A Technical Philosopher

Hilary Putnam

  • The Varieties of Reference by Gareth Evans, edited by John McDowell
    Oxford, 418 pp, £15.00, October 1982, ISBN 0 19 824685 4

Gareth Evans died of cancer when he was barely 34 years of age. He had been working on this book for several years; the task of completing it from his notes was carried out by John McDowell. (The first two chapters and the introduction were rewritten by Evans himself in the last months of his life.) Evans’s death at such an early age is a tragedy. We can have no real idea what his mature years might have brought forth, and this book is no substitute for those additional years of life and thought: it is the desperate attempt of a young man on his deathbed to leave a lasting mark on his profession. I do not like – in fact I hate – to say that he did not (in my opinion) succeed. And I apologise to those who will feel personally hurt at a negative review of a work produced by a vital person whom they knew and loved and miss today.

For reasons I shall discuss shortly, it is not possible to convey Evans’s main claims, conclusions and arguments in a review for a general audience, nor even to give the flavour of his mode of reasoning. Instead I shall confine myself to describing an example which he relies upon to illustrate and help establish certain main sub-points in his argument.

A man sees one of two whirling steel balls at time A. At time B he sees the other (imagine they cannot be distinguished by him). He has a memory of the second seeing, but not of the first (Evans postulates a bit of amnesia, but ordinary forgetting would do as well). Also, the man (I will assume the man is myself, for simplicity of discussion) does not know there are two different balls.

Can the man – that is, can I – think about the ball I remember? We recall that I have what seems to me to be a perfectly good memory of one (just one) steel ball whirling around. It is in fact a perfectly good memory (if we could trace the origin of the memory trace in my brain, it would go back to an event involving just one of the two original balls). I can believe that I am thinking: ‘that ball was whirling around terrifically fast.’ But, according to Evans – and this is a vitally important conclusion for this whole book – I am deceived. What is really going on is that I appear to myself to have thought a thought: but there was no thought of the kind I appeared to myself to have thought (no thought about a particular ball) to be thought in my situation.

In addition, in Evans’s view, a thought about a particular object which I perceive (or seem to myself to be perceiving) and which I identify by the use of an ordinary demonstrative (‘that ball’, ‘that eagle’ etc) presupposes the non-illusory existence of that object for its very existence: if the relevant ball, eagle or whatever is a mirage or an illusion or a hallucination, then the ‘thought’ is also an illusion. Russell believed (for very different reasons from the ones Evans gives) that there were propositions whose existence presupposed the existence of the objects they were about (for Russell these were all propositions about sense-data). For this reason, Evans calls a thought with this property – the property that it could not exist if the object it is about did not exist – a ‘Russellian thought’. To establish the existence of Russellian thoughts, Russellian Ideas (Idea, with a capital ‘I’, is Evans’s term for the concept of an individual thing – e.g. the concept which corresponds to the words ‘that eagle’ when I successfully think, ‘that eagle is flying in circles’), Russellian referring expressions, etc, is the avowed object of this book.

One thing a review cannot convey is the relentless technicality of the book. It could not be used in a graduate course in philosophy of language without several preliminary lectures on Davidson, on Kripke, etc (not to say a preliminary course in math logic). It is a book addressed to Evans’s fellow specialists, and only to them. Philosophy, as Evans pictures it, is as esoteric as quantum mechanics.

The model of philosophy as a Wissenschaft, if not actually a science, informs the treatment of Evans’s great predecessors as well. That Russell’s ‘propositions’ were not ‘thoughts’ (they were extra-mental objects which literally had ordinary objects as constituents) is an essential fact about Russell’s metaphysical system which Evans feels free to ignore; that Frege did assign referents to what Evans calls ‘empty’ descriptions in his ideal language, and that he regarded none of the more-or-less vague expressions we use in everyday language as having sense at all (in spite of what Evans calls the ‘viability’ of everyday language), are also not mentioned. Particular technical problems are simply ripped out of their place in the rich filigree of the metaphysical system as though they had a significance independent of their historical context.

In spite of this relentless technicality, it gradually becomes clear that Evans does have a philosophical picture of his own, and not just a set of technical claims about technical issues. The picture is roughly as follows: thoughts are not sentences, nor are they symbols in the mind or brain (they are not ‘sentence-analogues’ in a language of thought, as some Chomskyans think). Thoughts are exercises of structured systems of abilities. The thought that object a has property F involves an ability which is the Idea of a and an ability which is the idea of F.

This claim only has content if we can describe the Idea of, say, ‘that eagle’ otherwise than as ‘whatever capacity one has to have to make such judgments as “that eagle is flying in circles,” “that eagle is old,” etc. Evans’s tack, as I read him, is not to claim that we are in a position to do this completely, but to indicate at least in part what such a non-trivial characterisation of the Idea in question might look like.

According to Evans, I can only think about eagles (or any other sort of things) if 1. I possess a ‘Fundamental Idea’ of what it is to be an eagle (is a Fundamental Idea an analytically necessary and sufficient condition, or what? Evans does not say), and 2. I know what makes an object of the sort different from another object of the sort (in the case of eagles or other material objects this would be spatial location). When I pick out an object as ‘that eagle’ in a perceptual situation, my visual transaction with the object puts me in an ‘informational state’ which immediately gives rise to at least two abilities: an ability to roughly locate the object relative to my body (to locate the object in ‘egocentric space’, in Evans’s terminology); and an ability to locate the object in public space (‘objective space’, in Evans’s terminology).

Now we can say what is wrong when the object is not really there. If I am fooled by a mirage and I say, ‘That oasis is a good place to water my camels,’ then I think I have acquired an ability (the ability to locate the oasis in objective space) which I have not in fact acquired. This is what Evans’s notion of thinking I have a thought when there is no such thought to be thought cashes out to in such a case.

The problem with the whirling balls is similarly treated. If I describe the ball I remember as ‘whichever ball caused this memory’, then, in Evans’s view, I am simply passing the buck. I am not showing that I have an Idea of the ball (an ability to locate it in objective space, or to decide of objects which I can locate in objective space whether or not they are ‘that ball’), but simply indicating conceptually how I might get such an Idea. The fact that I cannot either locate ‘that ball’ in objective space nor decide which ball was ‘that ball’ (imagine I have since learned that there were two) shows that ‘that ball’ no more refers to a determinate real thing than ‘that oasis’ does when there is a mirage.

I find none of this convincing at all. (For example, if I don’t recall, even briefly, in which direction I was looking when I thought, ‘that eagle flew by very fast,’ then I did not acquire an ability to locate the eagle in either egocentric space or objective space: so it would follow I didn’t really think a thought on Evans’s view.) But this is not the place for a rebuttal of any of Evans’s claims. What bothers me is the conception of philosophy behind all of this. Evans appears to assume that philosophy is a technical discipline in which problems have a significance independent of the actual system of thought that generated them; and that one can make uncritical use of the notion of a ‘conceptual’ truth. These two ideas undergird each other: what gives substance to the notion of philosophy as a technical discipline for Evans is his amazing confidence that he is discovering indisputable conceptual truths and not just talking in a way he finds compelling.

If the picture that lies behind all of this – the picture of language as hooking onto the world through the medium of specifiable object-involving dispositions and capacities – could be shown to have some force, then one could, perhaps, throw away the technicalities and discuss that picture. But the impossibility of really specifying even one of these ‘capacities’ blocks that path (or so it seems to me, as it did to Wittgenstein and Quine before me). What specifiable ‘capacity’ must anyone have to have an ‘adequate Idea’ of an eagle, of an elm tree, of water? What reason is there to think that such a capacity could be specified in finitely many words in a non-question-begging way? Or, if it cannot be specified, why is this capacity-reductionism any better than the now abandoned sense-data-reductionism of Ayer and Russell? Evans does not discuss these questions at all.

It has been said that there are two sorts of analytic philosophers today: those who worry about whether a bundle of sticks would still be the same bundle if one stick were removed, and the therapists who try to cure the first sort of that kind of worry. With respect to Evans’s worries about the existence of ‘Russellian thoughts’, my stance is that of the therapist. But the larger question is what philosophy might be if it gave up the pretence of having a special authority of the kind a real technical discipline possesses. When Evans describes the phenomenology of egocentric space and objective space in one of his chapters (Six), he begins to move in the direction of such a philosophy: I wish he had done more of this and less over-elaborate ‘theory’.