Every so often one encounters a book with which one disagrees, wholly or in large part, but which one regards as a genuine contribution to philosophy precisely because it sets out views with which one disagrees, and does so with exemplary clarity and sophistication. For me, Galen Strawson’s Mental Reality is such a book, and any contemporary course of lectures on the philosophy of mind would be well advised to discuss it: the issues it deals with are important ones, and what Strawson has to say about them is original.
The philosophy of mind is a huge field, and Strawson has something to say about almost every one of the issues. I shall focus on three that are central. The first is the metaphysically remarkable character that mental states and occurrences have, according to Strawson. Although he calls himself a materialist, he believes that experiential properties (which, in his view, are the only distinctively mental properties there are) are totally inexplicable by physics as we know it. It will, he claims, take a revolution in fundamental physical science – a revolution in our most basic understanding of what matter is and is capable of – to account for them. Moreover, we cannot foresee the specific character that that revolution must take: on the contrary, one of Strawson’s theses is that we cannot at present understand how such a revolution is even possible. (I do not myself think it makes sense to say that we need an X unless we can to some extent conceive of what that X is, or at least say how it might help with our predicament, but I shall not press this objection here.) Given this thesis of the radical exceptionalism of the mental (relative to what we now think of as physical science), one might wonder why Strawson calls himself a materialist at all. He is quite forthright in admitting that his ‘agnostic materialism’ (as he calls it) has somewhat the character of an act of faith: ‘Why did I call myself a materialist rather than a “?-ist”? My faith, like that of many other materialists, consists of a bundle of connected and unverifiable beliefs ... I believe that the theory of evolution is true, that once there was no experience like ours on this planet ... and that there came to be experience like ours as a result of processes that at no point involved anything not wholly physical or material in nature.’
Strawson’s argument builds on arguments made famous by Thomas Nagel. According to Nagel, no amount of purely physical information about what goes on in a creature’s brain (he used a bat as his example) can tell us ‘what it is like’ to be that creature. In this sense, objective physical science radically omits an aspect of the world of fundamental importance and interest to us: this ‘subjective’ or ‘what-it’s-like’ aspect.
Strawson exploits Nagel’s discussion in two ways. First, he concludes from Nagel’s observation that this ‘what-it’s-like-ness’ (which he identifies with the experience itself) is necessarily different from all the properties of the organism that we (now) think of as physical. (Here he follows Nagel himself.) But he insists that we should not rest content with this explanatory gap’, as he sees it. Physics itself needs to change so that the explanatory gap can be filled.
Strawson raises the obvious difficulty: ‘how can the nature and existence of an experiential property be thought to be (fully) explained by reference to essentially non-experiential properties?’ This is all the more pressing for him, inasmuch as his programme requires the successor to present-day physics that he calls for to make it clear why the experience of hearing a particular note struck on the piano has the particular ‘qualitative character’ it does, and he will not allow any amount of information about the structure of the relations between experiences to count as a description of ‘what it is like’ to have them. ‘We are obliged to admit,’ he writes, ‘that we do not know how experience – experiential what-it’s-like-ness – is or even could be realised in the brain. This is why any serious statement of materialism needs to acknowledge that our current physical-science conception of the physical is radically incomplete.’
There are many objections one might make to this argument. I myself have argued that perceptual experience, at least, should not be thought of as experience of something ‘internal’ (in the brain or mind), but as a taking in of how it is ‘out there’. Strawson would presumably reply that the same problem of the metaphysically remarkable character of ‘what-it’s-like-nesses’ would still arise even if this view were correct, which he doesn’t agree that it is. It would arise even if (as in William James’s view, which Strawson surprisingly misunderstands) ‘what-it’s-like-nesses’ are ‘out there’ as well as ‘in here’, because the problem concerns the remark able nature of ‘what-it’s-like-ness’ itself, and not its location. Again, one might argue that whatever the nature of the problem of ‘explaining what-it’s-like-ness’ may be (if it is a genuine problem at all), it cannot be a problem for physics because physics concerns itself only with factors needed to predict the trajectories of particles and fields, and Strawson has given no reason to think that the additional ideas which he believes must be added to physics would improve its ability to do this job. But what I want to question is whether a materialist need grant that there is an explanatory problem here at all.
Part of the difficulty concerns Nagel’s original argument, and part concerns Strawson’s extension of it. Assume we do think of experiences as somehow ‘realised in the brain’ (as Strawson does), and that we have a description of a neurological property of the brain which is ‘correlated’ with the experience of hearing, say, high C struck on the piano – call the property N. Let E be the experiential property, the ‘what’s-it’s-like-ness’. My knowledge of E is non-discursive knowledge; it may consist in hearing the piano note right now, or of being able to imagine it, or of recalling it ‘vividly’. My knowledge of N is discursive knowledge. It consists in being able to theorise about N in a language which presupposes only such experiences as are necessary to do science (which is to say, presupposes the ability to read instruments, but not any particular sensory modality). Of course, these are different kinds of knowledge. But how can Nagel and Strawson validly infer from the fact that E and N are known so differently – E non-discursively and N discursively – that they cannot be identical? One popular reply used to be that identities between properties cannot be synthetic: if it were a truth that E and N are identical, it would have to be a conceptual truth; but this reply is now widely discredited. It seems that Nagel and Strawson move illegitimately from the existence of two modes of knowing to the existence of two different things to be known. (I say this not to defend the ‘identity theory’ of the experiential and the physical – I don’t think that we have given a sense to ‘identity’ or ‘non-identity’ in this context – but to argue that this particular argument should not move a classic materialist to concede that physics needs to be changed in any way.)
The second part of the problem with Strawson’s argument is that if Nagel’s arguments were right, it would show that ‘what-it’s-like-ness’ cannot be captured discursively – not unless the particular text which does the capturing already presupposes acquaintance with the what-it’s-like-ness in question. But physics is by definition a form of interpersonal discursive knowledge, one which can be grasped by anyone capable of performing experiments and calculations. It cannot require a complete palette of sensory experiences (nor for that matter could any one human being have a complete palette of sensory experiences) and still resemble physics. The whole idea of a discursive bridging of the gap between the discursive and the non-discursive, which seems to be what Strawson is calling for, looks like a confusion.
The second central issue raised by the book – indeed, the one on which Strawson lays the most stress – is the radical privacy of the mental. In sharp opposition to all the leading currents in contemporary analytic philosophy of mind, functionalist, neo-Wittgensteinian, Davidsonian etc, Strawson insists that we have private knowledge of experiential qualities (including both sensory qualities, such as the experience of a particular shade of red, and more ‘cognitive’ qualities, such as the experience of recognising that I understand a particular sentence), and that the identity and existence of these qualities are independent of all behaviour (and of all dispositions to behaviour) whatsoever.
In response, I wish to note two problems. First, one may reject the idea (which has almost disappeared from the American analytic philosophical scene) that there are logical truths connecting experiential concepts with actual and possible behaviour without concluding that our grasp of an experiential concept is independent of the role such a concept plays in ‘the stream of life’ (as Wittgenstein put it). I approve of many of Strawson’s arguments against strong forms of logical behaviourism/logical functionalism – indeed, I originated some of them! But logical behaviourism/functionalism and private knowledge simply do not exhaust the possibilities. Secondly, it is even doubtful that Strawson’s account of intentionality (which makes reference a matter of causal connections) is compatible with his conception of reference to experiences (which seems to involve just the sort of private dubbing that Wittgenstein criticised).
To bring out the incompatibility, let us suppose (Strawson regards this supposition as perfectly coherent) that my concept of, say, high C actually does not pick out ‘objectively similar experiential properties at different times’, although it seems to me that it does (my memories, reactions etc, all conform to the hypothesis that the experience of high C is at all times the same). Now the sameness of my reactions, the fact that the same memories are produced etc, may be caused, let us suppose, by the fact that these events all have some property in common; call the property they all have (the one that causes them to be incorrectly classified by me as having ‘objectively similar experiential properties’) Q. Since Q causes all these events to be classed as objectively similar experiences, and reference is a matter of causal connection in Strawson’s account, does it not follow that Q is the property we refer to as ‘objective experiential similarity’? And if that is what ‘objective experiential similarity’ refers to then it is objective experiential similarity. So the supposed coherence of the hypothesis that we in-detectably have qualitatively different experiences at different times cannot be maintained, and the idea of private dubbing collapses.
The third central issue raised by the book is the claim that there is no general problem about intentionality for materialists; the observation that reference is a matter of causal connections is supposed to take care of everything. I can only remark that the huge disagreement in the philosophy of mind/cognitive science community about whether any of the so far proposed attempts to reduce reference to causation (without smuggling in some intentional notion or other) strongly suggests that this ‘no problem thesis’ is simply untenable.
Send Letters To:
London Review of Books,
28 Little Russell Street
London, WC1A 2HN
Please include name, address, and a telephone number.