Chapter 1
In the first chapter I state the view, and defend its possibility, that sensations
are like ordinary objects inasmuch as the language which we use to talk about
sensations contains terms which designate the sensations in much the same
way as names of ordinary objects designate these objects: that sensations
are identified both by their intrinsic subjective quality and by the public
surroundings in which they occur; that these two kinds of criterion are jointly
necessary for identifying sensations; and that they apply interpersonally,
so that a sensation of a given type must have the same intrinsic quality
no matter whose sensation it is.
A number of arguments against the possibility of this view are considered.
They all impinge on the well known ‘argument from analogy’, whose intelligibility
and validity are essential to the view I defend. The first objection to the
argument from analogy which I consider attacks its validity: it, is claimed
that it is a weak argument. I reply that it is stronger than is sometimes
thought, and that even if it still falls short of some ideal standard, that
is no reason to reject it. Its acceptance is deeply embedded in our way of
thinking of other people, and we would need very persuasive arguments to
show that it must be be rejected.
Next I consider arguments against the premise of the argument from analogy
– the premise that I have subjectively similar experiences in objectively
similar circumstances. It is claimed either that this premise, depending
as it does on the veracity of memory, is too insecurely supported to be relied
on; or that it is actually unintelligible if thought of in terms of the view
of sensation language I adopt. Against the first kind of argument I maintain
that it is not unreasonable or irresponsible to rely on the accuracy of memory;
against the second, that the verificationist conditions of meaningfulness
on which it is based are irrelevantly severe. This involves me in a defence
of the view that the conditions of meaningfulness in question are indeed
verificationist; and in demonstrating that Wittgenstein, who put forward
an argument of this second kind, was a straightforward verificationist sceptic
about memory, rather than something more subtle.
Finally, having thus defended the intelligibility and truth of the premise
of the argument from analogy, I defend the intelligibility of its conclusion
– that different people have subjectively comparable experiences in comparable
circumstances. I do this by discussing various conceivable ways in which
the subjective quality of the experience of two people could be directly
compared.
Given that the conclusion of the argument from analogy is intelligible, and
that my defence of the validity of the argument is successful, it now follows
from the truth of its premise that the conclusion too is true. Hence, the
view I adopt at the beginning of the chapter is possible, at least. If it
is shown to be false, it will not be on a priori grounds.
Chapter 2
In chapter 2 I show how an alternative analysis of sensation language, which
appears to be immune to some of the criticisms levelled against the analysis
I adopt in chapter 1, and also avoids the behaviourist extreme, is not as
real a possibility as it at first seems. Its initial plausibility depends
on neglect of its more far-reaching implications.
According to this theory (‘C-subtle’) sensations are defined exclusively
in terms of their public surroundings, not by their intrinsic quality – though
this must be constant for any given person (except on an extreme interpretation
of the theory, which I quickly reject).
I give three main arguments against this theory. The first argument begins
by pointing out that, according to the theory, it should be possible for
someone to know the meaning of any sensation term without ever having experienced
the sensation in question, so that he might fail to recognise the sensation
when he first felt it, despite his knowledge of the meaning of its name.
I argue that this paradoxical situation is unacceptable for more general
sensation terms such as ‘pain’, even if it is tolerable for more specific
ones such as ‘headache’. This distinction is the occasion for some general
remarks about how much experience is required for knowing the meanings of
words.
The second argument, which arises out of the first, is that C-subtle’s requirement
of intrasubjective constancy of sensation quality for a given type of sensation,
together with some conclusions of Strawson’s, entails precisely the requirement
of intersubjective similarity which a-subtle wishes to eschew.
The third argument is that the consequences of applying C-subtle, not just
to individual sensation terms, but to the term ‘sensation’ itself, are unacceptable.
There are two main consequences in question. One is that we could not even
require that our sensations are in general like those of other people. And
the other consequence, which would be a sufficient objection to a-subtle
even if the first consequence did not matter, is that the term ‘sensation’
cannot be made intelligible at all on C-subtle’s view of its meaning, since
there is no more general public concept, already mastered, under which ‘sensation’
could be subsumed and in terms of which the meaning of ‘sensation’ could
be understood. We could not learn what sort of thing a sensation was unless,
contrary to C-subtle, we allowed ourselves to be instructed by our own personal
experience.
Chapter 3
The last chapter consists of a preliminary discussion of the question of
the actual – no longer merely the possible – meanings of sensation terms.
There is a range of possible views as to what sensation terms mean, from
the ‘intrinsicist’ extreme of holding that they are defined purely in terms
of inner quality, to the ‘extrinsicist’ extreme of holding that they are
defined exclusively in terms of their public surroundings. Between these
extremes lies the middle possibility that both intrinsic and extrinsic criteria
are involved.
These are the broad options. Finer discrimination is possible, and I try
next to sketch a framework within which a more detailed typology of sensation
terms could be established. I subdivide the categories both of extrinsic
and of intrinsic criteria, and suggest some questions that arise out of the
resulting classificatory scheme: questions as to what general truths about
the meanings of sensation terms might emerge, and questions about the explanation
of these truths, once discovered.
Next I outline, and demonstrate in action, two possible methods for investigating
the meaning of a sensation term, both of which can help to reveal not only
that element in the meaning of a term which is closely related to its functional
role in our lives, but also those elements in its meaning which arise from
the associative power of what, from a functional viewpoint, are inessentials.
The first method is to inspect the ‘teaching links’ of a sensation – the
public phenomena by reference to which the use of the term is taught. I discuss
what the various teaching links of sensation terms are, and what their twofold
connection with meaning is. I also mention two ways in which they are unreliable
guides to meaning: in particular, they can tell us nothing about intrinsic
criteria. [The other thing is that links are not necessarily implicated in
meaning.]
This is where the second method comes in. It is to ask what we would be disposed
to say if nature were to change in certain ways. I exemplify this method
in connection with ‘pain’: changes of intrinsic quality and of extrinsic
surroundings are imagined, in order to try and elicit the criteria which
govern the use of the term. Heavy reliance is placed, in considering what
we would say if such changes occurred, upon the deliverances of intuition:
this, I argue, is the only method available in the last resort for settling
the issues in question, though support for its rulings can be supplied, up
to a point, by examining the function that a sensation concept performs in
our lives – and this is one of the things that the two methodological tools
described above are designed to reveal.
Out of this discussion arises a general hypothesis, that all our sensation
concepts, being tailored to the world as we know it, apply only when both
intrinsic and extrinsic criteria are satisfied, and so only when private
sensation quality and public surroundings march in step. Were these two kinds
of criterion to part company, our concepts would need either to fall into
disuse, or to be given new definitions[/uses] that fitted the novel circumstances
– in which case they would not really be the same concepts at all. I show
this hypothesis in action by outlining a couple of applications of the framework
of enquiry drawn up earlier in the chapter.
|