‘Analytic philosophy’ has
been many things, but much of it can be — simplistically — characterized as
an attempt to do the philosopher’s work by quite literally analyzing the
meaningsof sentences and words into their constituent parts. These
constituent parts were said to be of two types only: 1) the concepts
corresponding to the symbols of symbolic logic, which we’ll not further remark
upon, and 2) the names of objects of direct experience, such as ‘red’ and
‘loud’. (Note that a successful analysis of say, ‘substance’ or ‘action’, into
nothing but names of objects of direct experience, such as ‘red’ and ‘loud’,
would be a considerable, and almost certainly, enlightening achievement.) The
objects of direct experience, which were called ‘sense-data’, were taken to be ‘given’;
they were just what they appeared to be and nothing else (thus, infallible
data) and they owed nothing to our conceptualization; all we could do is to
recognize and name them. Analytic philosophy, being strongly empiricist,
assumed we could ultimately know nothing but the sense-data given to our
senses, and certain truths about meanings and logic provided by our
conceptualization of the sense-data given to us; therefore, all meaningful
discourse could be analyzed into nothing but the names of sense-data,
and the logical particles that joined these sense-data into our concepts of
things and/or the meanings of the words that named those things, and the
latter, into propositions or sentences. Any meaningful sentence about anything
could be analyzed, in principle, into a sentence about sense-data. In time,
analytic philosophy rejected sense-data, sometimes even formal logic, but it
retained the basic assumption that the job of the philosopher is 1) to analyze
meanings into their constituents, and 2) to tie those meanings to empirical
criteria for their use.

Working on that assumption, Russell
showed that there need not be a king of France with some kind of shadowy being,
since we can — it seems — meaningfully talk about ‘him’; rather, for example,
‘The king of France is bald’ really means that there is one and only one thing
that is the King of France and it is bald. 
(Kings, France, and baldness do unproblematically exist, although a
complete analysis of them into sense-data would be difficult; the other
constituents of the proposition that the king of France exists are particles we
contribute.) Notoriously, Carnap showed by means of the analysis of language
that Heidegger’s "The nothing nothings." is meaningless.

And Ryle showed us how to exorcise
‘the ghost in the machine’, the ‘mind’, bequeathed to us by Descartes, by analyzing
talk about mind into talk about behavior; for example, to ascribe, say,
anger to someone is to say he is likely to yell, insult, avoid, etc. This view,
which may be called ‘analytic behaviorism’, seems to have two great advantages:
it explains how we know about other minds — by watching how people behave —
and it enables us to keep a wholly materialistic outlook by reducing the mind
to nothing but the behavior of the body. (Of course, the latter advantage is an
advantage only to materialists, but that includes almost all analytic
philosophers.) But there is a seemingly powerful objection to analytic
behaviorism! We are directly aware of certain things — they are givens —
that just cannot be reduced to behavior: 
our sense-data. (For various reasons, we now speak of ‘qualia’
rather than of sense-data.) . To see the problem, consider the sense-datum you
have when you see something red. You do not have to have just that sense-datum
whenever you see ripe tomatoes and fire engines in order to engage in such
overt behavior as saying that you saw something red or stopping for a red
light. You could have green sense data whenever you look at an object that
gives me red sense data. Indeed, we can imagine someone who behaves exactly
like we do when we see something red without having any sense data at all.
Sense data cannot be analyzed into behavior!

The materialist might respond that
while sense-data are not patterns of behavior, they are, nevertheless, material
things, namely brain events. Analytic philosophers have turned to the meanings
of the words ‘sense-datum’ and ‘brain event’ to show that they cannot be names
of the same kinds of things. It is true, in virtue of meaning, that a
sense-datum can be observed by only the one person who has it, and that
sense-data do not have spatial locations; brain events, on the other hand, can
be observed by many and do have spatial locations. Therefore, brain events
cannot be sense-data!

Since sense data cannot be analyzed
into behavior or identified with brain events, it seems that they cannot be
physical things subject to ordinary scientific study. This was a major problem
for analytic philosophy because it generally considered itself to aligned or
even continuous  with science. (As you
can imagine, these problems spawned a vast literature.)

Note well that this problem for
materialism is a consequence of the two basic assumptions of analytic
philosophy, the possibility of meaning analysis and the given. Sense data are given
and the very meaning of ‘material’ entails access to more than one
perceiver, while the very meaning of ‘sense datum’ excludes just that. Richard
Rorty, early in his career, offered a way to remove the problem by rejecting
the two basic assumptions of analytic philosophy, thereby clearing the way for
materialism while presenting a major challenge to analytic philosophy, one
which he has developed further through the years.

Rorty’s proposal was that since
nothing is simply given to experience without being taken into
some linguistically determined category, and since word meanings are at most
changeable conventions, we can simply deny that sense-data really exist, and
are necessarily private and without spatial location, and speak only of brain
events, which are, of course, perfectly material. What we used to call ‘pain’
is really nothing but the firing of various neurons. So speaking saves us much
trouble and we see that there is no reason not to when we rid ourselves of
notions of the given and of meanings as sources of necessary truth.

The first volume of Alan
Malachowski’s extensive four volume collection of papers on Rorty includes some
of the early papers reacting to this then novel approach to the ‘mind-body
problem’. At first, the reaction was, I think, uncomprehending, seeing Rorty’s
proposal within the assumptions of traditional analytic philosophy. It was
complained that we just are immediately aware of sense-data and that
they necessarily are either non-material or a major problem for materialism
demanding solution, that if we stopped talking about them, our description of
the world would be incomplete. In time, the wider implications of Rorty’s view
became apparent and discussion turned towards them and away from the
particularities of the traditional mind-body problem. Papers in this first
volume also record some of this development. Reactions to the later unfolding
of Rorty’s approach are mainly dealt with in later volumes.

© 2003 Alan N. Sussman

 Link: See the review
by Richard Matthews of Volume III of this set of books.

Alan N. Sussman received
his Ph.D. in philosophy from the University of Chicago. He has published a few
papers, including one in The Journal of Philosophy. He taught philosophy at
various colleges and universities in the US and Africa. At present he teaches
part time at Truman College, Chicago. His philosophical interests are primarily
in philosophy of mind.

Categories: Philosophical