have yet to invent
-- save we mean by "information" something technical.
[On
this, see Bennett and Hacker (2008, 2021) and Bennett, et al (2007).]
Moreover, as beekeepers pointed out in subsequent letters sent in to the
New Scientist, researchers observe bees in lighted conditions, whereas in
the hive there is no light. This throws into even more doubt whether bees can
actually see these 'dances'!
It
may be wondered why a series of signals (like those found in bird song) couldn't
be seen as a language of some sort, or which can't have developed into one. That
is because, once more, there is a fundamental difference between signalling and
linguistic communication. The latter sort of communication is a social phenomenon associated with a language
into which signals can be translated. It can't work the other way round. As
such, a signalling system doesn't seem capable of evolving into a language for reasons that
are
outlined in the
main body of
this Essay. [On this in general, see Bickerton (2009), once more. See also Note 23
andNote 85
for some background details.]
107.
This topic was discussed more extensively in Essay Twelve
Part One.
See also Note 102 above.
107a. On this, see, for example,
Chase (1976), Goliszek (2003), Gould (1981), Montague (2002, 2008),
Rose, et al (1984), Rosen (2004),
and Whitman (2017). See also
this page over at Dropbox, and
this Guardian article on Hitler's debt to American Eugenicists, which
is a short summary of Black (2012).
108.
Cf., Callinicos (1996, 1997), and John Parrington's reply in Parrington (1996).
109.
Hence, it is disconcerting to see Harman endorse Chomsky's 'Nativist'
theory of language:
"The ability to use language
is, according to the generally accepted theory of Noam Chomsky, a genetically
determined feature of all modern humans. The connection between language,
abstraction and human consciousness is spelt out in the books written by the
Russian Marxist Voloshinov during the 1920s, and in part two, Labour, of the
Ontology by the Hungarian Marxist
Georg Lukács."
[Harman (2008), p.621, note 6. Harman
is here referring to Voloshinov (1973), and Lukács (1978). Harman has
made a slight error here; Labour is part three of Lukács's work.
Italic emphases in the original; link added.]
Unfortunately, Harman failed to notice that
Chomsky's belief that language is "genetically determined" completely undermines
the Marxist view that language is a
social product.
Yet, if Voloshinov and
Lukács
are correct, Chomsky's theory can't be. That is, not unless we are
prepared to revise our ideas about human development and human genetics -- rejecting
Mendelian
theory in
favour of Lamarckism. Without such a revision there seems to be no way
to account for the geneticimprinting of the linguistic gains of
one generation (whether or not these had in turn been the result of the heroic
abstractive labours of our ancestors) for the benefit of -- and use by --
subsequent generations. Of course, recent work in Epigenetics might suggest this
could happen, but it is far from clear whether linguistic development can
be imprinted in this way, or that such changes last beyond a handful of generations. Epigenetic effects seem to be environmentally
motivated, anyway. Moreover, for them to be useful, these effects would have to
be expressed across an entirepopulation,
and identically (or nearly identically) in each individual, otherwise the problems
outlined in the main body of this Essay would kick in.
Moreover,
had they been presented with them, it is a pretty safe bet that Voloshinov and
Lukács
would have wanted to know how their social and abstractionist theories could
possibly be made
consistent with Chomsky's ideas about the existence of a "transformational grammar" (now "unbounded
Merge"). In
fact, anyone who reads and then agrees with, say, Part Two of Voloshinov (1973),
would
surely want to reject Chomsky's theory. To be sure, if there were such a
"transformational grammar" "unbounded merge") then the claim that language is based on abstraction
(heritable or not) would be completely misguided. What need would there be for
any sort of abstraction if language were hard-wired in the brain? The ability to perform
such 'abstractions' would in that case be about as useful as a snorkel on a
fish.
Of course, it could be objected that
Chomsky's theory simply commits socialists to the idea that it is the grammatical
form underlying language that is subject to genetic constraint; abstraction
simply provides the content. But, as we have
seen,
abstractionist ideas make a nonsense of grammar by reconfiguring general terms
(in
predicative propositions, for example) as
abstract proper names. Far from providing content to grammatical form,
abstraction obliterates it. This would mean, therefore, that the ability
to abstract would now be about as useful to human beings as a lead-lined
snorkel is on a fish!
"If there is a single
structure of language which is inscribed in our genetic inheritance, and if all
social or cultural differences are, from [the] standpoint of language,
irrelevant, [this follows]: each member of the human species is identical as
regards the faculty of language, because language is inscribed in her brain.
Language must therefore be studied in the individual: we are no longer dealing
with a system that is external to individual speakers and independent of
them..., but with a set of individuals endowed with the same capacities; and
language, at least as conceived by the science of language, has nothing to do
with social existence. In other words, the logical consequence of Chomskyan
naturalism is
methodological individualism, which is characteristic of liberal
thinking in economics and politics.
"And there is [another]
consequence. It is clear that language, derived from a mutation that constituted
the human species, has no history, or only the quasi-frozen history of the
evolution of the species over the very long term and by leaps: human language
has no history in the strict sense, since it cannot have changed since its
appearance at the dawn of humanity. Any historical phenomenon, any linguistic
change is superficial, and irrelevant for the scientific study of the language
faculty. Or, rather, there is linguistic change, but only at the level of the
individual whose competence passes from an innate 'initial state' to a 'steady
state', once parameters have been triggered by the linguistic environment.
"The transition from infancy
in the etymological sense to articulate language is therefore not effected by
learning (or only at a superficial level); and the sole temporality of language
is the retrospective time of recollection. The child who acquires (but does not
learn) speech is like the slave in the
Meno: he
remembers what he had always known, but did not yet know that he knew. Chomsky's
position at least possesses the merit of coherence in its idealism." [Lecercle
(2006), pp.21-22. Italic emphasis in the original; links added.]
[The rest of the same chapter
continues a sharp
critique of Chomsky's entire programme.]
As John Parrington himself argued, rather
perceptively:
"Of far more
immediate importance for us is the fact that while Chomsky's theory may be
compatible with the language creativity of individuals, it is nevertheless
unable to explain where such individual creativity comes from in the first
place. In reality, the lack of space that Chomsky's theory leaves for
understanding the material source of individual creativity means that sincere
followers of his approach like
Pinker
have ended up reintroducing it in a form impenetrable to further study. Thus
Pinker believes that we think in something called 'mentalese',
which we must then translate into social language for the benefit of those
around us. But despite this being a fancy term for thought, it tells us nothing
about the material nature of thought, surely a necessity if we are to understand
consciousness.
"Both Chomsky
and Pinker are similarly incapable of explaining the influence of social change
on language and thus of the development of language over time. Both these flaws
are linked to a much deeper and more fundamental one -- the separation between
individual thought and social language that we found in the earlier theories of
language." [Parrington
(1997), p.125. Links added.]
[However, we have already seen that reliance
on Voloshinov's theory creates similar problems for Marxists -- to say nothing
of the
Neo-Cartesian ideas retailed in the above quotation.]
Unfortunately for Harman, Chomsky now
seems to believe that not only the form but the content of language is
innate. Cf., Lowe (2000), pp.188-89.] If that is so, then perhaps, like
Jerry Fodor,
he thinks that our ancestors, tens of thousands of years ago, possessed the concept "carburettor"
and "quark"!
[No joke, Fodor, for example, does indeed think this! Fodor (1998d), p.28. See also Cattell
(2006), Chapter Four. What is more,
Chomsky appears to do so, too! On this, see Chomsky (2000b), p.75,
Chomsky (2000a), p.61, and Knight (2018), Chapter 18.
Here
is Chomsky:
"Furthermore, there is good reason to suppose
that the argument is at least in substantial measure correct even for such words
as carburettor and bureaucrat, which, in fact, pose the familiar
problem of poverty of stimulus if we attend carefully to the enormous gap
between what we know and the evidence on the basis of which we know it....
However unsurprising the conclusion may be that nature has provided us with an
innate stock of concepts, and that the child's task is to discover their labels,
the empirical facts appear to leave open few other possibilities." [Chomsky
(2000a), pp.65-66. Italic emphases in the original; spelling modified to
UK English.]
In
response, Hilary Putnam pointed out that if we, or children, had such concepts
that had been fixed in the Pleistocene, then:
"evolution would have had to be able to
anticipate all the contingencies of future physical and cultural environments.
Obviously it didn't and couldn't do this." [Putnam (1988), p.15.]
Daniel Dennett added:
"Thus Aristotle had the concept of an
airplane in his brain, and also the concept of a bicycle -- he just never had
occasion to use them!... [Maybe] Aristotle had an innate airplane concept, but
did he also have a concept of wide-bodied jumbo jet? What about the
concept of an APEX fare Boston/London round trip?" [Dennett (1991),
pp.192-93. Italic emphases in the original.]
Perhaps children 75,000 years ago had the words "π
meson", "2-4-dinitophenylhudrazine",
and "Andromeda
Galaxy" in their heads, too? But, which spellings of "labour"/"labor" and
"colour"/"color" did they possess? And did they have
"K2", "Mount
Godwin-Austen" or"Chhogori"
in there, too? What about the names of non-existent objects and substances, such
as
"Phlogiston",
"Caloric"
and "Quintessence"?
And if a child is taught to say a totally made-up word, did it already exist in
her head? So was "Supercalifragilisticexpialidocious"
imprinted in the human brain back in the Pleistocene long before it was
'invented' (or, re-discovered!) in Mary
Poppins?
So
many questions, so few answers.
On Fodor and Chomsky, see Cowie (2002), Glock (2009), and Williams
(1999). For Fodor's criticism of Pinker, and much else besides, see Fodor (2001).
Pinker's views can be found in Pinker (1994, 1997, 2003, 2007). Pinker (1997)
was
critically reviewed in
Blackburn (2002), where several of his key
assumptions were shown to be woefully false.
Finally, we have already
seen
that Chomsky accepts the view that language is primarily a vehicle for thought,
not communication.
[I will add several more
comments here on Chomsky, Pinker and Fodor in a later re-write of this Essay.
However, no one should assume I agree with everything Cowie says in her
articles and books. Her philosophy is openly empiricist, which philosophy, as I
have shown (here) is both
non-sensical and incoherent.]
110.
To be specific:
Helena Cronin. [Cf., Harman (1994), p.187, n.59.] Cronin is the theorist who
(so we are told) was almost
single-handedly responsible for making EP -- the recent
reincarnation of
Sociobiology
-- 'respectable', at least in the UK. On that, see this Guardian
article (from 28/08/1999). Doubters are also encouraged to read Cronin (1993), and
then the thoroughly reactionary
Cronin (2005). A brief outline of the ideological divide between left and
right in this area can be found in
Brown (1999a), but in more detail in Brown (1999b), and Malik (2001). See also Hill (2000) for
some of the more unsavoury aspects of
EP, including
Thornhill and Palmer's outrageous theory that
rape is 'natural'!
[On this in general, see Prindle (2009). Also see Fiona Cowie's remarks,
quoted earlier.]
However, the scientific credibility of Sociobiology sank
dramatically after the publication of Kitcher (1985); EP might follow suite
because of Buller (2005) and Richardson (2007). Unfortunately, however, that
outcome is far less certain in view of the hegemonic status EP has assumed
over the last fifteen or twenty years. [Buller (2007).] In fact, since this paragraph
was first written five years ago, Buller's work has bounced off the collective
EP-skull, and seems to have had little discernible effect.
Practically every TV programme these days
that strays into the area of 'human nature' and our evolutionary past seems to take EP's nostrums
for granted -- including
Inclusive Fitness -- along with the
despicable
idea that the
human foetus and placenta, for example, are a parasites! An excellent example of this (which featured the repeated unchallenged
and off-hand reference to the foetus as just such a parasite) occurred in House,
the TV programme, Series One, Episode One,
Maternity, and in Series Three, Episode Seventeen,
Fetal Position (with an aside thrown in for good measure that a fetus is
in effect a "tumour"). [The same objectionable opinion has also cropped up, for example, in Prasad (2012), p.127.]
In Series Three, Episode Seven of the same
programme, House -- Son of
Coma Guy -- one of the characters,
Dr
James Wilson, expressed the opinion that
"We have an evolutionary incentive to sacrifice ourselves for our offspring, our
tribe, our friends, to keep them safe...". Again, this went unchallenged as if
it represented a well-established scientific fact. Indeed, the main character,
Dr Gregory
House, regularly spouts EP-nostrums as if they were scientific truth, all
unchallenged.
Update June 2011: The BBC has just
shown a documentary series of three, one hour films with the unlikely title
All Watched Over By Machines Of Loving Grace, by
Adam Curtis.
The third and final episode in the series -- The Monkey In The Machine
And The Machine In
The Monkey -- was devoted to
the work of
William Hamilton and
George
Price (two of the god-fathers of Sociobiology and EP), alongside the theory
of Inclusive Fitness. The upshot of all this was the thoroughly right-wing
conclusion that, as the opening credits revealed, "No one believes you can
change the world for the better any more." This claim was based on the supposed
fact that we are all simply survival machines for our genes, and so our actions
are ultimately determined by our genetic make-up. Apparently, so we were told,
these genes will always act to maximise their own survival at whatever the
cost. Of course, this is just a secular version of the religious myth that there
is a fixed order to nature (and our nature, too) that we can't alter, which tale
is then
augmented with another religious fable: that we are
all dominated by superhuman forces over which we have no control -- so we can just forget
about socialism, can't we?
No coincidence either that the BBC ran this
programme at the beginning of the latest round of
attacks on the Welfare State in the UK (pushed through by the
Tory/Liberal
Democrat Coalition) -- at the same time as they ran a series called
Saints and Sinners, following an earlier series called Saints and
Scroungers (repeated many times, each season featuring twenty
episodes), which drew a stark contrast between 'benefit cheats'
("Sinners"/"Scroungers") and 'charity
workers' ("Saints"). Note the overtly religious symbolism!
[In fact, the BBC regularly repeats this
series; it is now (June 2013) into its fifth season, with a sixth following on
in 2014. If programming were
correlated with the amount of money lost to the tax-payer, there would by now
have been somewhere between 75 and 450(!!) similar series on rich tax dodgers --
i.e., between 1500 and 9000 episodes! 'Benefit fraud' costs the exchequer
approximately £900 million a year (2009 estimate); in
2013, the estimate was £1.3 billion. Estimates of the size of the Tax Gap
(the difference between what the UK Treasury expects to receive each year and what
it actually collects) vary from anywhere between
£30 billion and just short of
£70 billion -- some put it at
well over £90
billion. Indeed, the European Union is reported to lose
over a trillion eurosannually in this way! Globally, estimates put
this figure at approximately
one sixth of Gross World Product [GWP] -- in 2012, GWP was put at over
$84
trillion. This means that approximately $14 trillion is lost to rich tax
dodgers and tax avoiding/evading corporations each year! The extent of these
scams was leaked in 2015 via the Panama
Papers, and then again in 2017, via the Paradise
Papers.]
The BBC has never run a series on tax
dodgers, avoiders or evaders, even though the odd programme here and there
covers this topic. The latter are
mainly short news items (which report on other groups protesting tax dodgers), or
they form part of a comedy programme (for example, a short sketch on BBC Three's,
The Revolution Will Be Televised).This
UK
Parliamentary Report might suggest a reason for the BBC's selective
blindness in this regard.
This article shows that in the UK, although tax fraud is fifteen times
greater than benefit fraud, the media give the former 600% more coverage than
the latter.
Indeed, these regressive developments in
Biology (and cultural life in general) aren't unconnected with the resurgence of right-wing,
neo-liberal ideologies over the
last 50 years -- as Fiona
Cowie pointed out, earlier. [Lewontin and Levins (2007), pp.59-63 is also particularly good
on this.]
Even so, the New Scientist ran an article on
Inclusive Fitness back in 2010:
"An expectant silence has descended on the small
room in the Royal Netherlands Academy of Arts and Sciences in Amsterdam.
Alan Grafen, a theoretical biologist
from the University of Oxford, is taking his time to set up his presentation.
When he's ready, he denounces three of his colleagues as 'unscholarly' and
'transparently wrong', and wonders what could have led such 'talented, honest
biologists' to be so 'misguided'.
"It's day one of a
meeting on the evolution of conflict and cooperation,
and exchanges are fierce. At stake is one of the pillars of modern evolutionary
biology: the theory of inclusive fitness, which explains how altruistic
behaviour can spread through a population. Altruism, in this context, refers to
any behaviour which helps the chances of survival of others at the expense of
the altruistic individual. Honeybees, which sting intruders to protect their
hive and sign their own death warrant in the process, are a classic example.
"The conference is the latest stage of a
controversy that has been raging
over the work of
three Harvard University scientists: mathematical biologists
Martin Nowak
and
Corina Tarnita, and social insect guru and father of
sociobiology
Edward O. Wilson. Last month, they published a paper
in Nature attacking inclusive fitness (vol
466, p.1057). The details of their attack are technical and
mathematical, but the consequences could be far-reaching. They say inclusive
fitness is irrelevant to the real world and want to replace it with a series of
equations that could describe the evolution of cooperation in far more detail
than ever before.
"Their statements have infuriated many of their
colleagues, including Grafen, who say their approach has just as many problems
as inclusive fitness.
The story dates back to 1955, when British
geneticist
J. B. S.
Haldane
was asked if he would risk his life to save another. He supposedly
replied that he would only do so to save at least two brothers or eight cousins,
reasoning that this would preserve enough copies of his genes to justify his own
death. This idea -- that animals are more likely to show altruistic behaviour
towards individuals they are related to -- is called kin selection.
"Haldane's colleague William Hamilton later
drafted a mathematical description of the phenomenon, known as inclusive
fitness, which assigns numerical values to the costs and benefits of an animal's
actions. In theory, inclusive fitness makes it possible to calculate the extent
of the spread of a given altruistic behaviour -- such as staying with your
parents to raise your siblings -- through a population. Hamilton's maths has
been used for decades by biologists studying cooperation in animals and was a
major inspiration for Richard
Dawkins's The Selfish Gene.
"The problem, say Nowak and Tarnita, is that the
calculations just don't work in the real world because they rely on a limiting
set of conditions that nature does not stick to. For example, they are only
valid for interactions between pairs of animals, which is fine for solitary
species whose individuals rarely meet, but no use in studying thousands of ants
sharing a colony. What's more, they do not work for populations that are under
strong pressure to evolve. These and other limitations, Tarnita says, mean
that the maths of inclusive fitness is not relevant to the real world. Instead,
she says biologists should use the models of population genetics, which focus on
interactions between different gene variants. These models avoid the messiness
of predicting the consequences of behaviour and don't require any dubious
assumptions.
"Tarnita has shown that by using standard
population genetics equations, it is possible to produce an all-encompassing
model. In Amsterdam, she excitedly explained that when she plugged Hamilton's
conditions into her model, its equations simplified to those of inclusive
fitness. Hamilton's maths, she concludes, describes a special case of a broader
model of how all behaviours evolve: it is not wrong, but limited. Nowak points out that, in thousands of insect
species, daughters leave the nest despite being as closely related to each other
as the workers in an ant colony. This suggests there is some factor other than
kin selection keeping workers in the nest and driving altruistic behaviour.
"Some biologists have embraced the new ideas:
Michael Doebeli
of the University of
British Columbia in Vancouver, Canada calls them 'a great step in the right
direction'. But according to Grafen and many others, they are nothing new. They
say theoretical biologists have always known that inclusive fitness was an
approximation, though this seems not to have filtered through to experimental
biologists, who have tended to take it as gospel. What's more, in order to use Nowak and Tarnita's
model to study the evolution of a behaviour, you would need to know an enormous
amount about the genes involved -- their identity, location and interactions.
"'Lunacy!' cries Grafen. Even if you knew all of
this, he says, it would only illuminate the process for one species. So it would
be better to stick with inclusive fitness, rough and ready though it is, because
it will enable biologists to make predictions about how various species should
behave -- and indeed already has. Many of his peers agree, arguing that
inclusive fitness should still be used as a 'rule of thumb'.
"The argument seems set to run and run. As New
Scientist went to press, more than 140 leading biologists, including several
who were present in Amsterdam, had signed a letter to Nature criticising
Nowak's paper -- though the journal would 'neither confirm nor deny' that it had
received the letter. Nowak seems to have been taken aback by the fuss, saying:
'I didn't expect our work to be so controversial.'" [New
Scientist, 09/09/2010.
Quotation marks altered to conform with the conventions
adopted at this site; some links added. Several paragraphs merged.]
[Given the reception that
challenges to Scientific
Orthodoxy have always received, it is in fact Nowak's reaction that is the surprising
feature, here!]
The above was followed a few months later by
the following interview with Nowak:
"Using mathematics to tackle some of biology's biggest questions, Martin Nowak
has concluded that an ability to cooperate is the secret of humanity's success.
He talks to Michael Marshall about drawing fire from Richard Dawkins....
"Why are you so fascinated by our ability to help each other out?
"Cooperation is interesting because it essentially means that you help someone
else, someone who is a potential competitor. You reduce your own success in
order to increase the success of somebody else. Why should you do that? Why
should natural selection favour such behaviour? To answer these questions I use
evolutionary dynamics, evolutionary game theory and experimental tests of human
behaviour.
"You say there are five different ways in which we cooperate that give us an
edge, in terms of natural selection. Tell me about them.
"The first one is called direct reciprocity. This is when individuals have
repeated interactions, so if I help you now, you may help me later. There is
also indirect reciprocity, which takes place in groups. If I help you, somebody
else might see our interaction and conclude that I'm a helpful person, and help
me later. That's a reciprocal process relying on reputation.
"The third mechanism is when neighbours help each other -- cooperators survive in
clusters. This is called spatial selection, and it plays an important role, not
only for people but for bacteria, animals and plants. Then there is
group
selection: it may be that our group of cooperators is better off than
another group of defectors: here selection acts on two levels, because in our
group there is more cooperation.
"Group selection has had a tricky reputation, and has been attacked by
evolutionary biologists. Do you think it has now been rehabilitated?
"The introduction of the concept of group selection, some 40 years ago, was
imprecise. But recent mathematical models explain very clearly when group
selection can promote the evolution of cooperation. There must be competition
between groups and migration rates should be low.
"Unless I've lost count, there should be one mechanism left.
"The last one is
kin
selection, which can occur when you help a close relative.
"You published a paper on kin selection last year that caused a bit of
controversy.
"I have no problem with kin selection when it is properly formulated. My
criticism is directed against the current use of inclusive fitness theory, which
is the dominant mathematical approach used to study aspects of kin selection.
"Can you explain?
"Inclusive fitness theory assumes that the personal fitness of an individual can
be partitioned into components caused by individual actions. This restrictive
assumption implies that inclusive fitness theory is a limited approach that
cannot be used to describe typical situations that arise in social evolution.
The standard theory of natural selection does not make such a limiting
assumption. In that recent paper we showed that inclusive fitness theory is a
subset of the standard theory.
"Inclusive fitness is a key concept of evolutionary biology. No wonder that
many biologists, including Richard Dawkins, reacted negatively when you attacked
it (New Scientist, 2 October 2010, p.8). Do you think people are now coming
around to it?
"I feel that it is beginning to be appreciated. I would say the negative response
rests on a misinterpretation of the paper. People think that we are saying
relatedness is unimportant, but this is not at all what we said.
"People who are open-minded are beginning to realise that the results of our
paper are beautiful: simple mathematical models based on standard natural
selection are sufficient to explain the evolution of
eusociality
or other phenomena in social evolution. The strange mathematical contortions of
inclusive fitness theory are unnecessary. In other words if you are interested
in a mathematical description of evolution, a situation can never arise in which
you would need an inclusive fitness approach.
"You have also been involved in some other big debates. Can you tell me about
your work on punishment?
"Many people feel that punishment is a good thing, that it leads to human
cooperation. So their idea is that unless you cooperate with me, I punish you.
It might even cost me something to punish you, but I do it because I want to
teach you a lesson. One cannot deny that punishment is an important component of
human behaviour, but I am sceptical about the idea that it's a positive
component.
"I have analysed the role of punishment using mathematics and experiments. I
think that most uses of punishment are very much for selfish interests, such as
defending your position in the group. Punishment leads to retaliation and
vendettas. It's very rare that punishment is used nobly.
"Over the years you've applied mathematics to a lot of different areas of
biology. Is it your aim to put the whole field on a mathematical footing?
"Yes. It has happened in many disciplines of science. It's a kind of maturation
process. Without a mathematical description, we can get a rough handle on a
phenomenon but we can't fully understand it. In physics, that's completely
clear. You don't just talk about gravity, you quantify your description of it.
The beautiful thing about mathematics is that it can decide an argument. Some
things are fiercely debated for years, but with mathematics the issues become
clear....
"Profile
"Martin Nowak is professor of mathematics and biology at Harvard University. He
has a PhD from the University of Vienna, Austria, became professor of
theoretical biology at the University of Oxford aged 32, then moved to Princeton
University and later to Harvard. His book SuperCooperators: Altruism,
evolution, and why we need each other to succeed, co-authored with Roger
Highfield, New Scientist's editor, is out this month." [Italic
emphases in the original; some links added. Bold added.]
[The book mentioned above is in fact Nowak
and Highfield (2012). (See the Review in
Appendix A.) I hasten to add that I have quoted these passages from the
New Scientist not because I agree with
everything they contain (indeed, Nowak specifically accounts for
cooperation and altruism in Neo-Darwinian terms, which, of course, means that there
is no such thing as altruism -- as we might ordinarily understand this term --,
there are instead more complex mechanisms that ensure the maximal survival of
offspring/genes; on this, see Stove (2006), pp.115-247), but because it shows that the theory
that is current orthodoxy among EP-ers is likely to go the way of most other theories we have witnessed in the
history of science -- i.e.,
down the tubes.]
On EP, see also Buller (2005, 2007), Dupré (2001, 2002, 2003a,
2003b, 2012),
Barnes and Dupré (2008), Gasper (2004,
2005), and
Richardson (2007).
Of course, the 'philosophical foundations' of the collective slander on the
human race that is EP had already been exposed, and then demolished, in Stove
(2006). [Readers unfamiliar with David
Stove's writings should, however, read
this warning.] See also the works listed in Note
99, above. Earlier, left-leaning attacks on Sociobiology can
be found in, for example, Arditti, Brennan, and Cavrak (1980), Caplan (1978),
and Sahlins (1976).
Chomsky
is himself ambivalent about the
possibility of there being a Darwinian account of the origin of "grammar"
-- in fact, and from what little he has said, it seems his view of the origin of
language resembles the fable recorded in the Book of Genesis. Whereas that myth sees language as a gift from 'God', Chomsky perhaps sees it as a gift from outer space,
viacosmic rays! To be sure, Chomsky did add this rider: "This is a story not to be
taken literally" (a bit like the Bible story, too, then!), but he went on to
add the following:
"But it may be closer to
reality than many other fairy tales that are told about evolutionary processes,
including language....
"Specifically, the theory
that some time ago there were primates with pretty much our sensorimotor and
conceptual-intentional systems, but no language faculty, and some natural event
took place that brought about a mutation that installed a language faculty. Say,
a cosmic ray shower, or something that took longer, like the process that caused
a bone of the reptilian jaw to migrate to the inner ear, where it is wonderfully
designed for the use of language -- apparently something that has been going on
for about 160 million years....
"How did the basic principles
get into the genetic programme? Such questions go vastly beyond current
understanding, not just for language, but even for much simpler biological
systems." [Chomsky (2000b), pp.4, 61-62.]
So, this is in fact a "fairy tale"
that Chomsky seems to take
pretty seriously, despite his disclaimer. [On this unfortunate use of the word
"designed", see Jerry Fodor's comments,
quoted earlier. Chomsky's theory
of the 'semi-miraculous' origin of language, via a massive mutation of some
sort, certainly helps account for Fodor's
criticisms of natural selection,
since that process can't account for rapid, massive changes like this.]
Be this as it may, we
have evidence from the
fossil record that
charts the changes in the inner ear that Chomsky mentions, but
where is the evidence for the massive mutation that led to the formation
of the "language faculty"? [The first of these links to a PDF.] Naturally, exactly the same mutagen with the
same
effects must have been present and operative on several primates (or
proto-humans) all at
once, or this new faculty would have been useless. This is, of course,
independent of the insurmountable difficulties
--
rehearsed earlier in this
section -- faced by this approach to giving a Darwinian account of the origin of language.
It could be objected that
the ability to
think -- or, rather, the ability to use language, which resulted from this massive mutation
-- would have helped the
individual concerned survive much better and thus pass on more of his/her genes.
Hence, these genes would enter into and pass through the local population,
improving its fitness. However, as pointed out in the main body of this Essay,
such an individual would have woken up one morning uttering strange noises she
couldn't understand (rather like Brother John, mentioned
earlier), or she would begin
'thinking' strange 'thoughts' she had never encountered before and could not,
therefore, comprehend. These days, this individual would be classified as mentally
ill. So, if anything, this would hinder her survival.
Imagine how you, dear reader, would respond
if, one day, you woke up and began to 'think' in a language you didn't
already know (say
Navajo,
or even,
Klingon). You would most likely conclude (that is, if you could even understand yourself,
or work out what the hell was going on!) that you were going mad. But, you
already have the advantage of knowing a modern language (English!), and
people around you who speak a language. This
poor proto-linguist didn't have a language the day before it was stricken in this way.
How would or could he/she cope?
How, too, would those around you (friends,
relatives, neighbours, co-workers, strangers, etc.) respond today if you began to speak,
say, in Klingon, and nothing but Klingon. [I am, of course, assuming that
you don't belong to a Star Trek
fan club, and that none of those around you understand Klingon! Otherwise, you can
substitute for it any other language not known to you or to those around you.] Would that
improve your chances of finding a mate with whom you could reproduce? [Even
Star Trek geeks would tend to shun you if you only ever spoke in Klingon
and nothing else.
It isn't easy chatting up a member of the opposite sex if all you can do it
mutter incomprehensible noises at them.] Clearly, to
ask such questions is to answer them. But, unlike you, dear reader (whether or
not you spout Klingon or some other little known tongue) this afflicted creature would be
surrounded by those who had no language at all. If anything, they would run
away from this odd individual, screaming. Alternatively, they would simply banish,
ostracise, or even kill her/him. There wasn't much 'care in the community' one hundred
thousand years ago. Precious little mating would be going on with
an individual that socially-challenged,
either. This is quite apart from the fact that
most mutations are harmful. That is even more true with respect to large
mutations. Moreover, far as I am aware, no mutation of this sort has ever been
observed in human beings. Finally,
Cosmic Rays are hardly conducive to human health and well being.
However, with respect to the supposed Darwinian
origin of language, Chomsky's 'followers'
are much less cautious. On that, see Pinker (1994), Bickerton (1990, 1995),
and Calvin and Bickerton (2001). [Bickerton has in fact changed his mind
somewhat of late; see Bickerton (2009).]
[I will add several
comments about several of the above authors in a later re-write of this Essay.]
Perhaps the tide is beginning to turn against
Chomsky's views. For example, here is the opening paragraph of a recent review
of one of Chomsky's latest books:
"The Science of
Language,
published in the sixth decade of Noam Chomsky's
linguistic career, defends views that are visibly out of touch with recent
research in formal linguistics, developmental child psychology, computational
modeling (sic) of language acquisition, and language evolution. I argue that the
poor quality of this volume is representative of the serious shortcomings of
Chomsky's
recent scholarship, especially of his criticism of and contribution to debates
about language evolution. Chomsky creates the impression that he is quoting
titbits of a massive body of scientific
work he has conducted or is intimately familiar with. Yet his speculations
reveal a lack of even basic understanding of biology, and an unwillingness to
engage seriously with the relevant literature. At the same time, he ridicules
the work of virtually all other theorists, without spelling out the views he
disagrees with. A critical analysis of the
'Galilean method'
demonstrates that Chomsky uses appeal to
authority to insulate his own proposals against falsification
by empirical counter-evidence. This form of discourse bears no serious relation
to the way science proceeds.
"The
Science of Language: Interviews with James McGilvray, henceforth The
Science of Language, published in the sixth decade of Noam Chomsky's
linguistic career, should have been an impressive summary of the achievements of
one of the greatest intellectuals of our time. It is not. Chomsky's scholarship
has arguably been slowly deteriorating over decades, and this volume is
altogether representative of the problem. I shall argue in this review article
that uncritical acceptance of Chomsky's work despite its steadily declining
quality has been doing the field considerable damage."
[Behme (2014b),
pp.671-72. The work being reviewed is Chomsky (2012a). See also,
Pullum (2012).]
110a.
By this I don't mean that genes have absolutely nothing to do with
language (since that would be absurd!), only that any attempt to reduce language
to our genetic makeup is thoroughly misguided. However, the fine detail
connected with these claims will emerge as this main sub-section unfolds.
Admittedly, Harman's later remarks are
exceedingly brief, but they do express the state of current orthodoxy in the relevant
sciences (at least, as it developed throughout the 1990s), and so will be examined for that reason alone.
[On the early stages of this change, see Harris (1995).]
111.As
noted earlier, several rival theories will be examined here at a later
date.
112.Unfortunately, this is already a contestable move! In order for a sound to
count as a word (as opposed to a mere noise), let alone a name, a
sophisticated linguistic and social background must already be in place.
[See Essay Three Part One,
and Stroud (1991, 1996).] In
that case, all the serious problems I am about to highlight concerning the
hypothesised transition from the use of 'names' to the use of 'names' and 'verbs' apply equally, if not more so, to
the invention of 'names'. [On this see Baker and Hacker (2005a), pp.113-28, 227-49,
Schulte (2009), and Hanna and Harrison (2004), pp.63-158.
However, I hesitate to recommend wholeheartedly the latter work
since its authors have adopted the misguided 'causal
theory' of names.]
See also Note 118a, below.
Of
course, the material in the main body of this Essay doesn't pretend to be a summary of Chomsky's views,
merely of
the implications of Harman's ideas.
Even
so, the same sort of questions
could be asked of Chomsky's sketchy theory.
As noted above, he seems to believe that a language gene emerged
(relatively) suddenly in an individual, not a group -- and was fully formed,
too, like
Athena
bursting forth
from the head of Zeus --, which later spread through the relevant population, and which was
then co-opted to form a full-blown language. [On that, see Chomsky (2007), and
here. See also Note 114a, below.] It is relatively clear from this that Chomsky believes this
development was originally a
pre-adaptation of some sort, or perhaps even an
exaptation.
[See also here.]
In that case, many of the conclusions reached in this part of the Essay will
also apply to Chomsky's theory -- perhaps more so, since, for Chomsky, an entire
language (and not just verbs and nouns) seems to have emerged suddenly, all at once!
113.
W1 is assumed
to contain only 'nouns', whereas W2
is limited to 'nouns' and 'verbs' so that several historical and conceptual
possibilities may be explored. It isn't being suggested here that
language actually developed like this, or that the use of 'nouns' in fact
preceded that of 'verbs' -- or even that we will ever be able to say what
actually happened. Indeed, describing a system that uses only 'proto-nouns' as a
language is bizarre in itself. [Incidentally, the words "noun" and "verb" are in
scare quotes since it is controversial whether either could be so described
without the other, or even that they can be pictured this way in a proto-language.
Much of this was covered in Essay
Three
Part One.]
114.
Recall that MM was the first to innovate in their group. If so, how would MM
even understand him/herself?MM would wake up one morning uttering
strange sounds they had never heard before. MM would thus confront
him/herself as a foreign language speaker now confronts us today, only much more so
since we alreadyhave a language. No one had a language back then. [However, see Note 117a.]
114a0.
Of course, mutations can spread through a population quite rapidly if there is
horizontal gene transfer, or the population descends from a
single organism or from several that possess the mutation -- as happens, for
example, in
bacterial populations -- but that is a totally different matter. Here, we are
postulating a set of simultaneous, identical massive neural and
psychological, and possibly anatomical changes induced in a group of highly
complex organisms. As far as I am aware, this has never been observed anywhere. Nor is it
credible.
114a.
In fact, as we have seen, Chomsky argues that this genetic change appeared in a single individual
suddenly. This happy individual somehow had the capacity to form complex thoughts (but, how this
is even
conceivable Chomsky left
entirely mysterious (on this, see
Behme
(2014b), pp.766-69)), which gradually spread through the breeding
population:
"The core principle of language,
unbounded Merge, must have arisen from some rewiring of the brain,
presumably the effect of some small mutation. Such changes take place in an
individual, not a group. The individual so endowed would have had many
advantages: capacities for complex thought, planning, interpretation, and so on.
The capacity would be transmitted to offspring, coming to dominate a small
breeding group...." [Chomsky (2007), pp.23-24.]
"Small mutation"? Really? This 'just-so' story is even worse than
Harman's! Quite apart from the fact that the origin of complex thought in a single individual
isn't possible (the supposition that it might be is reminiscent of the
Robinsonades
that Marx lampooned),
its sudden and miraculous emergence is no less implausible.
Indeed, Terrence Deacon had this to say about
such fanciful ideas:
"This reminds me of a
wonderful piece of modern mythology from a recent film entitled
Short
Circuit. A sophisticated robot is suddenly transformed from a mechanism
that 'just runs programs' into a conscious, self-aware being as a result of
being struck by lightning. The power surge damaged its circuits in just the
right way.... As a cinematic device, the bolt of lightning accomplishes two
important things. The catastrophic and unpredictable nature of lightning
provided the vehicle for invoking drastic and unprecedented change, and its
intrinsically chaotic -- and by tradition, miraculous -- character obviates any
possibility of describing exactly what alterations changed a computer mechanism
into a human-type mind.... As an allegory of human mental evolution, it offers a
paradigm example of what biologists call 'hopeful monster' theory [which is now one
aspect of 'Saltation'
theory -- RL]: the evolutionary theorist's counterpart to divine intervention,
in which a freak mutation just happens to produce a radically different and
serendipitously better-equipped organism.
"The single most influential
'hopeful monster' theory of human language evolution was offered by the linguist
Noam Chomsky...." [Deacon (1997), p.35. Quotation marks altered to
conform with the conventions adopted at this site. Italic emphasis in the original.
Links added.]
As we have already seen, Chomsky appeals to an
even more surreal agent of this fortuitous mutation, cosmic rays!
"But it may be closer to
reality than many other fairy tales that are told about evolutionary processes,
including language....
"Specifically, the theory
that some time ago there were primates with pretty much our sensorimotor and
conceptual-intentional systems, but no language faculty, and some natural event
took place that brought about a mutation that installed a language faculty. Say,
a cosmic ray shower, or something that took longer, like the process that caused
a bone of the reptilian jaw to migrate to the inner ear, where it is wonderfully
designed for the use of language -- apparently something that has been going on
for about 160 million years....
"How did the basic principles
get into the genetic programme? Such questions go vastly beyond current
understanding, not just for language, but even for much simpler biological
systems" [Chomsky (2000b), pp.4, 61-62.]
Of course, only those
who confuse science fiction with genuine science will find the 'miraculous
origin of language theory' plausible -- on a par with events in
Mary
Shelley's
Frankenstein, Or The Modern Prometheus. [The 'monster' stitched together in this
novel, as it has been depicted in many films since, was also struck by
lightning, which miraculously brings it to life.
Chapter Two of the novel in fact hints that electricity is the vivifying
power at work in this case.] This is quite apart from the fact that Chomsky
offers no evidence that a single mutation (or set of them) has ever had, or
now has had, or is having, such profound consequences. [On this, also see Note 117.]
Independently of the
points raised in the main body of this Essay (and
earlier) -- concerning the
fact that such an individual would find him/herself uttering strange noises or
experiencing novel 'thoughts' one day that neither they nor anyone else would or
could understand (in which case, "planning" would be of little use because
these
"plans" couldn't be communicated to anyone), hence they would
have no 'survival value' --, this move by Chomsky
constitutes a major concession to Idealism since he has 'thought' preceding
language. For Chomsky, that might not be a decisive, or even a relevant, objection, but no Marxist could agree with
him.
It might be
countered that for Chomsky, thought
doesn't precede language, it is coincident with it. In fact,
Chomsky argues as
follows:
"If so, then it
appears that language evolved, and is designed, primarily as an instrument of
thought. Emergence of unbounded
Merge in human
evolutionary history provides what has been called a 'language of thought,' an
internal generative system that constructs thoughts of arbitrary richness and
complexity, exploiting conceptual resources that are already available or may
develop with the availability of structured expressions. If the relation to the
interfaces is asymmetric, as seems to be the case, then unbounded Merge provides
only a language of thought, and the basis for ancillary processes of
externalization." [Chomsky
(2007), p.22. This links to a PDF, access to which is open, but requires free
registration.]
"It seems that
we must either deprive the notion 'communication' of all significance, or else we must
reject the view that the purpose of language is communication.... It is
difficult to say what 'the purpose' of language is, except, perhaps, the
expression of thought, a rather empty formulation." [Chomsky (2005a), p.230. I owe this reference to
Millikan (2005b), p.24. However, Millikan must have been using a different
edition, since my copy reads as I have reproduced it above, but not as Millikan
presented it. Bold added. Quotation marks altered to conform with the conventions
adopted at this site.]
From this it is clear that Chomsky
does allow
for the appearance of conceptual resources (and thus for rudimentary 'thought')
before language developed, or emerged, in the first individual to possess
it (i.e., linguistic Adam). It is also hard to see how language could be an
"instrument of thought" if 'thought' wasn't already present to be
expressed by this medium, or which needed an 'instrument' in order to do just that.
And this is exactly what we find; Chomsky imagines a socially isolated primate
wandering around before it was zapped by radiation:
"It lacks the language organ, but it has
something like our brain and other organs, including sensorimotor systems
sufficiently close to ours, and also a conceptual-intentional system
sufficiently close to ours so that it can think about the world more or less
the way we do, in so far as that is possible without language. But it
doesn't have language and cannot articulate such thoughts -- even to itself."
[Chomsky, quoted in Knight (2010a). Bold added.]
Exactly what such language-less thoughts
could be was left mysterious.
Chomsky also believes that the possession of
language would confer on a lone speaker an adaptive advantage:
"Actually you can use language even if you
are the only person in the universe with language, and in fact it would even
have adaptive advantage. If one person suddenly got the language faculty, that
person would have great advantages; the person could think, could articulate to
itself its thoughts, could plan, could sharpen, and develop thinking as we do in
inner speech, which has a big effect on our lives. Inner speech is most of
speech. Almost all the use of language is to oneself, and it can be useful for
all kinds of purposes (it can also be harmful, as we all know): figure out what
you are going to do, plan, clarify your thoughts, whatever. So if one organism
just happens to gain a language capacity, it might have reproductive advantages,
enormous ones. And if it happened to proliferate in a further generation, they
all would have it." [Chomsky (2002), p.148.]
Despite what Chomsky says, it is hard to see what use language would be to a
lone proto-human. If it was alone, it wouldn't be able to compete with social
units of other, languageless proto-humans. As both Hume and Rousseau pointed
out, cooperative behaviour offers far greater advantage over a 'lone wolf'
enterprise. [On this, see Skryms (2004).] It could be argued that language would
make up the difference, conferring on an individual language user advantages
that far out-weighed co-operative action. Well, this is all highly speculative
and ignores the points made
earlier this Essay:
It could be objected that
the ability to
think -- or, rather, the ability to use language, which resulted from this massive mutation
-- would have helped the
individual concerned survive much better and thus pass on more of his/her genes.
Hence, these genes would enter into and pass through the local population,
improving its fitness. However, as pointed out in the main body of this Essay,
such an individual would have woken up one morning uttering strange noises she
couldn't understand (rather like Brother John, mentioned
earlier), or she would begin
'thinking' strange 'thoughts' she had never encountered before and could not,
therefore, comprehend. These days, this individual would be classified as mentally
ill. So, if anything, this would hinder her survival.
Imagine how you, dear reader, would respond
if, one day, you woke up and began to 'think' in a language you didn't
already know (say
Navajo,
or even,
Klingon). You would most likely conclude (that is, if you could even understand yourself,
or work out what the hell was going on!) that you were going mad. But, you
already have the advantage of knowing a modern language (English!), and
people around you who speak a language. This
poor proto-linguist didn't have a language the day before it was stricken in this way.
How would or could he/she cope?
How, too, would those around you (friends,
relatives, neighbours, co-workers, strangers, etc.) respond today if you began to speak,
say, in Klingon, and nothing but Klingon. [I am, of course, assuming that
you don't belong to a Star Trek
fan club, and that none of those around you understand Klingon! Otherwise, you can
substitute for it any other language not known to you or to those around you.] Would that
improve your chances of finding a mate with whom you could reproduce? [Even
Star Trek geeks would tend to shun you if you only ever spoke in Klingon
and nothing else.
It isn't easy chatting up a member of the opposite sex if all you can do it
mutter incomprehensible noises at them.] Clearly, to
ask such questions is to answer them. But, unlike you, dear reader (whether or
not you spout Klingon or some other little known tongue) this afflicted creature would be
surrounded by those who had no language at all. If anything, they would run
away from this odd individual, screaming. Alternatively, they would simply banish,
ostracise, or even kill her/him. There wasn't much 'care in the community' one hundred
thousand years ago. Precious little mating would be going on with
an individual that socially-challenged,
either. This is quite apart from the fact that
most mutations are harmful. That is even more true with respect to large
mutations. Moreover, far as I am aware, no mutation of this sort has ever been
observed in human beings. Finally,
Cosmic Rays are hardly conducive to human health and well being.
115.
Unless, of course, we suppose these innovators invented each letter (or
lexeme, or whatever)
one at a time!
116.Once more, this doesn't
amount to an admission that I think this was
the case, or
that I think genes 'control' language; it is simply an attempt to explore where
this argument might reasonably be expected to go.
117.
At best, this is all mere speculation; we have no idea
whether or not it is true. On this, see
Gould and Vrba (1982), and
McGarr (2003), pp.98-100. See
also,Pinker and Bloom
(1990).
117a.Admittedly, given the liberal assumptions allowed for in
Note 112 (and
here), some of the sounds made by
one or more individuals would
be 'recognisable' (i.e., perhaps the set of 'nouns'), but any new 'words', or
any new uses, wouldn't.
If those assumptions are tightened, or dropped, then even this would be unlikely.
Someone could argue that
children face these problems all the time, but manage to bumble through
somehow. The difference here is that children are born into a community that
already has a language; this wouldn't have been the case with these
proto-linguists, where no one would recognise most (or any?) of the sounds they heard
given this scenario. As they grow and develop children are trained or socialised by their carers and peers
to use sounds (to which they will have become accustomed in infancy) to say
things for themselves. Again, this can't have been the case with respect to these
proto-linguists. [However, how children actually learn their first language will have to be
set to
one side for now; on this see Erneling (1993), Greenspan and Shanker (2004), and
Williams (1999).]
117a0.
To be sure, this doesn't imply
that these hypothetical rules
must be explicit or even known to those who employ them (in the sense that they
are able to
quote them); all they need do is recognise the correct or
incorrect use of such words, or be capable of correcting one another (to some extent). For example, whether
or not any of us are qualified grammarians or linguists, the vast majority of us would
quickly spot this clear misuse of English:
M1: "Ok, it's agreed. We'll
meet at the station somewhere between 3 o'clock."
Nor does
this point imply that we are all expert
grammarians. But, some of us are, and we often look
to them for guidance. [On this, see Baker and Hacker (1984a).] Nor does it mean
that there is such a thing as perfect English. There are clearly
regional, ethnic, and class-based variations, just as there are well established dialects, but
all English
speakers faced with M1 would be puzzled by it. They would ask such things as "Don't you mean 'We'll meet at
3 o'clock'?" Or, "Do you mean 'We'll meet between, say, 3 and 4 o'clock'?" No
amount of regional or class-based variation would make M1 legitimate or
comprehensible --, that
is, without changing the meaning of, say, "between", so that it became the
equivalent of "at".
Furthermore, there would be no such thing as the
correct use of such words in the absence of a community that had already
grounded these rules in social practice. [This takes care of regional, ethnic
and class-based differences over how people use language.] To be sure, in working
class -- and many ethnic -- communities we often encounter sentences like this:
M2: "I ain't done
nothing!",
where it is quite clear that this is a way of
emphatically denying or rejecting something. [One example I saw in a TV
film recently went as follows "I don't know nothing about nobody".] No one from such a community would
think to correct such a use of these words since everyone understands what is meant. In this
instance,
double negation is being employed as other communities might use an emphatic "anything". We have to
view it this way, otherwise the double negative would imply the opposite
of what M2 was meant to repudiate!
But, no one from a working class community
(or any other, for that matter) would come out with this:
M3: "I didn't steal the
watch, officer. I paid nothing for it and left the shop with it in my bag!"
Which shows that we all know how not
to use this word. They would perhaps come out with M3a:
M3a: "I didn't steal the watch,
officer. I paid £50 for it and left the shop with it in my bag! Here's the
receipt."
Consider this conversation:
M4: MM -- "I know that Tony
Blair."
M5: MN -- "You know what
about Tony Blair?"
M6: MM -- "Nothing, I just
know that Tony Blair."
M7: MN -- "You aren't making
sense. You can certainly know Tony Blair (be acquainted with him, or be
friends with him), but if you are going to use a 'that' you must say what it is
that you know about him by using a sentence or clause, such as 'I know that Tony
Blair is a warmonger', or some such."
M8: MM -- "I can use words in
any way I please."
M9: MN -- "You most certainly
can, but you can't expect to be understood if you insist on using words in such a
puzzling way."
Few would disagree that
MM's use of English
here is incomprehensible, even if they might not always be able to say how it might be
corrected. [But, then again, few would object to the above
correction.]
In addition, all of us would react the same
way in the situation described in Essay Twelve Part One:
An example
taken from Wittgenstein's
Philosophical Investigations illustrates the radical difference between
number words and other terms we use (which, incidentally, also exposes one of
the core confusions motivated by
Semiotics -- that all words are signs, or operate as
"signifiers" of the "signified"; on this see Essay Thirteen
Part Three).
Wittgenstein encourages us to consider
an example where a customer enters a grocery
shop and asks the shop assistant for five red apples. The assistant
doesn't first go off in search of red things, nor yet collections of
five things. Manifestly, he or she will go and find apples first, or even red apples, and then count them.
This forms part of the
Fregean idea that number
words attach to concepts, not objects. Or, as Wittgenstein might have said,
number words express operations carried out on objects of a certain sort,
qualified by a
count noun -- like "three apples" or "five pears" (although, as far as I am
aware, Wittgenstein didn't use the phrase "count noun"
-- he did use a roughly equivalent term "substantive",
though).
Hence, the
assistant
will count apples: one apple, two apples..., and so on, as the concept
expression "ξ is an apple" is successively instantiated or
applied -- sometimes expressed demonstratively (typically to children) as: "This
is an apple, and that is another...".
Of course, this isn't to suggest that these are the words that this fictional
assistant will
actually use, or indeed that he/she will use any words at all, but they, or
words like them, will have been used in her/his childhood training, at
some point. No
one is just taught to count 'objects' -- but to count objects of a certain sort,
or objects identified demonstratively, governed by the use of concept expressions
(like "ξ is an apple"), or count nouns (i.e., "napples").
Novices who can
proceed along lines they have been trained are thus said to have grasped the use of number
words (and,
indeed, of concept expressions and/or count nouns). Subsequently, this linguistic skill
becomes automatic, which is indeed part of what we mean by "knowing how to count" --
or even how to serve in a grocers shop! [On
this, see Robinson (2003b). The use of Greek symbols, like those employed above, is explained
here.]
[This isn't to suggest,
either,
that knowing (implicitly) how to apply number words is sufficient to be
able to credit an individual with a minimal grasp of the concept of number. As
is well known (at least since Frege (1953) -- and as is implied by the above
comments), this requirement needs supplementing with what is called a "criterion
of identity" (that is, the individual concerned must be able to specify
whether or not, in this case, there are the same number of apples (or,
indeed, red apples) each time. That is, they must be proficient with the
practical and not just verbal
application of "same apple", i.e., with what counts as the same (sort of) apple. Cf.,
Wittgenstein (2009), §1,
pp.5e-6e.
(This links to a PDF.) See also,
Geach (1968), pp.39-40
this in fact links to the 3rd
(1980) edition, so the page numbers are different:
pp.63-64), Lowe (1989), and
Noonan (2014). For some of
the complexities involved in this area, see Epstein (2012).]
Now, the
whole point of this analysis is aimed at showing that not all words are names
and not all words function in the same way -- and, eo ipso,
that words can't be "signifiers" of the "signified" -- otherwise, the order in which
the above grocer looked for the items required by this customer would be
indifferent, and he/she could or would look for five things first, red things next,
apples last.
In
addition, it is also aimed at demonstrating
that we all know this to be so (i.e., in our practice -- in, say, our
automatic reaction to requests like the one the shopkeeper faced
--,
but not necessarily in our deliberations about such things, where we
often go astray). And, that is
why (whatever philosophical theory we hold, whatever ideology we assent to) not one of us
would dream of looking for something named by "five" first, or even "red",
and then "apples" last. On the other hand, if all words were names, we would typically
do this.
This
alone shows that Wittgenstein wasn't fixated on ordinary German (or even ordinary
English). No human being who has ever walked the planet would dream of
looking for something 'named' by "five" first, or even "red", and then "apples" last (always assuming they lived in a society with the requisite
social organisation and vocabulary,
etc.), whatever their language, social circumstances or ideological
commitments happened to be.
Now,this
is what Wittgenstein meant by "logical grammar": logical features expressed in
language, reflected in our practices, which illustrate how we all
react in social circumstances (or otherwise),
no matter what ideology or theory we
subscribe to, and no matter in what century we actually live. Indeed, they are so much part of our second nature, so much
part of what we do without thinking, that we fail to spot their significance --,
which is, of course, why they went unremarked upon for millennia (until Frege and
Wittgenstein pointed them out).
[Details concerning the above references can be accessed
here.]
So, we can safely predict that if we
randomly select, say, a member of the working class, or a billionaire, or a member of the
Central Committee of the UK-SWP, a Bishop, an
Imam, a Zen Buddhist, the editor of the Daily
Express, or a Conservative politician, and told them to do the following:
M10: "Go to your local park
and count all the objects there",
not one of them would know where to
begin, never mind having a clue what they had been asked to do. Whatever class we come from,
we all know how to use the word "object" (in such contexts, in our ordinary lives
-- but
maybe less so in our theoretical deliberations), and that would prevent us from
understanding how to proceed in the above case. None of us would need a grammarian, a
linguist, a logician or a psychologist to tell us that we can't just count
objects. All of us would know that M10 represents a radical misuse of
language. The same applies to speakers of all known languages that possess the
requisite vocabulary, at any point in human history. No Daoist monk, Roman
Centurion, Amazonian Indian, Medieval Knight, Chiricahua
Apache warrior,
Visigoth
King, student of Hegel (or, indeed, Hegel himself),
Cathar
'heretic', Bolivian miner,
Klondike gold prospector, member of the Central Committee of
the Bolshevik Party in October 1917, Babylonian Astrologer,
Christian Science 'healer', Greek slave, Peruvian
Shaman, Egyptian Pyramid
builder,
Medici Pope, Russian gangster, Vietnamese peasant, Chinese Emperor, Donald J
Trump supporter -- or even
Bob Avakian -- would know how to count
just objects, or even what was meant by a request that they should try to do
it. They might not be able to quote a grammatical, or even logical, rule that
prevents them from dong it, but they would all know such a task was impossible even to understand, let
alone carry out.
Which is, of course, why our ancestors
invented count nouns.
Some might think that they can, indeed, count
just objects.
For example, a tree is an object, so is a park bench. But are they? Is a
tree one object or many thousand? Do we count the leaves as objects, the bark,
each branch, each atom, each electron...? Notice, too, that in order to tell us
what is meant by an "object", these hypothetical 'object counters' would
find they had to
use count nouns to make themselves understood -- for example, by employing "tree" or "park bench"
(both of which are count nouns).
Indeed, one of the reasons why we have
such nouns in the language is to make counting instructions and requests clear. If someone now said:
M11: "Go to your local park
and count the ducks",
few of us would be puzzled. But they would be
if this were requested:
M12: "Go to your local park
and count the soil."
"Soil" is a
mass noun,
and so we all immediately recognise the ridiculous nature of M12. While most
speakers might not
be acquainted with the terms "mass
noun" or "count noun", they will all generally recognise
when they should and shouldn't be used. So, we would all respond to M12 with something like: "Don't you mean buckets
of soil?", "I think you mean 'piles of soil'", or "You can't count soil!
Don't you mean I should weigh it? What, all of it!?"
Of course, "bucket" and "pile" are count
nouns. To be sure, if a language lacks this distinction, then the individuals
using that language won't be able to form sentences like these or entertain such
(slightly more complex) thoughts. But, that is a separate matter.
Having said that,
the word "amount" now seems to be taking over from "number"; so we find ordinary
speakers saying things like "The amount of people who have caught 'flu this
winter...", or "The amount of books in the library continues to
fall...". Even politicians talk like this as do academics. Of
course, "amount" goes with mass nouns and "number" with count nouns. So, we might be losing a clear
distinction 'around the edges', so to speak. This isn't a pedantic point, either; much of
Statistics is based on
a clear distinction between discrete and continuous variables, that is,
on terms roughly covered either by count nouns or mass nouns. That branch of
mathematics would be crippled if this distinction were abandoned or blurred.
And it's not just mathematics that would suffer.
The importance of maintaining a clear distinction between our use of "number" and "amount"
is perhaps best illustrated by a plausible everyday example:
D1:
Doctor to patient, "I am worried about the amount of drugs you are taking."
["Drug" as a mass noun.]
D2:
Doctor to patient, "I am worried about the number of drugs you are taking."
["Drug" as a count noun.]
While
D2 necessarily implies the individual concerned is taking a range of different
drugs, D1 doesn't automatically carry that implication, and, if anything, implies
only one drug is being consumed, albeit in large quantities.
Any doctor who ran these two uses together, or was
oblivious of the difference, could cost someone their health or even their life.
So, while common use is eroding this difference, it is important that such
erosion, at least here, is resisted.
Similar
comments apply to the now almost universal erosion of the distinction between
"less" and "fewer":
D3: Doctor to patient, "I am worried about the
amount of drugs you are taking, so please take less." ["Drug" used as a mass
noun.]
D4: Doctor to patient, "I am worried about the
number of drugs you are taking, so please take fewer." ["Drug" used as a count
noun.]
Again, one of these might save your life, the other
might kill you.
Finally, the idea that such individuals
possess or follow rules 'unconsciously' (that is, they employ rules that have been 'wired' into
their
brains, a là Chomsky) makes even less sense. On that, see
Baker and Hacker (1984a), pp.243-368.
[The latter should, however, be read in
conjunction with Bloor (1997), Kripke (1982), Kusch (2006), and Robinson
(2003c).]
It could be argued that our language affects
the way we view the world, perhaps invoking the
Sapir-Whorf Hypothesis [SWH] in support. A recent exponent of a modified
and updated version of this approach can be found in Deutscher (2011).
[It is
worth pointing out here that it isn't easy to see how anyone who accepts Chomsky's theory
can also accept
the SWH.]
[I will return to
this topic in a later re-write of this Essay.]
Until
then, the reader is directed to
Cook (1978a, 1978b), and
Hutchinson (2012). As Hutchinson points out:
"However, there is another, alternative, way to undermine Whorf's thesis without
appealing to or subscribing to Chomsky's theories; here Whorf's argument is
identified as not residing in the claim that different grammars exist in
different natural languages but rather in a set of unacknowledged philosophical
assumptions that are operative in Whorf's hypotheses. This criticism of Whorf
brings to the fore the philosophical assumptions underlying Whorf's claim that
grammatical differences entail different metaphysics. In other words, rather
than focus on his grammatical pluralism one might focus on Whorf's claim that
one can read-off a metaphysics from the grammar of a language. For one thing of
note and all-too-often overlooked in discussions of Whorf's theory is that his
relativism is not entailed by his observations about the grammar of different
languages, and nor did Whorf claim it to be so. There is, rather, a
philosophical argument operative in Whorf's thesis, which is assessable
independently of his grammatical pluralism. A criticism addressed to this aspect
of Whorf's thesis is one that seeks to show that Whorf's thesis lacks
intelligibility. John W. Cook's critique is an exemplar of this approach.
"On
close examination, Whorf's claim that different metaphysics can be read-off
different grammars can be shown to fall short of his aims and this one can see
on examination of his remarks about his own language: English. A little like the
anthropologist who writes of the primitive superstitions of the tribe he is
studying while it remaining literally unremarkable to him that he then prays
before dinner and kisses a photograph of a loved one before going to sleep for
the night, Whorf's writings about the metaphysics he claims to read-off
Hopi
grammar
are accompanied by a distinctly superficial depiction of the grammar
of his own language. His arguments about grammatical categories determining the
metaphysics of the speakers of the language are consequent on his being led
astray by the surface grammar of his own language, because where Whorf took
himself to be identifying a metaphysics which can be read-off the grammar of
English (and, when he turned to study that, Hopi) he was rather reading-into
English a metaphysics which owed much to his own pre-existing, underlying,
metaphysical assumptions about English grammar. To paraphrase John W. Cook,
Whorf was not reading-off a metaphysics from the grammar of English, but rather
reading into English a metaphysics not there. For example, Whorf employs the
example of the concept of 'time' in support of his claim that metaphysics is
read-off grammar, but his assumptions about the metaphysics of time are simply
read-off his observation that in English 'time' is a noun and he seems to assume
that nouns must correspond to something. Whorf is therefore being led astray by
the surface (superficial) grammar of 'time'. When he then takes himself to have
read-off a metaphysics of time as being constituted of 'moments', 'time-slices'
or 'time-flows' he is actually being led astray by his unacknowledged assumption
that 'time' as a noun in English must correspond to something.
"To
summarise: The real problem faced for someone who is persuaded by the
Sapir-Whorf hypotheses is not that it is based on a now unfashionable theory of
grammar, in light of the prominence attained by Chomsky's work. Rather, the real
problem is that what had appeared to many (including Whorf) like a thesis
founded upon, if not logically forced upon one by, the results of empirical
studies of the grammar of natural languages, is actually a set of claims
emerging from the observer (such as Whorf) being in the grip of an
unacknowledged picture of grammar that leads him to both misrepresent his own
language and then proceed to misrepresent those he was studying." [Quoted from
here. Accessed 25/01/2013. Link added.]
[In the above, Hutchison and Cook are relying on
Wittgenstein's criticism of what has come to be known as "The Augustinian
Conception of Language". On that, see here
and here. Indeed, the
anecdote quoted
earlier
was designed to combat this view of language.]
As we will see, Deutscher is no less culpable
in this regard.
118.
On the other hand, if this had in fact have happened, our group of
proto-linguists would only have needed to be given a basic grounding in the new
tongue (How? By whom?), which they were then about to extend. That is because
(given this particular variation on the current theme) they would succeed in grasping their new
'language' from an early age
-- otherwise the gene or set of
genes responsible (manifestly) wouldn't be functioning aright. But, even then, all the
problems outlined in the main body of this Essay (and
above) would still apply. [On this, see
Note 117a0 and
Note 120.]
118a.
Communicating and
then understanding what speakers say to one another involves far more than
simply making noises, or mouthing words at our interlocutors. Any supposition to the contrary risks
accepting the idea that we represent things to ourselves first and then
try to communicate those thoughts to others second -- and that our
ancestors did the same. [As we have seen, communication would be impossible if that
were the case.] This is, of course, a point that Voloshinov himself
would have acknowledged (even if we part company over the exact details). On this, see Note
120, andNote 124b.
119.
Naturally, this is consistent with the view that language is a social, not an
individualistic, phenomenon, and that we can only make sense of the past by means
of the language we now have -- even if we have to use an appropriate level
of sensitivity when
attempting to do just that.
120.
Again, the situation is much worse than this. Pentecostal 'innovators' are
already (by all
accounts) sophisticated language users. In which case, they already have
in place the requisite linguistic resources and social skills by means of
which they can at least make some attempt to comprehend what others in their
congregation who are jabbering away are
trying to say. If, at present, these Pentecostals still fail to grasp
what the happy band of babblers in their midst is trying to say, novice
proto-linguists stand no chance.
It could be argued that
while Pentecostal babblers might confront one another with a novel barrage
of sounds, these proto-linguists will surely have done this gradually over many
months or years, perhaps in a piecemeal manner. But, whether a Pentecostal
ecstatic utters one new sound or several doesn't affect the incomprehension this
is now met with by their hearers -- even those who are already sophisticated language users, let alone
children or those with learning difficulties. Odd sounds (whether uttered singly or as a job lot, or even if they are gradually introduced over the course of many years) would be of no use to anyone unless their
meaning was already understood, or could be explained by someone who understood.
This can't have been the case with these
proto-linguists, if, as even
Voloshinov acknowledges (in his more coherent moments), meaning is a social
phenomenon, not a 'process in the mind' of an individual. [I have
already covered this topic, earlier;
readers are directed there for more details. On this, see also Note 23.]
In that case, there would be no advantage to
a lone noise-maker in producing such sounds (that is, if we concede for the
moment the idea that the possession of language confers some sort of
selection or survival advantage).
Consider again the aphasic we
met earlier:
a monk called Brother John who, in the 1970s, had a series of epileptic
seizures. Whenever these episodes occurred, he couldn't understand anything
that was said to him, and no one could comprehend his words, either -- no matter
how many times they listened to him. Longer seizures, though, had a more dramatic effect:
"When his spoken speech was tested
during a spell, similar sounding nonsense words tended to recur, often variants
of the nonsense sequence tuwari. When shown a picture of a
telephone...[he] said: 'That's it, there. The furi twar. No. Glarity tuware tuwa
tuware ari tuware tuware tuwarere tu tuware tu'." [Aitchison (1996), p.39. Italic
emphasis in the original; quotation marks altered to conform with the conventions
adopted at this site.]
No matter how slowly or gradually this individual
introduced these odd sounds/'words', or howsoever piecemeal this turned out to be
-- or, indeed, how
often they were repeated -- they still made no sense to anyone; and his
listeners were far more sophisticated language users than
these hypothesised proto-linguists are supposed to have been.
Indeed, in
relation to the
above nonsense words, it is worth asking sceptical readers in relation to, "Glarity
tuware tuwa tuware ari tuware tuware tuwarere tu tuware tu",which are the verbs and which the nouns?
Incidentally, this also shows that
communication isn't about barking noises at one another -- not even in an advanced linguistic community.
[See also Note 126a, below.]
120a.
There is an earlier, and somewhat cruder version of the approach adopted at this
site to be found in
Fritz
Mauthner's work; on this, see Weiler (1970), pp.6-89. [However, the reader
shouldn't assume that I agree with everything
Mauthner had to say. Far from it!] It is worth adding that Mauthner was an
important influence on Wittgenstein. [Again, on this see Weiler, op cit.]
121.
This is what it
means to take Marx seriously when he said that social being 'determines
consciousness' -- or, better: we can determine the ideas and beliefs that individuals are
capable of forming,
have formed, or have adopted, when we examine at the Mode of Production in which they lived and
the class to which they belonged, etc. This topic will
be explored in more detail in Essay Twelve (summary
here).
Some might claim that they can in fact
'think' in pictures/images, or without language. That odd idea was addressed in Note 61. [See also Note 85.]
122. In this Essay, I won't be discussing
memetic theory
-- which has become a rather fashionable nostrum of late, and not just in
connection with certain
sections of EP, but right across the Internet. The
use of the word
"meme" is now almost ubiquitous, having spread like
Japanese Knotweed. It now seems to have lost its
theoretical connotations (rather like "electricity" and "magnetic" have), and seems to
mean something like "widely copied or used word or concept".
However, on
this topic readers should consult Churchland (2002),
Kuper (2000), McGrath (2005), Plotkin (2003), and Sperber
(1996, 2000). [I
hesitate to reference McGrath, since he is a leading Christian
Theologian, but
his arguments against Dawkins's memes are, in my view, conclusive.
Although, there is a
YouTube
video of the two debating some of these issues.]
123. Admittedly,
all else having failed, a cultural account of the origin of language
would be equally difficult to construct -- but it wouldn't, I think, be
impossible. Such a task won't however be attempted here since it would
involve the present author in concocting several "just so stories" of her own.
One notable attempt to do this can be
found in Canfield (1993), but it isn't one with which I would agree since it makes
far too many concessions to 'animal thought'.
[Cf., Greenspan and Shanker (2004).] However, an interesting and
superficially persuasive account of the origin of language can be found in
Bickerton (2009).
[I will say more
about the above
rather impressive book in a later
re-write of this Essay.]
124. Although this hasn't yet been attempted,
as far as I know, but modern epigenetic theories of evolution might one day be
extended into this area. However, even if we had some idea how such changes could be
imprinted on the genome, that wouldn't affect the points made in this part of
the Essay. Once again: linguistic innovators would, on this scenario, still face one another
as Pentecostal babblers do today -- i.e.,
with total
incomprehension.
It could be objected that it has been shown -- for example, in the work of
William Samarin --
that (i) there is no grammatical structure to glossolalial 'tongues' and (ii)
that the 'phonemes' used are homophonic
to the native language of each ecstatic. As Samarin points out:
"[G]lossolalia consists of
strings of meaningless syllables made up of sounds taken from those familiar to
the speaker and put together more or less haphazardly.... Glossolalia is
language-like because the speaker unconsciously wants it to be language-like.
Yet in spite of superficial similarities, glossolalia fundamentally is not
language." [Quoted from
here.]
If
so, it could be argued that Pentecostal babble can't be used to criticise the
formation of a proto-language since the
latter does have an hypothesised grammatical structure, while the former
does not. Nevertheless, even though this might be true, it doesn't affect the points
being made in this Essay. That is because (a) those who have analysed
Glossolalia are already sophisticated language users who can tell the
difference between a grammatical and a non-grammatical barrage of sounds -- this
wouldn't have been be the case with these novice proto-linguists --, and (b) whether or not a language has
a grammatical structure, strange sounds would still be incomprehensible -- unless, once more, there already existed a sophisticated linguistic community
that enabled bemused listeners to decipher them.
It
could also be argued that it isn't true that no one understands Glossolalia --
many claim to be able to translate them. However, as the Skeptic's Dictionary
points out:
"Glossolalics behave in
various ways, depending on the social expectations of their community. Some go
into convulsions or lose consciousness; others are less dramatic. Some seem to
go into a trance; some claim to have amnesia of their speaking in tongues. All
believe they are possessed by the Holy Spirit and the gibberish they utter is
meaningful. However, only one with faith and the gift of interpretation is
capable of figuring out the meaning of the meaningless utterances. Of course,
this belief gives the interpreter unchecked leeway in 'translating' the
meaningless utterances. Nicholas Spanos notes: 'Typically, the interpretation
supports the central tenets of the religious community' (Spanos, 147)." [Quoted
from
here.
Quotation marks altered to conform with the conventions adopted at this
site. The reference is to Spanos (1996), a source I have not yet been able to
check.]
Since
such 'translations' can't be checked, they are in no way different from
completeinvention. Of course, as the story in the Bible makes
plain, the original point of this 'miracle' was to spread the Gospel:
"And when the day of
Pentecost was fully come, they were all with one accord in one place. And
suddenly there came a sound from heaven as of a rushing mighty wind, and it
filled all the house where they were sitting. And there appeared unto them
cloven tongues like as of fire, and it sat upon each of them. And they were all
filled with the Holy Ghost, and began to speak with other tongues, as the Spirit
gave them utterance. And there were dwelling at Jerusalem Jews, devout men, out
of every nation under heaven. Now when this was noised abroad, the multitude
came together, and were confounded, because that every man heard them speak in
his own language. And they were all amazed and marvelled, saying one to
another, Behold, are not all these which speak Galileans? And how hear we
every man in our own tongue, wherein we were born?Parthians, and Medes,
and Elamites, and the dwellers in Mesopotamia, and in Judaea, and Cappadocia, in
Pontus, and Asia, Phrygia, and Pamphylia, in Egypt, and in the parts of Libya
about Cyrene, and strangers of Rome, Jews and proselytes, Cretes and Arabians,
we do hear them speak in our tongues the wonderful works of God." [Acts
2: 1-11.
Bold emphases added; spelling modified to conform with UK English.]
In
this case, those new tongues were already in existence, which is why they
were readily understood. That isn't the case with contemporary Glossolalia. [Readers mustn't assume I
believe this tosh!]
As
the Skeptic's Dictionary rather tartly points out:
"This story is supposed to
support the notion that such an event really did occur and it was prophesied by
Joel that this kind of thing would happen in the last days. There is nothing in
Joel, however, that prophesied that, when the last days didn't come as
predicted, plan B would be to wait 1900 years and have a revival and claim that
when you speak gibberish it is a sign that God loves you." [Quoted from
here. Link
added.]
124a. This shouldn't be taken to
imply that I think that language has no biological basis. Plainly, had our
evolution taken a different course, and had human beings not inherited the
physical make-up we now have, it is unlikely that language would have developed.
Plainly, such biological facts are necessary, but not sufficient, to account
for the development of our capacity to talk and hence communicate.
124b. Naturally, this over-simplifies
things somewhat. To say that someone has understood something amounts to far
more than merely noting they use language as we do and that they share the same
vocabulary. Understanding is assessed in different ways in different
circumstances and in different areas of life; but one thing we don't do when we attempt to ascertain if our interlocutors have either
comprehended
what we have said, or have understood a task that has been assigned them, is examine their brains.
Typically, we monitor their overt reaction, performance and response. [On this, see
Baker and Hacker (2005a), pp.357-85, and Baker and Hacker (2005b), pp.305-56.
Also see, Note 51, above.]
124c. It could be objected that we
not only can, we do
manage to innovate non-socially today. New words are constantly being invented by individuals;
indeed, unless someone thought up a new use, or began to employ, a word in new ways
there would be no innovation.
Maybe so, but we can't assume that the
proto-linguists under scrutiny here innovated as we do today where
invention of new words is now situated against a background of shared
vocabulary, shared linguistic practices, and cultures, for these proto-linguists didn't
yet possess either a language or settled linguistic practices. Moreover, these hypothetical ancestral innovators and
the groups to which they belonged certainly hadn't gone through the lengthy
social development we have, which was responsible for the
creation of the sophisticated language we now possess and which allows us to innovate individually. To
suppose otherwise is to make the same mistake made by bourgeois theorists -- that we
are not only social atoms, our class, social or historical development and background
are
irrelevant in this respect.
So,
and to repeat, we manifestly do not innovate today against a background where there is no
language whatsoever.
125. Alex
Callinicos, for example, once called this a "truism";
cf., Callinicos (1998b), p.177. Readers
are referred to my discussion of this topic,
here.
126.Some
might be tempted to point out that this argument resembles creationist
criticisms of Darwinism. This isn't so, since the evolution of language isn't
being denied. All that is being claimed is that the theories we currently
possess can't account for it, and since we are already language users
it isn't likely that we will ever be able to account for it. [That is so for reasons
outlined in the main body of this
Essay.] Hence, the argument here doesn't appeal to "irreducible
complexity", but to (i) the fact that language depends on social
convention, and any explanation of its origin has to assume the very thing that
is to be explained -- social convention --, and (ii) the additional fact that as sophisticated
language users already, our theories (as social products themselves) can't
help but be coloured by social convention.
This
isn't to argue, either, that we are trapped by
language, or that it is some kind of 'prison', only that howsoever we try we can't even begin to recreate in our
imagination the conditions that existed before we
invented language without also going behind social convention -- thus trying to find a
naturalistic explanation for something that is irreducibly social -- and, what
is more, without
using language in order to do this. So, we aren't so much trapped by language, as limited by our social nature
and history as human beings.
[The general background
details to this topic were covered in Essay Twelve
Part One.
More details will be given in Essay Thirteen Part Two (which focuses on the nature of science
and its relation to DM),
when it is published sometime in 2025.]
126a.
Interesting medical opinion of this episode can be found
here. As the doctor at the latter site had this to say:
"In addition, the idea that
one can 'decode' aphasia is simply ludicrous."
Another, writing in the comments section on the same page, offered this opinion:
"As a speech-language pathologist, I can tell you that the
language output of people with aphasia cannot be decoded. It's a disorder of
symbols…even their gestures are affected (undifferentiated -- they tend to use
the same gestures for anything they are trying to communicate). Further, they
cannot use a dictionary to look up the words they are trying to say. People with
aphasia WISH there were such a nice answer…there is not. In the case presented
on House, the pt [patient -- RL] clearly thought his expressive language
was appropriate…he did NOT recognize his errors. How could a person like that
use a dictionary to find the right word? He thinks what he said was right! You
cannot decode aphasia." [Capitals in the original.]
Which
makes my point rather well. It isn't possible to comprehend radically
flawed, or novel, 'speech', even where this occurs in
homophonic discourse, as was
the case above -- the man afflicted with aphasia in this episode of House
uttered nothing but English words in grammatical sentences, but he still failed to make sense.
Of course, had the cast of House been novice linguists -- living perhaps
hundreds of thousands of years ago -- not
sophisticated speakers of English who also had expert knowledge of aphasics (etc.),
they would plainly have failed even to guess his speaker's meaning.
Hence, our ancestral innovators would face each
other as a group of aphasics face us, or one another, now -- only their
plight would have been far worse, since they would have no shared vocabulary, no norms of usage to fall back upon,
and no expert knowledge to guide them, unlike those of us today who are confronted by the odd things
some individuals utter, especially aphasics.
And,
it isn't to the point to suggest that, contrary to the opinion advanced in the above
quotation, we might one day learn to interpret
aphasics with ease, for if we ever manage to do that, it would merely be testimony to our
enhanced interpretive skills, shared vocabulary, norms of application and
advanced medical/psychological expertise. Manifestly, these weren't available to
proto-linguists and novice 'language decoders' hundreds of thousands of years ago.
128. On
this, see Stove (2006), pp.258-306, and Fodor (1998c), partially quoted
here.
[However, readers unfamiliar with David Stove's work should also heed
this warning.]
129a. Of course, for those who
believe in 'natural necessity' this is no big deal. But, they are closet
Platonists anyway. On that, see Note 130 andNote 133. [On 'determinism', see
here and the
Appendix.]
Even
so, as Note 93and Note
130 also show, Harman is indeed just such a believer.
Further proof that Harman is a necessitarian-still-in-the-closet can to be
found in the following passage from Engels, which Harman quotes approvingly:
"'Men-in-the-making arrived
at the point where they had something to say to each other. Necessity
created the organ; the undeveloped larynx of the ape was slowly but surely
transformed by modulation to produce constantly more developed modulation, and
the organs of the mouth gradually learned to pronounce one articulate sound
after another.'
"Parallel with
this there was a necessary development of the brain:
'The reaction
of labour and speech on the development of the brain and its attendant senses,
of the increasing clarity of consciousness, power of abstraction and of
conclusion, gave both labour and speech an ever renewed impulse to further
development.'" [Harman (1994),
p.85; quoting Engels (1876),
p.356. Bold emphases added.]
This
seems pretty clear.
As we
have seen, in order to make traditional accounts of causation work,
theorists have to appeal to some form of necessity operating in nature. But,
that just
anthropomorphises the material world, making
it look as if it were under the influence of a Cosmic Will of some sort.
This topic will be discussed in more detail
presently, and in Essay Three Part Five.
Until then, see here and the
Appendix. The metaphysical background to this way of looking at the world was
examined in
Essays Eight Parts
One and
Two, Twelve Part One, and the
rest of Twelve (summary
here). Cf., also
Bobro (2017), and
Carlin (2008).
See also Note 150, and the links
in Note 133.
130a.
It isn't too clear what a non-literal interpretation of these metaphors might
look like, anyway! I
will attempt to say more about this in Essay Three Part Five.
131. As
noted above, without these metaphors, Harman's entire theory falls apart.
Previous theorists soon discovered that in order to remove 'mind' from nature,
and for nature still to function in the way they imagine it should, they
had to smuggle it back in again somewhere else, using an entire dictionary of
fetishised words to that end. They
found they were forced
to do this so they could explain how the universe worked 'lawfully'. [On this, see Essay
Twelve, once more -- summary
here and here.]
As noted in the main body of this Essay, often this was (and still is) achieved by the use of inappropriate metaphors and analogies, compounded by the
anthropomorphisation of nature and the fetishisation of language.
132. This
is particularly apparent in Harman (1994); e.g.,
p.188, n.73, among several other instances.
132a. Although,
Harman doesn't use the word "law", he does use "determine", and he refers
his readers to Engels's first 'Law' -- p.102. He also managed to piece together a rather good impression of a
'necessitarian'.
133.More
on this below: here, here,
and here.
[Once
again, this topic will be discussed in more detail in Essay Three Part Five.]
133a. Indeed,
languages change far faster than genes.
134. Except
perhaps because of the ideological blessing seemingly given to Darwinism by Marx and Engels. On this see
Gerratana (1973). [However, Gerratana accepts the myth that Marx wanted to
dedicate Volume Two of Das Kapital to Darwin (pp.79-80); on that, see
below.] A different view can be found in Ball (1979), who alleges that
it was Engels who incorrectly linked Marx and Darwin after Marx's death.
Others think it was all the fault of that complete waste of space,
Aveling. [Browne (2003), p.403.]
However,
Steven Jay Gould probably strikes the right balance in an article nominally about
Ray
Lankester, one of the few individuals who attended Marx's funeral:
"If Lankester showed so little
affinity for Marx's worldview, perhaps we should try the opposite route and ask
if Marx had any intellectual or philosophical reason to seek Lankester's
company. Again, after debunking some persistent mythology, we can find no
evident basis for their friendship.
"The mythology centres upon a notorious,
if understandable, scholarly error that once suggested far more affinity between
Marx and Darwin (or at least a one-way hero worshiping of Darwin by Marx) than
corrected evidence can validate. Marx did admire Darwin, and he did send an
autographed copy of Das Kapital to the great naturalist; Darwin, in the
only recorded contact between the two men, sent a short, polite, and basically
contentless letter of thanks. We do know that Darwin (who read German poorly and
professed little interest in political science) never spent much time with
Marx's
magnum opus. All but the first 105 pages in Darwin's copy of
Marx's 822-page book remain uncut (as does the table of contents), and Darwin,
contrary to his custom when reading books carefully, made no marginal
annotations. In fact, we have no evidence that Darwin ever read a word of Das
Kapital.
"The legend of greater contact
began with one of the few errors ever made by one of the finest scholars of
this, or any other, century --
Isaiah Berlin, in his 1939 biography of Marx. Based on a dubious inference
from Darwin's short letter of thanks to Marx, Berlin concluded that Marx had
offered to dedicate volume 2 of Kapital to Darwin and that Darwin had
politely refused.
"This tale of Marx's proffered
dedication then gained credence when a second letter, ostensibly from Darwin to
Marx but addressed only to 'Dear Sir,' turned up among Marx's papers in the
International Institute of Social History in Amsterdam. This letter, written
on October 13, 1880, does politely decline a suggested dedication: 'I Shd.
prefer the Part or Volume not be dedicated to me (though I thank you for the
intended honour) as it implies to a certain extent my approval of the general
publication, about which I know nothing.' This second find seemed to seal Isaiah
Berlin's case, and the story achieved general currency....
"To shorten a long story, two
scholars, working independently and simultaneously in the mid-1970s, discovered
the almost comical basis of the error (see Margaret A. Fay, 'Did Marx offer to
dedicate Capital to Darwin?' Journal of the History of Ideas39,
1978, and Lewis S. Feuer, 'Is the "Darwin-Marx correspondence" authentic?'
Annals of Science32, 1975). Marx's daughter
Eleanor
became the common-law wife of the British socialist Edward Aveling. The couple
safeguarded Marx's papers for several years, and the 1880 letter, evidently sent
by Darwin to Aveling himself, must have strayed into the Marxian collection.
"Aveling belonged to a group
of radical atheists. He sought Darwin's official approval, and status as
dedicatee, for a volume he had edited on Darwin's work and his (that is,
Aveling's, not necessarily Darwin's) view of its broader social meaning
(published in 1881 as The Student's Darwin, volume 2 in the
International Library of Science and Free-thought). Darwin, who understood
Aveling's opportunism and cared little for his antireligious militancy, refused
with his customary politeness but with no lack of firmness. Darwin ended his
letter to Aveling (and not to Marx, who did not treat religion as a primary
subject in Das Kapital) by writing:
'It appears to me (whether rightly or wrongly) that direct
arguments against christianity and theism produce hardly any effect on the
public; and freedom of thought is best promoted by the gradual illumination of
men's minds which follows from the advance of science. It has, therefore, been
always my object to avoid writing on religion, and I have confined myself to
science.'
"Nonetheless, despite this correction, Marx might still have regarded himself as
a disciple of Darwin and might have sought the company of a key Darwinian in the
younger generation -- a position rendered more plausible by Engels's famous
comparison (quoted earlier) in his
funerary oration. But this interpretation must also be rejected. Engels
maintained far more interest in the natural sciences than Marx ever did (as best
expressed in two books, Anti-Dühring and Dialectics of Nature).
Marx, as stated above, certainly admired Darwin as a liberator of knowledge from
social prejudice and as a useful ally, at least by analogy. In a famous letter
of 1869, Marx wrote to Engels about Darwin's Origin of Species: 'Although
it is developed in the crude English style, this is the book which contains the
basis in natural history for our view.'
"But Marx
also criticized the social biases in Darwin's formulation, again writing to
Engels, and with keen insight:
'It is remarkable how Darwin
recognizes among beasts and plants his English society with its division of
labour, competition, opening up of new markets, invention and the
Malthusian
'struggle for existence.' It is
Hobbes'sbellum omnium contra omnes [the war of all against all].' [Marx
to Engels, 18/06/1862.]
"Marx
remained a committed evolutionist, of course, but his interest in Darwin clearly
diminished through the years. An extensive scholarly literature treats this
subject, and I think that Margaret Fay speaks for a consensus when she writes
(in her previously cited article):
'Marx...though
he was initially excited by the publication of Darwin's Origin...developed a
much more critical stance toward Darwinism, and in his private correspondence of
the 1860s poked gentle fun at Darwin's ideological biases. Marx's
Ethnological Notebooks, compiled circa 1879-81, in which Darwin is cited
only once, provide no evidence that he reverted to his earlier enthusiasm.'" [Gould
(2002c), pp.123-25. Spelling adjusted to conform with UK English; formatting
and Quotation marks altered to conform with the conventions adopted at this site. I have
not been able to check the articles Gould cites. I have, however, added the references to
Marx's correspondence. Links also added.]
It is rather odd, though, that Gould ignored a letter Engels
sent to Lavrov, which is far more
negative in its opinion of
Darwin:
"1) Of the
Darwinian doctrine I accept the theory of evolution, but Darwin's
method of proof (struggle for life, natural selection) I consider only a first,
provisional, imperfect expression of a newly discovered fact. Until Darwin's
time the very people who now see everywhere only struggle for existence
(Vogt,
Büchner,
Moleschott, etc.) emphasized precisely cooperation in organic
nature, the fact that the vegetable kingdom supplies oxygen and nutriment to the
animal kingdom and conversely the animal kingdom supplies plants with carbonic
acid and manure, which was particularly stressed by
Liebig.
Both conceptions are justified within certain limits, but the one is as
one-sided and narrow-minded as the other. The interaction of bodies in nature --
inanimate as well as animate -- includes both harmony and collision, struggle
and cooperation. When therefore a self-styled natural scientist takes the
liberty of reducing the whole of historical development with all its wealth and
variety to the one-sided and meagre phrase 'struggle for existence,' a phrase
which even in the sphere of nature can be accepted only cum grano salis
[with a grain of salt -- RL], such a procedure really contains its own
condemnation.
"...I should
therefore attack -- and perhaps will when the time comes -- these bourgeois
Darwinists in about the following manner:
"The whole
Darwinists teaching of the struggle for existence is simply a transference from
society to living nature of Hobbes's doctrine of bellum omnium contra omnes
[from Hobbes's De Cive and Leviathan, chapter 13-14] and of
the bourgeois-economic doctrine of competition together with Malthus's theory of
population. When this conjurer's trick has been performed (and I questioned its
absolute permissibility, as I have indicated in point 1, particularly as far as
the Malthusian theory is concerned), the same theories are transferred back
again from organic nature into history and it is now claimed that their validity
as eternal laws of human society has been proved. The puerility of this
procedure is so obvious that not a word need be said about it. But if I wanted
to go into the matter more thoroughly I should do so by depicting them in the
first place as bad economists and only in the second place as bad
naturalists and philosophers.
"4) The
essential difference between human and animal society consists in the fact that
animals at most collect while men produce. This sole but
cardinal difference alone makes it impossible simply to transfer laws of animal
societies to human societies....
"At a certain
stage the production of man attains such a high-level that not only necessaries
but also luxuries, at first, true enough, only for a minority, are produced. The
struggle for existence -- if we permit this category for the moment to be valid
-- is thus transformed into a struggle for pleasures, no longer for mere means
of subsistence but for means of development, socially produced
means of development, and to this stage the categories derived from the animal
kingdom are no longer applicable. But if, as has now happened, production in its
capitalist form produces a far greater quantity of means of subsistence and
development than capitalist society can consume because it keeps the great mass
of real producers artificially away from these means of subsistence and
development; if this society is forced by its own law of life constantly to
increase this output which is already too big for it and therefore periodically,
every 10 years, reaches the point where it destroys not only a mass of products
but even productive forces -- what sense is their left in all this talk of
'struggle for existence'? The struggle for existence can then consist only in
this: that the producing class takes over the management of production and
distribution from the class that was hitherto entrusted with it but has now
become incompetent to handle it, and there you have the socialist revolution.
"...Even the
mere contemplation of previous history as a series of class struggles suffices
to make clear the utter shallowness of the conception of this history as a
feeble variety of the 'struggle for existence.' I would therefore never do this
favour to these false naturalists....
"6) On the
other hand I cannot agree with you that the 'bellum omnium contra omnes'
was the first phase of human development. In my opinion, the social instinct was
one of the most essential levers of the evolution of man from the ape. The first
man must have lived in bands and as far as we can peer into the past we find
that this was the case...." [Engels to
Lavrov,
17/11/1875.
Spelling adjusted to UK English; formatting and Quotation marks altered to
conform with the conventions adopted at this site.
Links also added.]
Which seems to me to get things about right.
However, it is worth adding a comment about this famous comment by Marx:
"Although
it is developed in the crude English style, this is the book which contains the
basis in natural history for our view." [Marx
to Engels, 19/12/1860, in Marx and Engels (1985), p.232. The published
version I have referenced is obviously a different translation to the one Gould
used.]
Even before I became a revolutionary, this throw-away remark
puzzled me since it didn't appear to be consistent with
HM, and might seem to many to be
advocating some form of
Social Darwinism. However, the struggle envisaged by Darwin had nothing to
do with the class war, which, in the end, leads to its own demise. Nor does the
class war have anything to do with reproductive advantage, or survival value.
There is nothing like this in the natural world. So, it might be more to the
point to interpret this to mean that Marx was referring to their joint atheism
and their naturalistic view of extra-social reality. That interpretation
is clearly supported by Marx's use of the phrase "natural history".
134b.Recall, I am not questioning Harman's argument here (i.e., the one that appears on
pp.96-99), merely
pointing out that we need to give this far more careful
consideration than Harman might have had available space, or time.
135. The
locus classicus of this line of thought is, of course, Davidson (1963).
On this in general, see Hacker (2007a), pp.199-232; on Davidson, see pp.226-32. Since this approach to human action (among
others things) has gained some credibility among certain sections of the 'Marxist
intelligentsia', I have devoted this section of the Essay to it.
A recently unsuccessful attempt to make Roy
Bhaskar's ideas comprehensible can be found in Archer, et al (1998). Bhaskar's own thought,
however, seems to have
slipped back into the mystical swamp from which it originally slithered -- see,
for instance, Bhaskar (2003a, 2003b, 2003c).
136. Cf.,
Collier (1994), pp.209-21, and Pateman (1987).
[I will examine
Pateman's more important arguments here at a later date.]
Readers might take exception to this
rejoinder,
claiming that the present author often appeals to the typical (or "everyday")
use of certain words without producing "the yet-to-be-published, innovative research into
the sociolinguistics of the terms in question". If an omission like this is OK
for Ms Lichtenstein, why is it a problem for Collier?
However, the hypothetical objector omitted this
significant rider:
This
query isn't being raised out of mere cussedness or even 'pedantry'; if someone
makes an empirical claim about the "typical" way the rest of us are supposed to
interpret (or use) a certain word that is in fact not even remotely like
the way they would naturally employ it, some supporting evidence is the least
one should expect.
When an appeal is made to "everyday use" in
these Essays it isn't the case that that use is "not even remotely like
the way they would naturally employ" certain words. Quite the reverse, in fact.
Throughout this site I quote scores (if not hundreds) of easily recognisable
everyday sentences. Collier does not. Who uses, or even implies, the words "tipped the balance"
when speaking about what caused what in everyday life? Indeed, as we are about
to see, we don't even "a-typically" regard causes this way when we reflect about
them.
138.Online dictionaries also fail to mention
'balance-tipping' in connection with causes:
"cause
"n.
"1.
"a. The
producer of an effect, result, or consequence.
"b. The
one, such as a person, event, or condition, that is responsible for
an action or result.
"2. A basis
for an action or response; a reason: The doctor's report gave no cause
for alarm.
"3. A goal or
principle served with dedication and zeal: 'the cause of freedom versus
tyranny' (Hannah Arendt).
"4. The
interests of a person or group engaged in a struggle: 'The cause of
America is in great measure the cause of all mankind' (Thomas Paine).
"5. Law
"a. A
ground for legal action.
"b. A
lawsuit.
"6. A subject
under debate or discussion.
"tr.v. caused,
causing, causes
"1. To be the
cause of or reason for; result in.
"2. To bring
about or compel by authority or force: The moderator invoked a rule
causing the debate to be ended." [Quoted from
here.]
"noun
"1. a person, thing, event, state, or action that produces an effect
"2. grounds for action; motive; justification ⇒ she had good cause to shout like
that
"3. the ideals, etc, of a group or movement ⇒ the Communist cause
"4. the welfare or interests of a person or group in a dispute ⇒ they fought for
the miners' cause
"5. a matter of widespread concern or importance ⇒ the cause of public health
"6.
1. a ground for legal action; matter giving rise to a lawsuit
2. the lawsuit itself
"7. (in the philosophy of Aristotle) any of four requirements for a thing's
coming to be, namely material (material cause), its nature (formal cause), an
agent (efficient cause), and a purpose (final cause)....
"verb
"9. tr to be the cause of; bring about; precipitate; be the reason for...." [Quoted
from
here.]
138a.
It could be
argued that this is an absurd objection since Collier and Bhaskar are talking
about causes that are specific to the events in hand, not a cause of everything.
That
response will be dealt with presently.
138b.
In the example given in the main body of this Essay (concerning the slow acceleration of
the bus because of the driver's lack of attention), it could be argued that one
of the reasons why NN caught the bus was the driver's inattention. Indeed, but
it can't be NN's reason. I return to this topic in the
next sub-section of this Essay.
139.
Even so, when we are told that a certain word typically means another
term,
the two are surely interchangeable. In that case, for Collier and Bhaskar,
reasons are causes just as much as causes are reasons:
"What does it mean to say
that reasons can be causes? Bhaskar suggests that:
'When something is cited as
a cause it is, I think, most typically being viewed as that factor which, in the
circumstances that actually prevailed, "so tipped the balance of events to
produce the known outcome"…'….
"Intentional actions involve
beliefs and desires…. It is hardly open to dispute that, given a desire for
something, coming to have a belief about the way to get it may 'tip the
balance', and so be naturally described as 'the cause'." [Collier (1994),
p.152-53. Bold emphases added.]
It could be objected that the above quotation
doesn't say that the word "reason" means the same as "cause",
it merely says that reasons can be causes, which is fair enough. But, that
seemingly innocent first move only leads the reader toward the words highlighted
above
(in bold). Now,
that part of the quotation is altogether less neutral. Beliefs and desires
are there connected with reasons which can now be "naturally described" (not,
note,
sometimesdescribed, but "naturally described") as "the cause".
This clearly implies that a reason (or a set of reasons) is the cause of
our actions, just as the cause of an (intentional?) action is a reason. It is
difficult, therefore, to resist the conclusion that for Collier, the words
"reason" and "cause" (used in such a context) are synonymous.
139a.
On this, see Child (1996), pp.91-92. However, it should be pointed out that Child
in the end argues for a view that is diametrically opposed to the approach
adopted
in this Essay.
"A main cause of
philosophical disease -- a one-sided diet: one nourishes one's thinking with
only one kind of example." [Wittgenstein (1958), §593, p.155e.]
141.
Anyone tempted to answer "Yes" to all of these questions on the grounds that something
must have 'tipped the balance' at some point would plainly be using Bhaskar's thesis as a
"form of representation" -- i.e., as a logical device that sanctions
specific inferences and conclusions before the relevant evidence becomes
available (that is, if it ever does!), allowing the said objector to assume there
must always be such a 'balance-tipping' cause.
[The use of the word "must" here in response to this objection, by the way,
is a further give-away. Where there is a "must" there is an inference.]
"Form of representation" will be explained more fully in
a later Essay. Until then, the reader is directed to Glock (1996), pp.129-35.
See also here.
It
could be argued that this is merely an
inference to the best explanation [IBE]. IBEs (or "abductive
inferences") will be discussed in a later
Essay, too, but it is sufficient to point out here that the best explanation we can
give for an intentional action involves a consideration of the reasons why the agent
concerned performed it.
Assuming this is always a cause would be to
beg the question, therefore.
142.
Note, this might help us understand how reasons could be
'balance-tipping' with respect to the examples listed in
C1-C6, too.
142a0.
Some might dispute whether an agent always knows why he/she does
something. That worry will be allayed presently.
142a. It could be argued that these
examples are highly contrived, and so are of little relevance. But, that isn't so with
respect to (ii) and (iii). However, for extra, less 'controversial', examples, see Schaffer
(2004).
142b. And good luck answering that
question! On this, see McPherson (1990), especially pp.854-59. Readers can also
listen to Professor McPerson's lecture on this topic,
here. [See also Goodwin (2013) and Keegan (2010).]
After
sorting out the 'balance-tipping' cause of the defeat of the Confederacy,
ambitious readers can then like to try their hand at finding the
'balance-tipping' cause that led to the Nazi Wehremacht losing the
Battle of
Kursk in July and August 1943, and then the 'balance-tipping' cause of the
decline and fall of the Roman Empire. [For assistance on the latter, see Heather (2006).]
143.
I have outlined the case against such 'laws' in Essay Thirteen Part Two (to be
published in 2025), and Essay Three Part Five (to be published at a later date).
Nevertheless, this way of viewing reasons and causes isn't just a
'Wittgensteinian prejudice', it can be found in Hegel, too -- cf.,
MacIntyre (1976b). The so-called "Critical Realists" -- who apparently
hold Hegel in high regard -- seem not to have noticed this rather damning fact.
Of course, this introduces the vexed question
whether there can be singular causes. On that, see Note 144.
144.
For recent work on singular causes, etc., see
Psillos (2002), Chapter Two. For how these are to be distinguished from the
coincidental, or the merely adventitious,
see Anscombe (1981b). See also Cartwright (2000), Teichmann (2008), pp.177-90,
and Hacker (2007a), pp.57-121 for an ordinary language account of causation that
doesn't lapse into necessitarianism. A brave
attempt to distinguish causes from coincidences can be found in Owens (1992).
144a.
It could be argued that if we hope to give a naturalistic account of human
action, motives and intentions, then we are forced to appeal to natural laws
governing certain processes in the
CNS of the individual
concerned. This will then by-pass questions about the uniqueness of each human
being, since it will focus on natural processes that aren't exclusive to one
individual. But,
even if we accept this view of human action (and it should be clear by now from
what has gone before in this Essay that I don't), this won't work either. That
is because if the human action concerned is unique (and it must be so for the
reasons explored in the main body of this Essay), then whatever natural or physical
processes led up it being performed will be unique, too.
145.
Again, it is worth pointing out here that the philosophical thesis that causes
(or causal laws) are 'particulars/tropes/individuals'
is neither being asserted nor denied here (for reasons outlined several times at
this site -- especially in Essay Twelve
Part One). What is of
concern, however, is the fact that a metaphysical view of causation
oscillates alarmingly between two extremes: (i) the belief that
laws are irreducibly general and (ii) the idea that they are essentially particular.
[These days the latter view seems to have morphed into an updated version of
Essentialism. More on this in Essay Three Part Five.]
Nor is this Essay aimed at questioning the laws that scientists
have constructed. It is
aimed at undermining the amateur Metaphysics they often indulge in when they
write prose about these laws.
[The word "prose" here is a Wittgensteinian term,
and characterises the sorts of things mathematicians and scientists say about their
theories when they attempt to translate them into ordinary, or even semi-technical,
language. Invariably, they find they have to indulge in amateur Metaphysics or
use inappropriate metaphors and analogies.]
145a.
Dogmatism in Philosophy was tackled in
Essay Two and Essay Twelve
Part One. [On this, see Kuusela
(2008).]
145b.
I am well aware that this view of Hume has come under sustained attack of late.
[On this, see, for example, Strawson (1996). Also, cf., Psillos (2002),
pp.22-23.] Nevertheless, this recent development doesn't affect the argument in the main body of this Essay
since Collier appears to accept the traditional view of Hume.
145c.
This explanation of causation seems to owe nothing to
Hegel's
theory (or to Lenin's), which is odd in view of the fact that
'Critical Realists' are supposed to be Hegelians of one sort or another (upside
down or 'the right way up').
146.
This is, of course, a predicament that has afflicted Traditional Philosophy from
the get-go.
[Several of Bhaskar's key arguments
will be examined here at a later date. In the meantime, the reader is directed
to my earlier comments,
here
and
here.
Many of his arguments, and those of the 'Critical Realists' depend on the
response to Hume advanced in Harré and Madden (1975). I will also examine
this book at a later date. In the meantime, see
Note 150.]
147.
On this, see Essay Eight
Part One.
The ideas of a handful of modern 'necessitarians' will be examined in Essay Three Part
Five. [A general attack on this way of viewing any philosophical theory was
developed in Essay Twelve
Part One.]
148a.
Indeed, this is the line of
reasoning that led
Leibniz to re-write causation along just such 'rational' lines, as the action
of 'tiny minds', or
monads.
[On this see,
Bobro (2017), and
Carlin (2008).]
In fact, and amusingly, this picture of causation often re-surfaces in
children's films when an attempt is made by someone to perform a feat of magic. The 'magician'
or 'wizard'
concerned will
just wave
his or her hands about, or make facial gestures -- in the 1960s TV programme,
Bewitched,
'Samantha', the 'witch', only had to twitch her nose --, utter a few 'magic
words', and objects
effortlessly moved
in the 'required manner' as if they not only understood exactly what was
intended by such
histrionics they knew precisely how to go about obeying these coded 'commands'.
Clearly, this fanciful picture is predicated on the idea that inanimate objects
are intelligent and quintessentially dutiful, if addressed in the right manner. The
Harry Potter film franchise is a classic example of this genre.
But, this is precisely the fanciful picture of nature that
dominates Traditional Theories of causation -- albeit hidden behind page
after page of verbose jargon!
149. We have
already noted that Bhaskar has reverted
to an overt form of mysticism; we can now add Collier's name to this list, too!
[Cf., Collier (1999, 2001, 2003).] But,
there are many others, as Lenin predicted there would be (in
MEC). Why
DM-fans are quite so susceptible in this regard is revealed in Essay Nine
Part Two.
150. For a neo-Humean account of causation,
see Mackie (1980). For a recent survey, see Psillos (2002). For an
account of physical law that doesn't rely on 'natural necessity', see Swartz (1985,
2009), and for a conventionalist
account of necessity, see Sidelle (1989).
Also
see the little-known and unjustly neglected article written by Bertrand Russell
[i.e., Russell (1917c)], which, apart from the metaphysical views it expresses,
presents an account of causation not significantly different from the one adopted at this site. [On
that, see also Price and Corry (2007).] For a more Wittgensteinian account of
our use of "cause" (and its related vocabulary), see the
references in Note 144, above.
Once more, Bhaskar's (and thus Collier's) theory
leans heavily on Madden and Harré (1975). I will consider their
arguments in Essay Three Part Five. In the meantime, the reader is
directed to Hacker (2007a), pp.57-121 and Sidelle (1989).
151.
There is a rather effective demolition of Bhaskar's theories in Pleasants (1999).
151a.
It is worth warning the reader here that Sampson is a right-wing Tory who holds
more than his fair share
offensive reactionary and racist opinions, which, naturally, I have neither used nor
referenced in my discussion of Chomsky. Despite this, Sampson's technical
criticisms of the latter's work seem to me to be valid, and, as I noted with
respect to the opinions of David Stove (here),
his right-wing views should no more prevent us from
reading his (non-reactionary) criticisms of Chomsky than dialecticians allow
Hegel's right-wing ideas to prevent them from studying his 'Logic'.
However, Sampson's
alternative hypothesis -- that children learn their first language by
trial-and-error and by hypothesis formation -- is hardly any better than
Chomsky's. For a child to form an hypothesis it would already have to have
language (not to mention possessing a rudimentary knowledge of 'the
scientific method'!); in which case, this theory can't account for first
language acquisition. Moreover, it is quite clear that Sampson's theory is consistent with his
right-wing ideology, since he sees each of us as socially isolated and atomised
individuals, teaching ourselves to speak, as opposed to our having been
trained or socialised in its use by relatives, careers, peers and (in many cases)
teachers. It isn't surprising, therefore, to discover that Sampson was heavily
influenced by
Karl Popper and the Empiricist tradition, which further confirms
my allegation that in this area, theories of human cognition haven't advanced
much beyond those of the 17th
century -- and thus smack of
bourgeois
individualism.
152.
Even though, in the end, Cowie makes rather too many concessions to Nativism.
Cf.,
Sampson (2005, 2008), and Cowie (1997, 2002,
2008),
along with the discussion
here (accessed 16/08/2004).
For Marxist criticisms of Chomsky, see
Knight (2003,
2005,
2006,
2007a,
2007b,
2007c,
2010a,
2010b,
2011, 2016a,
2016b), and Lecercle (2006). On the general background to this often
heated debate, see Harris (1995). Since this Essay was originally written, Bickerton has changed
his mind somewhat; on that, see Bickerton (2009).
In this section, I intend to reproduce reviews
of some of the books referenced in this Essay alongside other relevant material.
In what follows, bold emphases and links have
been added. Italic
emphases are in the original. Minor typos have been corrected; spelling altered
to conform with UK English. Readers mustn't assume I agree with every view
expressed in this material.
What do colon cancer, ant colonies, language and
global warming have in common? This might sound like the front end of a joke,
but in fact it's a serious challenge to the standard view of evolution. Martin
A. Nowak, the director of the Program for Evolutionary Dynamics at Harvard, has
devoted a brilliant career to showing that Darwin, and particularly his
followers,
batted only two for three. Random mutation and natural selection have indeed
been powerful motors for change in the natural world -- the struggle for
existence pitting the fit against the fitter in a hullabaloo of rivalry. But
most of the great innovations of life on earth,Nowak argues, from genes to
cells to societies, have been due to a third motor, and "master architect," of
evolution: cooperation.
"SuperCooperators" (written with Roger Highfield,
editor of New Scientist magazine) is an absorbing, accessible book about the
power of mathematics. Unlike Darwin with his brine bottles and pigeon coops,
Nowak aims to tackle the mysteries of nature with paper, pencil and computer. By
looking at phenomena as diverse as
HIV infection and
English irregular verbs, he has formally defined five distinct mechanisms that
have helped give rise to cooperative behaviour, from the first molecules that
joined to self-replicate, to the first cells that formed multicellular
organisms, all the way to human societies, which exhibit a degree of cooperation
unmatched in all creation. In Nowak's view, figuring out how cooperation comes
about and breaks down, as well as actively pursuing the "snuggle for existence,"
is the key to our survival as a species.
At the heart of
Nowak's ideas is the haunting game of
Prisoner's
Dilemma. The game involves two accomplices who
are caught for a crime, interrogated separately
and offered a deal. If one player incriminates
the other, or "defects," while the second
remains silent, or "cooperates," he will be
given a sentence of one year, while the other
player gets four. If both remain silent, they
will be sentenced to only two years, but if both
defect, they will receive three years. The
rational choice for either prisoner is to
defect, getting three years -- though had both
cooperated, they'd have been out in two. In the
absence of trust, reason can be
self-destructive.
In the 1990s,
Nowak and Karl Sigmund, building on work by
Robert Axelrod, showed that the Prisoner's
Dilemma, played over and over, could describe
cycles of behaviour in which strategies of
selfishness ("Always Defect") are beaten out by
cooperation ("Tit for Tat"), then overtaken by
even more cooperative behaviour ("Generous Tit
for Tat," summarized as "Never forget a good
turn, but occasionally forgive a bad one"), only
to be invaded once more by egoists until the
cycle begins anew. These "evolutionary dynamic"
models, made more realistic by introducing an
element of randomness, demonstrate that under
the right conditions, competition can lead to
teamwork. They also show how fragile that
balance can be.
In
"SuperCooperators," Nowak argues that two of his
mechanisms, indirect reciprocity and group
selection, played an important role in human
evolution. Think of a proto-simian trying to
figure out whether to trust another in an
exchange: Should I provide sex now for food and
protection later? The proto-simian may have
observed the behaviour of its prospective
partner, or it may not have; chances are good
that others have, though. Reputation becomes
important. The proto-simian evolves into a
hominid, with a bigger brain allowing for more
precise communication about reputation. Moral
instincts evolve to produce shame, guilt, trust,
empathy; social intelligence and conscience are
born. Before you know it,
Yogi Berra is summing it all up: "Always go
to other people's funerals, otherwise they won't
come to yours." Language, cognition and
morality, Nowak argues, are evolutionary
spinoffs of the fundamental need of social
creatures to cooperate.
Cooperation also
breeds division of labour, as any ant or gene
will tell you. When group size and structure,
benefits and costs, all align just right,
Nowak's models show, the red talons of nature
turn green. Sixty million years of fungus
gardening by leaf cutter ants is one example,
genes stacked on
chromosomes another.
Nowak is one of
the most exciting modellers working in the field
of mathematical biology today. But a model, of
course, is only as good as its assumptions, and
biology is much messier than physics or
chemistry. Nowak tells a joke about a man who
approaches a shepherd and asks, "If I tell you
how many sheep you have, can I have one?" The
shepherd agrees and is astonished when the
stranger answers, "Eighty-three." As he turns to
leave, the shepherd retorts: "If I guess your
profession, can I have the animal back?" The
stranger agrees. "You must be a mathematical
biologist." How did he know? "Because you picked
up my dog."
Nowak does his
best to avoid dogs, but "SuperCooperators" gives
little sense of the debates that have raged for
years between two traditions of modelling
evolution. One school considers the intricate
complications of genetics, like the ways
different versions of a gene interact in sexual
reproduction. The other treats organisms as if
they have a single set of chromosomes and
reproduce asexually. At the heart of the debate
lies the crucial question of whether natural
selection is always maximizing the spread of an
organism's genes. Nowak belongs to the camp that
assumes it is.
Nowak has also
ignited controversy with a paper in the journal
Nature, written with
E. O. Wilson and Corina Tarnita, arguing
that "inclusive fitness" -- the idea that
organisms cooperate with relatives because it
helps pass on shared genes -- is not necessary
to explain the birth of complex societies like
bees and ants, or altruism towards kin in
humans. Nature recently published five critical
letters, including one with 137 signatories, one
of whom denounced the paper's mathematics as not
worth "wasting time" over.
Nowak gives
little hint of these fierce debates in this
cheerful book, instead offering this striking
claim: "The way that we human beings collaborate
is as clearly described by mathematics as the
descent of the apple that once fell in Newton's
garden." It seems significant to Nowak that,
according to his models, the interest of groups
can override the interests of individuals if
"the ratio of the benefits to cost is greater
than one plus the ratio of group size to number
of groups," and that cooperation can prevail if
altruists cluster together in particular
topographies. If only we could take such facts
into account, as special cancer-preventing
"crypt" formations in our colons have
unthinkingly done, perhaps we might work
together to combat global warming.
Near the end of
the book, Nowak describes
Gustav Mahler's efforts, in his
grandiloquent Third Symphony, to create an
all-encompassing structure in which "nature in
its totality may ring and resound," adding, "In
my own way, I would like to think I have helped
to give nature her voice too." But there remains
a telling gap between the precision of the
models and the generality of the advice Nowak
offers for turning us all into supercooperators.
We humans really are infinitely more complex
than falling apples, metastasizing colons, even
ant colonies. Idealized accounts of the world
often need to ignore the messiness of reality.
Mahler understood this. In 1896 he invited
Bruno Walter to
Lake Attersee to glimpse the score of the
Third. As they walked beneath the mountains,
Walter admonished Mahler to look at the vista,
to which he replied, "No use staring up there --
I've already composed it all away into my
symphony!" [Quoted from
here.
Several links added.]
Here follows
a summary of my ideas on 'determinism', but comrades shouldn't expect a
water-tight solution to this knotty problem in a few paragraphs. I am only
posting this because I was asked to do so.
To tell you the truth, I have so far only written on this topic tangentially. I
will however be publishing an essay specifically about this subject in the next few
years, where I will substantiate what I have to say below far more expansively.
This issue has always revolved around the use of jargon drawn from Traditional Philosophy (appropriating words such as "determined",
"controlled", "will", "free", and the like), the
employment of
which bears no relation to how these words are used in ordinary speech.
For example, "determine" and its cognates are typically employed in sentences like
these: "The rules determine what you can do in chess", "The time of the next train
can be determined from the timetable", or "I am determined to go on the
demonstration", and so on. Hence, this word is normally used in relation to
what human beings can do, apply, infer, or bring about.
As we will see, their use in Traditional Thought inverts this, making nature
the agent and human beings the patient.
[Added on edit: "patient" here is being used in its older sense -- i.e., to
refer to that which is acted on, not that which acts.] No wonder then that the 'solution'
to this spurious problem (i.e., 'determinism' versus 'free will') has eluded us
for over 2000 years.
An analogy might help illustrate how inappropriate such questions
are: would we take seriously anyone who wondered when the King and Queen in
chess got married, and who then wanted to know which priest conducted the ceremony? Or,
whether planning/zoning permission had been sought for that Castle over in the corner?
Such empty questions, of course, have no answer since they depend on a decidedly
odd use and understanding of the language associated with this particular game. The same applies to the traditional questions
asked about the use of "determine" and other words connected with human choice.
This is perhaps more difficult to see in relation to the
'problem' at hand, in relation to 'determinism', but it is nonetheless the result of similar confusions. So,
it is my contention that this 'problem' has only arisen because
ideologically-motivated theorists (who lived many centuries ago) asked such
empty questions, based on a similar misconstrual of language. [More on this
below.]
When the details are worked out, what we now call 'determinism' can only be made to seem to work
-- or, indeed, prompt the asking of what appear to be comprehensible
questions -- if nature is anthropomorphised, whereby natural laws,
for instance, are said to 'determine' the course of events in 'reality' as well
as in the central nervous system, thus 'controlling' what we do
or can think.
But, this is to take concepts that properly apply to what we do, think or
decide, and read them into natural events, suggesting that nature itself is the
expression of, or is controlled by, a Mind or a Cosmic Will of some sort. [Why
that is so will
be explained presently.]
In which case, concerning these 'laws' it is natural to ask: Where have they been written?
And who wrote them? More to the point, why should unintelligent matter 'obey'
them? Can atoms and electrons really comprehend what they are 'supposed to do' when
a
law 'orders' them to do it?
Of course, the answer to the first two questions above might be "No one" and "Nowhere", but then
how can something that doesn't exist control anything?
Theists will, naturally, answer such questions differently, but they have yet to
explain how 'god's will' can move electrons and atoms about the place, and how
such inanimate objects never disobey and behave everywhere in
the same way that they have always done. Are they intelligent? Theists
have an all-purpose and convenient answer to such questions, "It's a mystery".
Isn't it also a mystery that inanimate matter isn't intelligent but can 'obey'
laws?
It could be replied that natural law is just a summary of how things have
proceeded up to now. In that case, such 'laws' are descriptive not
prescriptive. But it is the latter of those two alternatives that determinists
need; their 'laws' need to prescribe what happens in nature and the
CNS. Of course, the descriptive aspect of such 'laws' sidesteps the following
even more awkward question: How does this view of physical law -- as
regularities -- distinguish them from accidental
generalisations? The answer to that question would, of course, require an
appeal to some form of 'natural necessity'.
[Added on edit: an 'accidental generalisation' would be something like this: the
coins in my pocket all contain copper; whenever I have coins in my pocket, I
have copper in my pocket, too. But does anyone think there is a causal
connection between 'being in my pocket' and 'made of copper'? If so, a gold coin
slipped into my pocket without my knowledge must now be made of copper! Exactly
why the 'prescriptive' nature of such laws (employing terms like "must" or
"necessarily") is required by determinists will also be explained.]
However, the introduction of modal notions here (such as 'must' and 'necessary')
can't be justified on the basis the descriptive nature of these 'laws' -- i.e., not without
re-introducing the untoward anthropomorphic connotations mentioned earlier.
So, if we say, concerning two events, A and B, that A is always followed
by B, we can't now
say Bmust follow A unless we attribute to A some
form of control over B (and recall B has not yet happened, so what
A is supposed to be controlling is somewhat obscure). And, if we now try
to say what we mean by 'control' (perhaps along lines that things 'couldn't be otherwise', or,
maybe that
'AmadeB happen' -- or even that 'A necessitates B') we
would need to explain how A is capable of
doing this. How could A prevent, say, C happening instead? How is
A able to make sure B, andonly B,
happens? And, in some cases, does so everywhere and at all times across the entire universe.
The use of "obey" here would further give the game away (i.e., if employed in sentences like "A
and B obey natural law"), since, if this word is used with connotations
that go beyond mere description, it would imply that events like B
'understand' the 'law' (like any good citizen in civil society), and always 'behave
in ways they have been ordered to act' whenever A 'determines' what they must do, everywhere, right across the entire
universe, for all of time. It further implies that such 'laws' must exist in some
form or other in order to make objects and processes 'obey' them.
Of course, if it
doesn't mean this, what does it mean?
Now, I maintain that any attempt to supply these extra (missing) details will again introduce
notions of will and intelligence into the operation of A on B (and
also negatively on C) -- which is why theorists have found they have
had to employ anthropomorphic concepts (such as 'determine', 'obey', 'law'
and 'control') to bridge this gap, failing to notice that the use of
such language does indeed imply there is a 'Mind' or a 'Will' of some sort operating in nature.
[But, note the qualifications I have introduced below. There were in fact ideological reasons
why words like these were chosen. Naturally, the introduction of a
'Cosmic Will' here solves nothing since it still leaves mysterious how inanimate
matter manages to 'do what it is old', unerringly, right across the universe,
for all of time. Again, in effect, this credits matter with some form of
intelligence.]
If this is denied then, once more, "determine" (etc.) can only be working descriptively,
not prescriptively, and we are back to square one. In that case, questions will
naturally arise: why choose this particular word if it doesn't mean that AdeterminesB? Why not just say B always seems to follow
A?
Incidentally, in relation to human action, the above problems can't be side-stepped by the introduction of
biochemical, neurological or physiological objects and processes into the
equation. The same
questions apply there as elsewhere: how can, for example, a certain chemical
'control' what happens next unless it is in some way intelligent and purposeful?
Reducing this to physics is even worse; how can 'the field', a set of
differential equations or whatever
control events still in the future? 'The field' is a mathematical object and no more
capable of controlling anything than a
Hermite Polynomial is. Of course, and once more, to argue otherwise would be
to anthropomorphise molecules and mathematical objects --, which is why I made the argument above
abstract, since it covered all bases.
This also explains why theorists (and particularly scientists who attempt to
popularise their work) find they have to put 'scare' quotes around certain
words, and use misleading
analogies and inappropriate metaphors in order to try to
explain themselves.
As I noted earlier, this whole way of looking at the 'will' inverts things.
We are denied a will (except formally) and nature is granted one. As some
readers might now be able to see: this is yet another aspect of the alienating nature of
Traditional Thought, where certain words are fetishised (i.e., they are
attributed with mysterious powers) and we are dehumanised.
This shouldn't surprise us since such questions were originally posed
theologically (and thus ideologically). Traditional Theorists were quite happy to
alienate to 'God' control both over nature and, in some cases, even over
our (supposed) 'free will'. And they did this in order to rationalise the status quo; if
the nature of the state and class domination have been 'determined', or even 'ordained',
by 'God', they can't be opposed or questioned -- at least, not 'legitimately'.
Hence, Marxists find that they, too, will have to use such distorted,
anthropomorphic terminology if
they follow traditional patterns of thought in this area.
"Feuerbach's
great achievement is.... The proof that philosophy is nothing else but
religion rendered into thought and expounded by thought, i.e., another form
and manner of existence of the estrangement of the essence of man; hence equally
to be condemned...." [Marx
(1975b), p.381. I have used the on-line version, here.]
"The philosophers have only to dissolve their language
into the ordinary language, from which it is abstracted, in order to recognise
it, as the distorted language of the actual world, and to realise that neither
thoughts nor language in themselves form a realm of their own, that they are
only manifestations of actual life." [Marx
and Engels (1970), p.118.]
And:
"The ideas of the ruling class are in every epoch the ruling ideas, i.e. the
class which is the ruling material force of society, is at the same time its
ruling intellectual force. The class which has the means of material production
at its disposal, has control at the same time over the means of mental
production, so that thereby, generally speaking, the ideas of those who lack the
means of mental production are subject to it. The ruling ideas are nothing more
than the ideal expression of the dominant material relationships, the dominant
material relationships grasped as ideas; hence of the relationships which make
the one class the ruling one, therefore, the ideas of its dominance. The
individuals composing the ruling class possess among other things consciousness,
and therefore think. Insofar, therefore, as they rule as a class and determine
the extent and compass of an epoch, it is self-evident that they do this in its
whole range, hence among other things rule also as thinkers, as producers of
ideas, and regulate the production and distribution of the ideas of their age:
thus their ideas are the ruling ideas of the epoch. For instance, in an age and
in a country where royal power, aristocracy, and bourgeoisie are contending for
mastery and where, therefore, mastery is shared, the doctrine of the separation
of powers proves to be the dominant idea and is expressed as an 'eternal law.'"
[Ibid.,
pp.64-65.]
These concepts will "rule" us, too, if we uncritically appropriate traditional
thought-forms like the above.
~~~~~~oOo~~~~~~
Many of the
above
ideas aren't original to me (although a Marxist re-application of them most
definitely is). They
first saw light of day, as far as I am aware, in Bertrand Russell's essay:
Russell, B. (1917a), 'On The Notion Of A Cause', in Russell (1917b), pp.132-51.
--------, (1917b), Mysticism And Logic (George Allen & Unwin).
Although there is a hint of this in David Hume's work.
They can also be found
more explicitly stated (but not from a
Marxist angle) in the following:
Gallop, D. (1962), 'On Being Determined', Mind71, pp.181-96.
I have also adapted the following analysis of the phrase 'physical law':
Swartz, N. (1985), The Concept Of A Physical Law (Cambridge University
Press).
--------, (2009), 'Laws
Of Nature', Internet Encyclopedia of Philosophy.
Influential Wittgensteinian criticisms of modern scientistic philosophies of
'mind' can be found here:
Bennett, M., and Hacker, P. (2008), History Of Cognitive Neuroscience (Blackwell).
--------,
(2022), Philosophical Foundations Of Neuroscience
(Blackwell).
Also see:
Hacker, P. (2007), Human
Nature, The Categorial Framework (Blackwell). [A PDF of Chapter One can be downloaded
from
here.]
--------, (2013a), The Intellectual Powers. A Study Of Human
Nature (Wiley Blackwell).
Those who think an appeal to ordinary language is inappropriate in this area should
re-read what Marx said above, consult the first half of
the following, and then
perhaps think again:
Button, G., Coulter, J., Lee, J., and Sharrock, W. (1995), Computers, Minds
And Conduct (Polity Press). [A key passage from this book has been quoted
here.]
The bottom line is that Marxists have been
only too eager to
appropriate concepts and forms-of-thought bequeathed to us by Traditional
Thinkers,
without subjecting them to sufficient critical scrutiny. Unfortunately, this means that
while our politics might appear to be radical, our theory (both here and in
relation to dialectics, for example) is thoroughly traditional -- and, if
I may say so, inappropriately conservative.
I explain why I allege this in the first few sections of the following Essay:
I have also explained why I reject Traditional Thought as incoherent non-sense,
here -- summarised
here.
Finally, I
would try to get (more polished versions of) this material published in Marxist journals, etc., but I
am in general treated as a pariah, and face highly emotional and irrational
hostility wherever I attempt to present these ideas. It seems, therefore, that
"ruling ideas" rule comrades who are
also editors!
Finally, the above implies that 'determinism' either makes no sense (since it depends on
the projection of human capacities onto the world, and even onto the brain!), or
it is a misnomer. Moreover, because the traditional doctrine of 'free
will' depends for its force on being the obverse of 'determinism', it too makes
no sense (since it abstracts human beings from history and society).
So, do we have 'free will'? That question is far too abstract. What we need to
pay detailed attention to is how we use words associated with our psychological
make-up and the decisions we make. I won't attempt that here.
Several of Marx and Engels's works listed below have
been linked to the Marxist Internet Archive, but since Lawrence & Wishart
threatened legal action over copyright infringement many no
longer work.
However, all of their work can now be accessed
here.
Aarsleff, H. (1970), 'The History Of
Linguistics And Professor Chomsky', Language46, pp.570-85;
reprinted in Aarsleff (1982), pp.101-19.
--------, (1982), From Locke To Saussure.
Essays On The Study Of Language And Intellectual History (Athlone Press).
Ahmed, A. (2010) (ed.), Wittgenstein's
Philosophical Investigations. A Critical Guide (Cambridge University Press).
--------, (1985), 'Wittgenstein And The
Vienna Circle: The Exaltation And Deposition Of Ostensive Definition', Teoria
5, pp.5-33, reprinted in Hacker (2001), pp.242-67.
--------, (1988), Wittgenstein. Rules,
Grammar And Necessity Volume Two (Blackwell, 2nd
ed.).
--------, (2005a), Wittgenstein: Understanding And Meaning.
Volume One Of An Analytic Commentary On The Philosophical Investigations, Part I
-- Essays (Blackwell,
2nd
ed.).
--------, (2005b), Wittgenstein: Understanding And Meaning. Volume One Of An
Analytic Commentary On The Philosophical Investigations, Part II -- Exegesis
§§1-184 (Blackwell,
2nd ed.).
Ball, T. (1979), 'Marx And Darwin.
A Reconstruction', Political Theory7,
pp.469-83.
Barnes, J. (1995) (ed.), The Cambridge Companion
To Aristotle (Cambridge University Press).
--------, (2009), Truth, Etc. Six Lectures
On Ancient Logic (Oxford University Press).
Barnes, B., and Dupré, J. (2008), Genomes
And What To Make Of Them (Chicago University Press).
Beaken, M. (1996), The Making Of Language
(Edinburgh University Press).
Behme, C. (2014a), 'A
Potpourri Of Chomskyan Science'. [Unfortunately, this link is now
dead, but a corrected version of this paper can be
accessed here.
This links to a PDF.]
Benacerraf, P., and Putnam, H. (1964) (eds.),
Philosophy Of Mathematics. Selected Readings (Blackwell).
--------, (1983),
Philosophy Of Mathematics. Selected Readings (Cambridge University Press,
2nd
ed.).
Bennett, M., and Hacker, P. (2008), History Of Cognitive
Neuroscience (Blackwell).
--------,
(2021),
Philosophical Foundations Of Neuroscience (Blackwell, 2nd
ed.).
Bennett, M., Dennett, D., Hacker, P., and
Searle, J. (2007), Neuroscience And Philosophy. Brain, Mind And Language
(Columbia University Press). [Part of this can be found
here (this links to a PDF), and there is a .wav recording of this debate
available
here (this takes a few minutes to load!).]
Block, I. (1981) (ed.), Perspectives On
The Philosophy Of Wittgenstein (Blackwell).
Bloomfield, B. (1987) (ed.), The Question
Of Artificial Intelligence (Croom Helm).
Bloor, D. (1997), Wittgenstein, Rules And
Institutions (Routledge).
Bobro, M. (2017), 'Leibniz
On Causation',
Stanford Encyclopedia of Philosophy, edited by Edward N. Zalta (Fall 2017 Edition).
Bogen, J. (1972), Wittgenstein's
Philosophy Of Language (Routledge).
Bono, J. (1995), The Word Of God And The Language Of Men, Ficino To
Descartes, Volume One: Interpreting Nature In Early Modern Science And Medicine
(University of Wisconsin Press).
Born, R. (1987) (ed.), Artificial
Intelligence. The Case Against (Croom Helm).
Bouveresse, J. (1995), Wittgenstein Reads
Freud. The Myth Of The Unconscious, translated by Carol Cosman (Princeton
University Press).
Boyd, R., and Richerson, P. (2005), The
Origin And Evolution Of Cultures (Oxford University Press).
Braverman, H. (1999),
Labor And Monopoly
Capitalism. The Degradation Of Work In The Twentieth
Century (Monthly Review
Press, 2nd
ed.).
Bridges, J. (2010), 'Wittgenstein vs
Contextualism', in Ahmed (2010), pp.109-28.
Broad, W., and Wade, N. (1985), Betrayers
Of The Truth. Fraud And Deceit In The Halls Of Science (Oxford University
Press).
Brook, A., and Ross, D., (2002) (ed.),
Daniel Dennett (Cambridge University Press).
--------, (1999b), The Darwin Wars (Simon &
Schuster).
Brown, D. (1970), 'Knowing How And Knowing That, What', in Wood
and Pitcher (1970), pp.213-48.
Browne, J. (2003),
Charles Darwin. The Power Of Place. Volume Two Of A Biography
(Pimlico).
Budd, M. (1989), Wittgenstein's Philosophy
Of Psychology (Routledge).
Buller, D. (2005), Adapting Minds.
Evolutionary Psychology And The Persistent Quest For Human Nature (MIT
Press).
--------, (2007), 'Varieties Of Evolutionary
Psychology', in Hull and Ruse (2007), pp.255-74.
Burkert, W. (1972), Lore And Science In
Ancient Pythagoreanism (Harvard University Press).
--------, (1985), Greek Religion, translated by
John Raffan (Blackwell).
Burling, R. (2005), The Talking Ape. How
Language Evolved (Oxford University Press).
Burnyeat, M. (1987), 'Wittgenstein
And Augustine De Magistro', Proceedings of the Aristotelian
Society, Supplementary Volume61, pp.1-24. [This links to a
PDF.]
Button, G., Coulter, J., Lee, J., and
Sharrock, W. (1995), Computers, Minds And Conduct (Polity Press).
Carruthers, P. (1984), 'Critical Study: Baker
And Hacker's Wittgenstein', Synthèse58, pp.451-79.
--------, (2006), The Architecture Of The Mind (Oxford
University Press).
Carruthers, P., and Boucher, J. (1998)
(eds.), Language And Thought. Interdisciplinary Themes (Cambridge
University Press).
Carruthers, P., Laurence, S., and Stich, S.
(2005) (eds.), The Innate Mind. Volume One: Structure And Contents
(Oxford University Press).
--------, (2006) (eds.), The Innate Mind.
Volume Two: Culture And Cognition (Oxford University Press).
--------, (2007) (eds.), The Innate Mind.
Volume Three: Foundations And The Future (Oxford University Press).
Carstairs-McCarthy, A. (1999), The Origins
Of Complex Language. An Inquiry Into The Evolutionary Beginnings Of Sentences,
Syllables, And Truth (Oxford University Press).
Cartwright, N. (2000), 'An Empiricist Defence Of Singular
Causes', in Teichmann (2000), pp.47-58.
Cattell, N. (2006), An Introduction To
Mind, Language And Consciousness (Continuum).
Cavell, S. (1971a), 'Must We
Mean What We Say?', in Lyas (1971), pp.131-65, and Cavell (1976), pp.1-43.
--------, (1971b), 'The
Availability Of Wittgenstein's Later Philosophy', in Lyas (1971), pp.166-89, and
Cavell (1976), pp.44-72.
--------, (1976), Must We
Mean What We Say? (Cambridge University Press).
--------, (1996), 'Notes And
Afterthoughts On The Opening Of Wittgenstein's "Investigations"', in Sluga and
Stern (1996), pp.261-95.
Charles, D., and Child, W. (2001) (eds.),
Wittgensteinian Themes (Oxford University Press).
Chase, A. (1976), The Legacy Of Malthus. The Social Costs Of The New
Scientific Racism (Alfred Knopf).
Child, W. (1996), Causality, Interpretation And The Mind
(Oxford University Press).
Chomsky, N. (1965), Aspects Of A Theory Of
Syntax (MIT Press).
--------, (1972), Language And Mind
(Harcourt Brace Jovanovich, 2nd
ed.).
--------, (1988), Language And the Problems Of Knowledge. The Managua
Lectures (MIT Press).
--------, (2000a), New Horizons In The
Study Of Language And Mind (Cambridge University Press).
--------, (2000b), The Architecture Of
Language (Oxford University Press).
--------, (2002), On Nature And Language (Cambridge University Press).
--------, (2005a), Rules And Representations
(Columbia University Press, 2nd
ed.).
--------, (2005b), 'Some Simple Evo Devo
Theses. How True Might They Be For Language?' Paper presented at the Alice V.
and David H. Morris Symposium on Language and Communication, Stony Brook,
SUNY, New York. [I haven't yet been able to check this source.]
--------, (2007), 'Minds
And Language', Biolinguistics1, pp.9-27. [This links to a PDF, but
access requires (free) registration.]
--------, (2009a), Cartesian Linguistics. A
Chapter In The History Of Rationalist Thought (Cambridge University Press, 3rd
ed.).
--------, (2009b), 'Concluding Remarks', in
Piattelli-Palmarini, et al (2009), pp.379-401. [I haven't yet been able to
check this source.]
--------, (2012a),
The Science Of Language: Interviews With
James McGilvray
(Cambridge University Press).
[I haven't yet been able to check this source.]
Collins, C. (1999),
Language, Ideology And Social Consciousness: Developing A Sociohistorical
Approach (Ashgate).
--------, (2000), 'Vygotsky On Language And
Social Consciousness: Underpinning The Use Of Voloshinov In The Study Of Popular
Protest', Historical Materialism7, pp.41-69.
--------, (2003), 'Review Of Bodies Of
Meaning: Studies On Language, Labor And Liberation, By David McNally',
Historical Materialism11, 2, pp.227-38.
--------, (2004), 'Marxism And Language: A
Response To McNally', Historical Materialism12, 2, pp.169-87.
Cook, J. (1978a), 'Whorf's Linguistic
Relativism I', Philosophical Investigations 1, 1, pp.1-30.
--------, (1980), 'The Fate Of Ordinary Language
Philosophy', Philosophical Investigations3, 2, pp.1-72.
--------, (1982), 'The Illusion Of Aberrant
Speakers', Philosophical Investigations5, pp.215-26.
Copleston, F. (2003), A History Of
Philosophy, Volume Two: Medieval Philosophy (Continuum).
Cornforth, M.
(1965), Marxism And Linguistic Philosophy (Lawrence & Wishart).
--------, (1974), Dialectical Materialism,
Volume Three: The Theory Of Knowledge (Lawrence & Wishart, 3rd
ed.).
--------, (1976), Materialism And The
Dialectical Method (Lawrence & Wishart, 5th ed.).
[A copy of the 1968 edition is available
here.]
Cottingham, J. (1992) (ed.), The Cambridge
Companion To Descartes (Cambridge University Press).
Coulter, J. (1983), Rethinking Cognitive
Theory (Macmillan).
--------, (1989), Mind In Action (
Humanities Press).
--------, (1993), 'Consciousness: The
Cartesian Enigma', in Canfield and Shanker (1993), pp.173-94.
--------, (1997), 'Neural Cartesianism:
Comments On the Epistemology Of The Cognitive Sciences', in Johnson and Erneling
(1997), pp.293-301.
Coulter, J., and Sharrock, W. (2007),
Brain, Mind, And Human Behavior In Contemporary Cognitive Science. Critical
Assessments Of The Philosophy Of Psychology (The Edwin Mellen Press).
Cowie, F. (1997), 'The Logical Problem Of
Language Acquisition', Synthèse111, pp.17-51.
--------, (1995), Darwin's Dangerous Idea (Allen
Lane).
Descartes, R., (1997a), Descartes Key
Philosophical Writings, translated by E. S. Haldane and G. T. R. Ross
(Wordsworth Classics).
--------, (1997b), Discourse On The Method,
in Descartes (1997a), pp.71-122. [This links to a different translation.]
Deutscher, G. (2011), Through The Language Glass. Why The
World Looks Different In Other Languages (Arrow Books).
Dever, J. (2008), 'Compositionality', in Lepore and Smith,
pp.633-66.
Devitt, M. (2006), Ignorance Of Language (Oxford
University Press).
Diamond, C. (1995), The Realistic Spirit
(MIT Press).
--------, (1997), 'Realism And Resolution: Reply To
Warren Goldfarb And Sabina Lovibond', Journal of Philosophical Research
22, pp.75-86.
--------, (1999), 'How Old Are These Bones.
Putnam, Wittgenstein And Verification', Proceedings of the Aristotelian
Society Supplementary Volume73, pp.99-134.
--------, (2010), 'Inheriting From Frege: The Work Of Reception,
As Wittgenstein Did It', in Potter and Ricketts, pp.550-601.
Dreyfus, H. (1992), What Computers Still
Can't Do (MIT Press). [There is an extended summary of this book
here, and a discussion
here.]
Dummett, M. (1956), 'Nominalism',
Philosophical Review65, 491-505, reprinted in Klemke (1968),
pp.321-36.
--------, (1959), 'Wittgenstein's Philosophy Of Mathematics', Philosophical Review68, pp.324-48, reprinted in Benacerraf and Putnam (1964), pp.491-509, in
Dummett (1978), pp.166-85, and in Pitcher (1968), pp.420-47.
--------, (1978), Truth And Other Enigmas
(Duckworth).
--------,
(1981a), Frege. Philosophy Of Language (Duckworth, 2nd
ed.).
--------, (1981b), The Interpretation Of
Frege's Philosophy (Duckworth).
--------, (1981c), 'Frege And Wittgenstein',
in Block (1981), pp.31-42, reprinted in Dummett (1991b), pp.237-48.
--------, (1984), 'An Unsuccessful Dig', Philosophical Quarterly34,
pp.379-401, reprinted in Wright (1984), pp.194-226, and, with a few changes, in Dummett (1991b), pp.158-98.
--------, (1987), 'Frege And The Paradox Of Analysis', a lecture given at
Bologna University, reprinted in Dummett (1991b), pp.17-52.
--------, (1988), 'A Reply To "Dummett's Dig", by Baker And Hacker',
Philosophical Quarterly38, pp.87-103; revised and reprinted as
'Second Thoughts' in Dummett (1991b), pp.199-216.
--------, (1990),
'Language And Communication', in George (1990), pp.192-212; reprinted in Dummett
(1993a), pp.166-87.
--------, (1991a),
Frege. Philosophy Of Mathematics (Duckworth).
--------, (1991b), Frege And Other
Philosophers (Oxford University Press).
--------, (1993a), The Seas Of Language (Oxford
University Press).
--------,
(1993b), Origins Of Analytic Philosophy (Duckworth).
Dunbar, R., Knight, C., and Power, C. (1999)
(eds.), The Evolution Of Culture (Edinburgh University Press).
Dupré, J. (1991), 'Conversations With Apes.
Reflections On The Scientific Study Of Language', in Hyman (1991), pp.95-116,
reprinted in Dupré (2002), pp.236-56.
--------, (2001), Human Nature And The
Limits Of Science (Oxford University Press).
--------, (2002), Humans And Other Animals
(Oxford University Press).
--------, (2003a), Darwin's Legacy. What
Evolution Means Today (Oxford University Press).
--------, (2003b), 'On
Human Nature', Journal of
the Slovakian Academy of Sciences 13, pp.109-22, and Human
Affairs, 13, 2, pp.109-22.
--------, (2012), Processes Of Life.
Essays On The Philosophy Of Biology (Oxford University Press).
Eco, U. (1997), The
Search For The Perfect Language (Fontana).
Egidi, R. (1995) (ed.),
Wittgenstein: Mind And Language (Kluwer Academic Press).
Eldredge, N. (1996), Reinventing Darwin.
The Great Evolutionary Debate (Phoenix Press).
--------, (2004),
Why We Do It. Rethinking Sex And The Selfish Gene (W. W. Norton & Co.).
Erneling, C. (1993), Understanding
Language Acquisition (State University of New York Press).
Erneling, C., and Johnson, D. (2005) (eds.),
The Mind As A Scientific Object. Between Brain And Culture (Oxford
University Press).
Evans, V. (2014), The Language Myth: Why Language Is Not An Instinct
(Cambridge University Press).
Everett, D. (2008), Don't Sleep, There Are
Snakes. Life And Language In The Amazonian Jungle (Profile Books).
--------, (2012), Language: The Cultural Tool (Profile
Books).
Everson, S. (1995), 'Psychology', in Barnes (1995), pp.168-94.
Fann, K. (1978) (ed.),
Ludwig Wittgenstein: The Man And His Philosophy (Harvester Press).
Feyerabend, P., and Maxwell, G. (1966) (eds.), Mind, Matter,
And Method, Essays In The Philosophy Of Science In Honor Of Herbert Feigl
(University of Minnesota Press).
Field, H. (2008),
Saving Truth From Paradox (Oxford University Press).
Fischer, E. (2011a), 'Diseases Of The Understanding And The Need For
Philosophical Therapy', Philosophical Investigations34, 1,
pp.22-54.
--------, (2011b) Philosophical Delusion
And Its Therapy: Outline Of A Philosophical Revolution (Routledge).
Floyd, J. (1991), 'Wittgenstein On 2,2,2…: On
The Opening Of Remarks On The Foundations Of Mathematics', Synthèse87, pp.143-80.
Fodor, J. (1975), The Language Of Thought
(Harvard University Press).
--------, (1980), 'Methodological Solipsism Considered As A
Research Strategy In Cognitive Psychology', The Behavioral and Brain Sciences,
3, pp.63-73.
--------, (1998a), In Critical Condition. Polemical
Essays On Cognitive Science And The Philosophy Of Mind (MIT Press).
--------, (1998b), 'Review Of Richard Dawkins Climbing Mount
Improbable', in Fodor (1998a), pp.163-69.
--------, (1998c), 'Deconstructing Dennett's Darwin', in Fodor
(1998a), pp.171-87.
--------, (1998d), Concepts: Where
Cognitive Science Went Wrong (Oxford University Press).
--------, (2001), The Mind Doesn't Work
That Way (MIT Press).
--------, (no date), 'Against
Darwinism'. [As far as I am aware, this paper is unpublished, but the author
tells me that it has now been incorporated into Fodor and Piattelli-Palmarini (2010).]
Fodor, J., and Piattelli-Palmarini, M.
(2010), What Darwin Got Wrong (Profile Books). [Section One is available
here as a PDF.]
Fogelin, R. (1987),
Wittgenstein (Routledge, 2nd
ed.).
Francis, R. (2012), Epigenetics: How Environment Shapes Our
Genes (W. W. Norton & Co.).
Frawley, W. (1997), Vygotsky And Cognitive
Science. Language And The Unification Of The Computational Mind (Harvard
University Press).
Frege, G.
(1914), 'Logic In Mathematics', reprinted in Frege (1979), pp.203-50.
--------, (1923), 'Compound Thoughts', translated by R. H. Stoothoff, reprinted in Frege (1977), pp.55-77,
and Klemke (1968), pp.537-58.
--------, (2000), 'More Things In Heaven And Earth', in Rose and
Rose (2000), pp.85-105; reprinted in Gould (2006a), pp.438-60.
--------, (2001), 'The
Evolutionary Definition Of Selective Agency, Validation Of The Theory Of
Hierarchical Selection, And Fallacy Of The Selfish Gene', in
Singh, et al (2001),
pp.208-34.
--------, (2002a),
The Structure Of Evolutionary Theory (Harvard University Press).
--------, (2002b), I Have Landed. Splashes
And Reflections In Natural History (Jonathan Cape).
--------, (2006a), The Richness Of Life.
The Essential Stephen Jay Gould, edited by Paul McGarr; Introduction
by Steven Rose (Jonathan Cape).
--------, (2006b), 'Challenges To
Neo-Darwinism And Their Meaning For A Revised View Of Human Consciousness',
Tanner Lecture 1984, reprinted in Gould (2006a), pp.222-37.
Gould, S., and Vrba, E. (1982), 'Exaptation
-- A Missing Term In The Science Of Form', Paleobiology8, 1,
pp.4-15; reprinted in Hull and Ruse (1996), pp.52-71, and in Gould (2006a),
pp.417-37.
Greenspan, S., and Shanker, S. (2004), The
First Idea. How Symbols, Language, And Intelligence Evolved From Our Primate
Ancestors To Modern Humans (Da Capo Press).
Grice, P.
(1957), 'Meaning',
Philosophical Review66, 3, pp.377-88 (this links to a PDF),
reprinted in Grice (1989), pp.213-23.
--------,
(1989), Studies In The Way Of
Words (Harvard University Press).
Griffin, D. (2001), Animal Minds. Beyond
Cognition To Consciousness (University of Chicago Press, 2nd
ed.).
Guldberg, H. (2010), Just Another Ape? (Imprint Academic).
Guthrie, W. (1950), The Greeks And Their
Gods (Methuen).
--------, (1990), A History Of Greek
Philosophy, Volume Six,Aristotle: An Encounter (Cambridge University
Press).
Guttenplan, S. (1975) (ed.), Mind And
Language (Oxford University Press).
Haack, S. (1978), Philosophy Of Logics (Cambridge
University Press).
Hacker, P. (1987), Appearance And Reality
(Blackwell).
--------, (1991), 'Seeing, Representing And
Describing', in Hyman (1991), pp.119-54.
--------, (1993a), Wittgenstein. Meaning And Mind,
Volume Three Part One (Blackwell, 2nd
ed.).
--------, (1993b), 'The Inner And The Outer',
in Hacker (1993a), pp.127-41.
--------, (1993c), 'Thinking: Methodological
Muddles And Categorical Confusions', in Hacker (1993a), pp.143-60.
--------, (1993d), 'Thinking: The Soul Of
Language', in Hacker (1993a), pp.161-82.
--------, (1993e), 'Behaviour And Behaviourism', in Hacker
(1993a), pp.97-126.
--------, (1993f), 'The Mind', in Hacker (1993a), pp.64-67.
--------, (1993g), 'Homunculi And Brains', in Hacker (1993a),
pp.69-72.
--------, (1996),
Wittgenstein's Place In Twentieth Century Analytic Philosophy (Blackwell).
--------, (1997), Insight And Illusion
(Thoemmes Press, 2nd
ed.).
--------, (2000a), 'On Carnap's Elimination
Of Metaphysics', Deutsche Zeitschrift für Philosophie (2000), pp.469-86;
reprinted in Hacker (2001a), pp.324-44.
--------, (2000b), Wittgenstein. Mind And
Will. Volume Four Part One (Blackwell, 2nd
ed.).
--------, (2000c), Wittgenstein. Mind And
Will. Volume Four Part Two (Blackwell, 2nd
ed.).
--------, (2001a), Wittgenstein:
Connections And Controversies (Oxford University Press).
--------, (2001b), 'Eliminative Materialism',
in Schroeder (2001a), pp.60-84.
--------, (2001c), 'Philosophy', in Glock (2001), pp.322-47.
Hacking, I. (1975), Why Does Language
Matter To Philosophy? (Cambridge University Press).
--------, (1983), Representing And
Intervening. Introductory Topics In The Philosophy Of Science (Cambridge
University Press).
Hale,
B., and Wright, C. (eds.) (1999), A Companion To The Philosophy Of Language
(Blackwell).
Hallett, G. (1967), Wittgenstein's
Definition Of Meaning As Use (Fordham University Press).
--------, (1977), A Companion To
Wittgenstein's 'Philosophical Investigations' (Cornell University Press).
--------, (1984), Logic For The Labyrinth
(University Press of America).
--------, (1988), Language And Truth
(Yale University Press).
--------, (1991), Essentialism: A
Wittgensteinian Critique (State University of New York Press).
--------, (2008), Linguistic Philosophy.
The Central Story (State University of New York Press).
Hanfling, O. (1981), Logical Positivism
(Blackwell).
--------, (1989), Wittgenstein's Later
Philosophy (Macmillan).
--------, (2000), Philosophy And Ordinary
Language (Routledge).
--------, (2001a), 'Thinking', in Glock
(2001), pp.138-55; reprinted in Hanfling (2002a), pp.133-48.
--------, (2001b), 'Consciousness: "The Last
Mystery"', in Schroeder (2001a), pp.36-59; reprinted in Hanfling (2002a),
pp.104-27, but in a revised form as Hanfling (2002b)
--------, (2002a), Wittgenstein And The
Human Form Of Life (Routledge).
--------, (2002b), 'Wittgenstein And The
Problem Of Consciousness', in Hanfling (2002a), pp.104-27.
Hanna, P., and Harrison, B. (2004), Word
And World. Practice And The Foundations Of Language (Cambridge University
Press).
Hanson, N. (1971a),
What I Do Not Believe, And Other Essays (Reidel).
--------, (1971b), 'Mental Events Yet Again;
Retrospect On Some Old Arguments', in Hanson (1971a), pp.190-212.
Hanson, S., and Bunzl, M. (2010) (eds.), Foundational Issues In Human
Brain Mapping (MIT Press).
Hardin, C. (1993), Color For Philosophers
(Hackett Publishing Company, 2nd
ed.).
Hark, M. (1990), Beyond The Inner And The
Outer: Wittgenstein's Philosophy Of Psychology (Kluwer Academic Press).
--------, (1995), 'Electric Brain Fields And
Memory Traces: Wittgenstein And Gestalt Psychology', Philosophical
Investigations18, pp.113-38.
Harman, C. (1992), 'Blood Simple', International
Socialism54, pp.169-75.
--------,
(2006b), A New History Of Western Philosophy, Volume Three: The Rise Of
Modern Philosophy (Oxford University Press).
--------, (2009), 'Cognitive Scientism', in Glock and Hyman
(2009), pp.250-62.
Kirschner, M., and Gerhart, J. (2005), The Plausibility Of Life. Resolving
Darwin's Dilemma (Yale University Press).
Kirwan, C. (2001),
'Augustine's Philosophy Of Language', in Stump and Kretzmann (2001), pp.186-204.
Kitcher, P.
(1985), Vaulting Ambition. Sociobiology And The Quest For Human Nature
(MIT Press).
--------,
(1998), 'A Plea For Science Studies', in Koertge (1998), pp.32-56.
Kitching, G., and Pleasants, N. (2002)
(eds.), Marx And Wittgenstein. Knowledge, Morality And Politics
(Routledge).
Klemke, E. (1968) (ed.), Essay On Frege (University of Illinois Press).
Knight, C. (1995),
Blood Relations. Menstruation And The Origins Of Culture
(Yale University Press).
--------, (2003), 'Noam
Chomsky: Politics Or Science?', What Next?26, pp.17-29. [The
What Next? site appears to be down. This
article is now available
here, and
here (the second link takes the reader to a downloadable PDF).]
--------, (2018), Decoding Chomsky: Science And Revolutionary Politics
(Yale University Press, 2nd
ed.). The Preface to this edition can be accessed
here, although visitors will need a Microsoft account to read it.]
Knight, C., Studdert-Kennedy, M., and
Hurford, J. (2000) (eds.), The Evolutionary Emergence Of Language. Social
Function And The Origins Of Linguistic Form (Cambridge University Press).
Koertge, N. (1998) (ed.), A House Built On
Sand (Oxford University Press).
Kölbel, M., and Weiss, B.
(2004) (eds.),
Wittgenstein's Lasting Significance (Routledge).
Kretzmann, N. (1993), 'Philosophy Of Mind', in
Kretzmann and Stump (1993), pp.128-59.
Kretzmann, N., and Stump, E. (1993) (eds.),
The Cambridge Companion To Aquinas (Cambridge University Press).
Kripke, S. (1982), Wittgenstein On Rules And Private
Language (Blackwell).
Kuper, A. (2000), 'If Memes Are The Answer, What Is The
Question?', in Aunger (2000), pp.175-88.
Kurzweil, R. (2013),
How To Create A Mind: The Secret Of Human Thought Revealed
(Duckworth).
Kusch, M. (2006), A
Sceptical Guide To Meaning And Rules. Defending Kripke's Wittgenstein
(Acumen).
Kuusela, O. (2005), 'From Metaphysics And
Philosophical Theses To Grammar: Wittgenstein's Turn', Philosophical
Investigations28, 2, pp.95-133.
--------, (2006), 'Do
Concepts Of Grammar And Use In Wittgenstein Articulate A Theory Of Language Or
Meaning', Philosophical Investigations 29, 4, pp.309-41.
--------, (2008), The Struggle Against
Dogmatism. Wittgenstein And The Concept Of Philosophy (Harvard University
Press).
Kuusela, O., and McGinn, M. (2011), The
Oxford Handbook Of Wittgenstein (Oxford University Press).
Kuusinen, O. (1961) (ed.), Fundamentals Of Marxism-Leninism
(Lawrence & Wishart).
Lambert, M. (1992), Medieval Heresy.
Popular Movements From The Gregorian Reform To The Reformation (Blackwell, 2nd
ed.).
Land,
S. (1974), From Signs To Propositions. The Concept Of Form In
Eighteenth-Century Semantic Theory (Longman).
Lantolf, J., and Appel,
G. (1994) (eds.), Vygotskian Approaches To Second Language Research
(Ablex Publishing).
Laudan, L. (1977), Progress And Its
Problems (University of Californian Press).
Leahy, T. (2005), 'Mind As A Scientific Object: A
Historical-Philosophical Exploration', in Erneling and Johnson (2005), pp.35-78.
Lecercle, J-J. (2006), A Marxist
Philosophy Of Language, translated by Gregory Elliott (Brill).
Lecercle, J-J., and Riley, D. (2004), The
Force Of Language (Palgrave).
Malcolm, N. (1968), 'The Conceivability Of
Mechanism', Philosophical Review78, pp.45-72; reprinted in
Malcolm (1977a), pp.1-40.
--------, (1977a), Thought And Knowledge (Cornell University Press).
--------, (1977b), Memory And Mind (Cornell
University Press).
--------, (1980), '"Functionalism" In
Philosophical Psychology', Proceedings of the Aristotelian Society80,
pp.211-29; reprinted in Malcolm (1995a), pp.27-44.
--------, (1984), 'Consciousness And Causality' in Armstrong and
Malcolm (1984), pp.3-101, 193-204.
--------, (1986a), Wittgenstein: Nothing
Is Hidden (Blackwell).
--------, (1986b), 'Thoughts', in Malcolm
(1986a), pp.63-82.
--------, (1986c), 'Mind And Brain', in
Malcolm (1986a), pp.182-200.
--------, (1995a), Wittgensteinian Themes
(Cornell University Press).
--------, (1995b), 'Thinking', in Malcolm (1995a), pp.1-15.
--------, (1995c), 'The Mystery Of Thought', in Malcolm (1995a),
pp.182-94.
--------, (1997a), Wittgenstein. A
Religious Point Of View?, edited with a response by Peter Winch (Routledge).
--------, (1997b), 'Underlying Mechanisms' in
Malcolm (1997a), pp.48-57.
--------,
(2001), Ludwig Wittgenstein: A
Memoir (Oxford University Press, 2nd
ed.).
Malik, K. (2001), Man, Beast And Zombie.
What Science Can And Can't Tell Us About Human Nature (Phoenix Press).
Mandel, E. (1979), Introduction To Marxism (Pluto
Press, 2nd ed.).
Marcuse, H. (1968),
One Dimensional Man. Studies In The Ideology Of Advanced Industrial
Society (Abacus Books).
Markus, R. (1970), 'Marius Victorinus And
Augustine', in Armstrong (1970a), pp.331-419.
Martin, R. (1987), The Meaning Of Language
(MIT Press).
Martinich, A., and Sosa, D. (2005) (eds.),
A Companion To Analytic Philosophy (Blackwell).
Monk, R. (1990), Wittgenstein. The Duty Of
Genius (Vintage).
--------, (2007), 'Bourgeois, Bolshevist Or
Anarchist? The Reception Of Wittgenstein's Philosophy Of Mathematics', in
Kahane, Kanterian and Kuusela (2007), pp.269-94.
Montague, A.
(2008), Man's Most Dangerous Myth: The Fallacy Of Race (Whitley Press, 2
nd
ed.).
--------, (2002), (ed.) Race And IQ.
Expanded Edition (Oxford University Press).
Morris, M. (2007), An Introduction To The
Philosophy Of Language (Cambridge University Press).
Moss, L. (2003), What Genes Can't Do
(MIT Press).
Niles, I. (1992), 'Wittgenstein And Infinite
Linguistic Competence', in French, et al (1992), pp.193-213.
Nowak, M., and Highfield, R. (2012), Super
Cooperators: Beyond The Survival Of The Fittest. Why Cooperation, Not
Competition, Is The Key To Life (Canongate).
Onians, R. (2000), The Origins Of European
Thought About The Body, The Mind, The Soul, The World, Time, And Fate
(Cambridge University Press, 2nd
ed.).
Orr, H. (1996a), 'Dennett's Dangerous Idea', Evolution50, 1,
pp.467-72.
Otto, W. (2001), The Homeric Gods. The
Spiritual Significance Of Greek Religion (Ayer Company Reprints).
Owen, G. (1966), 'The Platonism Of Aristotle', Proceedings of the British
Academy51, pp.125-50, reprinted in Owen (1986), pp.200-20.
--------, (1986), Logic, Science And Dialectic. Collected Papers In Greek
Philosophy (Duckworth).
Owens, D. (1992), Causes And Coincidences
(Oxford University Press).
Palmer, A. (1988), Concept And Object. The
Unity Of The Proposition In Logic And Psychology (Routledge).
--------, (2011), 'Propositions, Properties And Relations: Wittgenstein's
Notes On Logic And The Tractatus', Philosophical Investigations34, 1, pp.77-93.
Pasnau, R. (2002), 'Human Nature', in McGrade
(2002), pp.208-30.
--------, (2015), 'Divine
Illumination', Stanford Encyclopedia of Philosophy, edited by Edward
N. Zalta (Spring 2015 Edition).
Passmore, J. (1966), A Hundred Years Of
Philosophy (Penguin Books).
Pateman, T. (1987),
Language In Mind And Language In Society (Oxford University Press).
Pepperberg, I. (2008), Alex And Me. How A Scientist And A Parrot Discovered
A Hidden World Of Animal Intelligence -- And Formed A Deep Bond In The Process
(Harper Collins).
Perloff, M. (1996), Wittgenstein's Ladder
(University of Chicago Press).
Peters, S., and Westerathåhl, D.
(2008), Quantifiers In Language And Logic (Oxford University Press).
Piattelli-Palmarini, M., Uriagereka, J., and
Salaburu, P. (2009) (eds.), Of Minds And Language: A Dialogue With Noam
Chomsky In The Basque Country (Oxford University Press). [I haven't yet been
able to check this source.]
Pinker, S. (1994), The Language Instinct (Penguin
Books).
--------, (1997), How The Mind Works
(Penguin Books).
--------, (2003), The Blank Slate. The
Modern Denial Of Human Nature (Penguin Books).
--------, (2007), The Stuff Of Thought
(Allen Lane).
Pinker, S., and Bloom, P. (1990), 'Natural Language And Natural
Selection', Behavioral and Brain Sciences13, 4, pp.707-84. A pre-print of this paper is available
here.
Pitcher, G. (1968)
(ed.), Wittgenstein. The Philosophical Investigations (Macmillan).
--------, (1978), 'Wittgenstein, Nonsense, And Lewis
Carroll', in Fann (1978), pp.315-35.
Plato (1997a), The Complete Works,edited by John Cooper (Hackett Publishing).
--------, (1997b),
Theaetetus, in Plato (1997a), pp.157-234.
--------, (1997c),
Sophist, in Plato (1997a), pp.235-93.
Platts, M. (1997), Ways Of Meaning
(MIT Press, 2nd
ed.).
Pleasants, N. (1999),
Wittgenstein And The Idea Of A Critical Social Theory (Routledge).
Pleh, C. (no date), 'Was Wittgenstein A
Connectionist, After All?' [This can be downloaded as a .DOC file from
here.]
Plekhanov, G. (1908), Fundamental Problems Of Marxism
(Lawrence & Wishart). [The Appendix to this work -- which in fact
formed part of Plekhanov's Introduction to Engels (1888) -- can be
accessed
here, under the title 'Dialectic and Logic'. It can also be found in
Plekhanov (1976), pp.73-82.]
--------, (1976), Selected Philosophical
Works, Volume Three (Progress Publishers).
Plotkin, H. (2003), The Imagined World
Made Real. Toward A Natural Science Of Culture (Penguin Books).
Potter, M. (2009), Wittgenstein's 'Notes On Logic' (Oxford University
Press).
Potter, M., and Ricketts, T. (2010) (eds.),
The Cambridge Companion To Frege (Cambridge University Press).
Preston, J. (1997)
(ed.), Thought And Language (Cambridge University Press).
Price, H., and Corry, R. (2007) (eds.),
Causation, Physics, And The Constitution Of Reality: Russell's Republic
Revisited (Oxford University Press).
Priest, G. (2002),
Beyond The Limits Of Thought (Oxford University Press, 2nd
ed.).
--------, (2004),
'Wittgenstein's Remarks On Gödel's Theorem', in Kölbel and Weiss, pp.206-25.
--------, (2006), In Contradiction
(Oxford University Press, 2nd
ed.).
--------, (2007), 'Reply To
Slater', in Béziau, Carnielli and Gabbay (2007), pp.467-74.
Putnam, H. (1967), 'The "Innateness Hypothesis" And
Explanatory Models In Linguistics', Synthèse17, pp.12-22;
reprinted in Putnam (1975a), pp.107-16, and Stich (1975), pp.133-44.
--------, (1975a), Mind Language And
Reality. Philosophical Papers Volume Two (Cambridge University Press).
--------, (1975b), 'Language And Philosophy',
in Putnam (1975a), pp.1-32.
--------, (1988), Representation And
Reality (MIT Press).
--------, (2002),
'Travis On Meaning, Thought And The Ways The World Is', Philosophical
Quarterly52, pp.96-106.
Quine, W. (1981),
Theories And Things (Harvard University Press).
Radick, G. (2007), The Simian Tongue. The
Long Debate About Animal Language (University of Chicago Press).
Rappe, S. (1999), 'Self-Knowledge And
Subjectivity In The Enneads', in Gerson (1999), pp.250-74.
Roeper, T., and Spears, M. (2014) (eds.),
Recursion: Complexity In Cognition,
Studies In Theoretical Psycholinguistics (Springer-Verlag). [I haven't yet
been able to check this source.]
--------, (1987d), Wittgenstein And The Turning-Point In The
Philosophy Of Mathematics (State University of New York Press).
--------, (1988), 'The Dawning Of Machine
Intelligence', Philosophica42, pp.93-144.
--------, (1995), 'Turing And The Origins Of
AI. The Mechanisation Of Reason', Philosophica Mathematica3,
pp.52-85.
--------, (1996a) (ed.), Philosophy Of Science, Logic And Mathematics In The
Twentieth Century. Routledge History Of Philosophy Volume Nine (Routledge).
--------, (1996b), 'Descartes Legacy; The Mechanist/Vitalist Debate', in Shanker
(1996a), pp.315-75.
--------, (1997), 'Reassessing The Cognitive
Revolution', in Johnson and Erneling (1997), pp.45-54.
--------, (1998), Wittgenstein's Remarks
On The Foundations Of Artificial Intelligence (Routledge).
Shields, C. (2016), 'Aristotle's
Psychology', Stanford Encyclopedia of Philosophy, edited by Edward N.
Zalta (Winter 2016 Edition).
Shirokov, M., et al
(1937), A Textbook Of Marxist Philosophy (Victor Gollancz). [This book in fact has no publication date,
but internal evidence suggests that it was published in the mid- to late-1930s,
so I have arbitrarily assigned it the given date. The entire book can now be
accessed
here (as a PDF), and parts of it,
here.]
Sidelle, A. (1989), Necessity, Essence, And Individuation. A Defence Of
Conventionalism (Cornell University Press).
Singh, R., Krimbas, C., Paul, D., and Beatty, J. (2001) (eds.),
Thinking About Evolution. Essays In Honour Of Richard Lewontin, Volume Two:
Historical, Philosophical, And Political Perspectives (Cambridge University
Press).
Skinner, B. (1992), Verbal Behavior (Copley
Publishing Group).
Skorupski, J. (1976), Symbol And Theory
(Cambridge University Press).
--------, (1997), 'Why Did Language Matter To
Analytic Philosophy?', in Glock (1997), pp.77-91.
Skryms, B. (2004), The Stag Hunt And The EvolutionOf Social Structure
(Cambridge University Press).
Slater, H. (2007a), The
De-Mathematisation Of Logic (Polimetrica).
--------, (2007b),
'Dialetheias Are Mental Confusions', in Slater (2007a), pp.233-46. [This paper can also
be found in Béziau, Carnielli and Gabbay (2007), pp.457-66.]
--------, (2007c), 'Response
To Priest', in Béziau, Carnelli and Gabbay (2007), pp.475-76.
Sluga, H., and Stern, D. (1996) (eds.),
The Cambridge Companion To Wittgenstein (Cambridge University Press).
Smiley, T. (1995) (ed.),
Philosophical Dialogues (Oxford University Press).
Snell, B. (1982), The Discovery Of The Mind In Greek
Philosophy And Literature (Dover Books).
Soames, S. (1989), 'Semantics And Semantic Competence',
Philosophical Perspectives, 3, pp.575-96. [I haven't yet been able to
check this source.]
Sober, E. (1993a), The Nature Of
Selection. Evolutionary Theory In Philosophical Focus (University of Chicago
Press).
--------, (1993b), The Philosophy Of
Biology (Oxford
University Press).
Sober, E., and Wilson, D. (1998), Unto
Others. The Evolution And Psychology Of Unselfish Behaviour (Harvard
University Press).
Sokolov, A. (1975), Inner Speech And
Thought (Plenum Publishing).
Somerville, J. (1946), Soviet Philosophy (Philosophical Library).
--------, (1967), The Philosophy Of Marxism (Random
House).
Sorabji, R. (1982), 'Myths About Non-Propositional Thought', in
Schofield and Nussbaum (1982), pp.295-314.
--------, (2004) (ed.), The Philosophy Of The Commentators
200-600AD. A Sourcebook, Volume Three: Logic And Metaphysics
(Duckworth).
--------, (2005) (ed.), The Philosophy Of The Commentators. A
Sourcebook 200-600AD, Volume One: Psychology, Ethics And Religion
(Duckworth).
Spanos, N. (1996), Multiple Identities And False Memories: A
Sociocognitive Perspective (American Psychological Association). [I have not
yet been able to check this reference.]
Sperber, D. (1996),
Explaining Culture. A Naturalistic Approach (Blackwell).
--------, (2000), 'An Objection To The Memetic Approach To
Culture', in Aunger (2000), pp.163-73.
Squires, R. (1974), 'Silent Soliloquy', in
Vesey (1974), pp.208-25.
Stainton, R. (1996), Philosophical
Perspectives On Language (Broadview Press).
--------, (2006a), Words And Thoughts. Subsentences, Ellipsis,
And The Philosophy Of Language (Oxford University Press).
--------, (2006b) (ed.), Contemporary Debates In Cognitive
Science (Blackwell).
Stanley, J., and Williamson, T. (2001), 'Knowing How', Journal
of Philosophy98, pp.411-44.
Steinberg, D., and Jakobovits, L. (1971)
(eds.), Semantics (Cambridge University Press).
Sterelny, K. (2003), Thought In A Hostile
World. The Evolution Of Human Cognition (Blackwell).
--------, (2007), Dawkins Versus Gould. Survival Of
The Fittest (Icon Books, 2nd
ed.).
Stern, D. (1995), Wittgenstein On Mind And
Language (Oxford University Press).
--------, (1996), 'The
Availability Of Wittgenstein's Philosophy', in Sluga and Stern (1996),
pp.442-76.
Stich, S. (1975) (ed.), Innate Ideas (University of
California Press).
Stich, S., and Warfield, T. (1994) (eds.), Mental
Representation. A Reader (Blackwell).
Stove, D. (2006),
Darwinian Fairytales
(Encounter Books, 2nd
ed.). [A PDF of the first edition can be accessed
here. (The first edition actually contains fewer typos than the
second!)]
Strawson, G. (1996), The Secret Connexion.
Causation, Realism, And David Hume (Oxford University Press).
Stroud, B. (1990),
'Wittgenstein On Meaning, Understanding And Community', reprinted in Stroud
(2000a), pp.80-94.
--------, (1991), 'The
Background Of Thought', reprinted in Stroud (2000a), pp.131-50.
--------, (1996),
'Mind, Meaning And Practice', in Sluga and Stern (1996), pp.296-319; reprinted
in Stroud (2000a), pp.170-92.
--------, (2000a), Meaning, Understanding
And Practice (Oxford University Press).
--------, (2000b), The Quest For Reality
(Oxford University Press).
Stump, E., and Kretzmann, N. (2001) (eds.), The Cambridge
Companion To Augustine (Cambridge University Press).
Suter, R. (1989), Interpreting
Wittgenstein. A Cloud of Philosophy, A Drop Of Grammar (Temple University
Press).
Szabó, Z. (2017), 'Compositionality',
Stanford Encyclopedia of Philosophy, edited by Edward N. Zalta (Summer 2017 Edition).
Swartz, N. (1985), The Concept Of A
Physical Law (Cambridge University Press). [A PDF of this book
is available
here.]
--------, (2009), 'Laws
Of Nature',
Internet Encyclopedia of Philosophy.
Tait W. (1997) (ed.),
Early Analytic Philosophy. Frege, Russell, Wittgenstein (Open Court).
Taylor, C. (1964), The Explanation Of Behaviour
(Routledge).
Teghrarian, S. (1994) (ed.), Wittgenstein
And Contemporary Philosophy (Thoemmes Press).
Teichmann, R. (2000) (ed.), Logic, Cause
And Action (Cambridge University Press).
--------, (2008), The Philosophy Of Elizabeth Anscombe
(Oxford University Press).
--------, (2015), Wittgenstein On Thought And Will (Routledge).
Teske, R. (2001), 'Augustine's Theory Of
Soul', in Stump and Kretzmann (2001), pp.116-23.
Travis, C. (1989), The Uses Of Sense
(Oxford University Press).
--------, (2000), Unshadowed Thought
(Harvard University Press).
--------, (2004), 'The Silence Of The Senses', Mind 113,
449, pp.57-94.
--------, (2006), Thought's Footing:
Themes In Wittgenstein's Philosophical Investigations (Oxford University
Press).
Trigger, B. (1992), 'Comment on Tobias "Piltdown, The Case
Against Keith"', Current Anthropology33, pp.274-75. [I have not
yet been able to check this source.]
--------, (1973), Problems Of Everyday Life (Monad Press).
--------, (1986), Notebooks 1933-35 (Columbia University
Press).
Ushakova, T. (1994), 'Inner Speech And Second Language
Acquisition: An Experimental-Theoretical Approach', in Lantolf and Appel (1994),
pp.135-56.
Uschanov, T. (2002), 'Ernest Gellner's
Criticisms Of Wittgenstein And Ordinary Language Philosophy', in Kitching and
Pleasants (2002), pp.23-46. [A greatly expanded version of this article is
available
here.]
Uttal, W. (2001), The New Phrenology. The Limits Of Localising
Cognitive Processes In The Brain (MIT Press).
--------, (2004), Dualism: The Original Sin Of Cognitivism (Routledge).
--------, (2008), Distributed Neural Systems: Beyond The New Phrenology (Sloan
Educational Publishing).
--------, (2011), Mind And Brain: A Critical Appraisal Of
Cognitive Neuroscience (MIT Press).
Vernant, J-P. (1990), Myth And Society In
Ancient Greece (Zone Books).
Vesey, G. (1974) (ed.), Understanding
Wittgenstein (Macmillan).
--------, (1994), 'Wittgenstein On
Psychological Verbs', in Teghrarian (1994), pp.89-107.
Vickers, B. (1984a) (ed.), Occult And
Scientific Mentalities In The Renaissance (Cambridge University Press).
--------, (1984b), 'Analogy Versus Identity:
The Rejection Of Occult Symbolism', in Vickers (1984a), pp.95-163.
Voloshinov, V. (1973), Marxism And The
Philosophy Of Language (Harvard University Press). [The first two chapters
can be accessed
here.]
--------, (1987), Freudianism. A Critical
Sketch (Indiana University Press).
Von Wright, G. (1963), The Varieties Of
Goodness (Routledge).
Wiggins, D. (1971), 'On
Sentence-Sense, Word-Sense And Difference Of Word-Sense. Towards A Philosophical
Theory Of Dictionaries', in Steinberg and Jakobovits (1971), pp.14-34.
--------, (2009), 'Knowing How To And Knowing
That', in Glock and Hyman (2009), pp.278-98.
--------, (2012), 'Practical Knowledge:
Knowing How To And Knowing That', Mind121, 481, pp.97-130.
Williams, B. (1990), Descartes. The Project Of Pure Inquiry (Penguin
Books).
Williams, C. (2009), 'Show
Me The Honey', New Scientist203, 2726, 19/09/09, pp.40-41.
[The on-line article has a different title.]
Williams, G. (1966) Adaptation And Natural
Selection (Princeton University Press).
--------, (1996), Plan And Purpose In
Nature (Weidenfeld & Nicolson).
Williams, M. (1999), Wittgenstein, Mind
And Meaning. Toward A Social Conception Of Mind (Routledge). [Chapter Seven
of this book can be downloaded as a PDF from
here.]
Wilson, B. (1998), Wittgenstein's Philosophical
Investigations. A Guide (Edinburgh University Press).
Wittgenstein, L. (1913a), 'On Logic And How
Not to Do It', Cambridge Review34, p.351; reprinted in
Wittgenstein (1993), pp.2-3.
--------, (1913b), 'Notes On Logic', reprinted in Wittgenstein (1979b),
pp.93-107; there is a new, critical edition printed in Potter (2009), pp.276-95.
--------, (1958), The Philosophical
Investigations, translated by G. E. M. Anscombe (Blackwell, 2nd
ed.). [This links to a PDF.]
--------, (1969), The Blue And Brown Books
(Blackwell). [The Blue Book can be accessed
here.]
--------, (1974a), Philosophical Grammar,
edited by Rush Rhees, translated by Anthony Kenny(Blackwell).
--------, (1974b), On Certainty(Blackwell),
edited by G. E. M. Anscombe and G. H. von Wright.
--------, (1975), Philosophical Remarks,
edited by Rush Rhees,translated by Roger White and Raymond Hargreaves (Blackwell).
--------, (1976), Wittgenstein's Lectures
On The Foundation Of Mathematics: Cambridge 1939, edited by Cora Diamond(Harvester
Press).
--------, (1978), Remarks On The
Foundations Of Mathematics, translated by G. E. M. Anscombe, edited by G. H.
von Wright, R. Rhees and G. E. M. Anscombe (Blackwell, 3rd ed.).
--------, (1979a), Wittgenstein's
Lectures: Cambridge, 1932-1935, edited by Alice Ambrose (Blackwell). [Pp. 2-40
can be accessed
here.]
--------, (1979b), Notebooks 1914-1916, edited by G. H. von Wight and G.
E. M. Anscombe (Blackwell, 2nd
ed.)
--------, (1980a), Remarks On The
Philosophy Of Psychology, Volume One, edited by G. E. M. Anscombe and G. H.
von Wright (Blackwell).
--------, (1980b), Remarks On The
Philosophy Of Psychology, Volume Two, edited byG. H. von Wright and
Heikki Nyman, translated by C. G. Luckhardt and M. A. E. Aue (Blackwell).
--------, (1980c),
Remarks On Colour, edited by G. E. M. Anscombe, translated by Linda L.
McAlister and Margarete Schättle
(Blackwell).
--------, (1980d), Wittgenstein's
Lectures: Cambridge, 1930-1932, edited by Desmond Lee (Blackwell).
--------, (1981), Zettel, edited by
G. E. M. Anscombe and G. H. von Wright (Blackwell, 2nd
ed.).
--------, (1982), Last Writings On The
Philosophy Of Psychology, Volume One, edited byG. H. von Wright and
Heikki Nyman, translated by C. G. Luckhardt and M. A. E. Aue (Blackwell).
--------, (1989), Wittgenstein's Lectures
On Philosophical Psychology 1946-7, edited by Peter Geach (University of
Chicago Press).
--------, (1992), Last Writings On The
Philosophy Of Psychology, Volume Two, edited byG. H. von Wright and
Heikki Nyman, translated by C. G. Luckhardt and M. A. E. Aue (Blackwell).
--------, (1993), Philosophical Occasions,
1912-1951, edited by James Klagge and Alfred Nordmann (Hackett Publishing).
--------, (1998), Culture And Value,
edited byG. H. von Wright and Heikki Nyman, revised text by Alois
Pichler, translated by Peter Winch
(Blackwell, 2nd ed.).
--------, (2009),Philosophical
Investigations, translated by G. E. M. Anscombe, Peter Hacker and Joachim
Schulte (Blackwell, 4th
ed.). [This links to a PDF.]
--------, (2013), The Big Typescript: TS 213,
edited and translated by C. G. Luckhardt and M. A. E. Aue (Blackwell).
Wood, O., and Pitcher, G. (1970) (eds.),
Ryle. A Collection Of Critical Essays (Doubleday Anchor).
Woodward, J., and Cowie, F. (2004), 'The Mind Is Not (Just) A
System Of Modules Shaped (Just) By Natural Selection', in Hitchcock (2004),
pp.312-34.
Woolfson, A. (2000), Life Without Genes
(Harper Collins).
Wray, A. (2002) (ed.), The Transition To Language. Studies In
The Evolution Of Language (Oxford University Press).
Wright, C. (1984) (ed.) Frege. Tradition And Influence (Blackwell).
Wright, J., and Potter, P. (2000), Psyche
And Soma. Physicians And Metaphysicians On The Mind-Body Problem From Antiquity
To Enlightenment (Oxford University Press).