In theory, it is the highest virtue of the philosopher to be constantly receptive to criticism, always willing to abandon his own views upon hearing a better argument. In practice, students tend to become exasperated when an important philosopher changes his mind. It suits their doxographic purposes best to use the philosopher’s name to denote the monolithic set of doctrines which initially made him famous. Bertrand Russell is an example of an important and influential philosopher who changed his mind several times and thereby induced exasperation. Hilary Putnam is another. Just when people have finished writing a devastating critique of Putnam, they discover that Putnam has written a similar critique of his own previous views. This refusal to serve as an unmoving target has sometimes led to attempts to dismiss him as a reed shaken by every new wind of doctrine. But such attempts fail, for Putnam is one of the most vigorous and thoughtful representatives of the second generation of analytic philosophers.
This generation (which also includes Quine, Goodman, Sellars and Davidson) separated itself from the generation of Russell and Carnap by developing doubts about the atomism which had linked these earlier figures to the British Empiricists. Russell and Carnap had shared a vision of language as tied down to the world at lots of discrete points, and of philosophy as offering an account of the relations between these points by exhibiting ‘the logical syntax of language’. This was reminiscent of Hume’s vision of human nature as something which philosophy could split up into lots of psychological atoms (‘impressions and ideas’), and then describe in terms of the relations between these atoms. Carnap’s and Russell’s disciples, however, realised that there are as many ways to tie language down to the world as there are languages. Goodman’s example of ‘grue’, and Quine’s of the various ways to translate ‘gavagai’, suggested that there were as many ‘logical syntaxes of language’ as there were vocabularies. Contemplation of such examples led to holistic theories of meaning, reference and justification and thus to the view that all these are ‘relative to a choice of language’ or to ‘a background theory’ or to ‘linguistic practices’ – suggestions which chime with the older Wittgenstein’s repudiation of the atomism of the Tractatus and his mockery of the idea that ‘logic is something sublime.’
In the early Sixties, Putnam identified himself with this movement towards holism and relativism, but in the late Sixties he began to have doubts about it. He then briefly affiliated himself with a third-generation backlash in favour of the atomism of the grandfathers – the anti-Quine, anti-Wittgenstein movement spearheaded by Kripke. This backlash protested the deconstruction of the model of philosophy common to Hume and Russell – philosophy as piecemeal analysis. It tried to reformulate the ‘traditional problems of philosophy’ which the second generation had been inclined to dissolve, and in particular to breathe life into the notion of ‘reference’ which Quine thought indeterminate and Davidson dispensable. In Volume Two of his Philosophical Papers, published in 1975, Putnam joined Kripke in suggesting the desirability of a new, non-intentional, ‘realistic’ and physicalistic theory of reference – a way of viewing the ties between language and world which would be immune to Quinean and Davidsonian dissolutions.
In 1976, however, Putnam formally recanted his alliance with the rising generation, causing consternation among his followers and inspiring articles with such titles as ‘Realism and the Renegade Putnam’. In the Introduction to this latest volume of his papers he explains the reasons for his shift, and in the essays that follow he tries to steer a middle course between the ‘metaphysical realism’ typical of the third generation and the ‘relativism’ which he thinks follows from going too far in the direction pointed by his own, second generation. The volume as a whole shows Putnam at the top of his form – moving easily back and forth between brilliantly original arguments on small points in the philosophy of language, and equally original diagnoses of large-scale cultural trends. This book is analytic philosophy at its best. It makes one realise that Putnam is, among contemporary analytic philosophers, the one who most resembles Russell: not just in intellectual curiosity and willingness to change his mind, but in the breadth of his interests and in the extent of his social and moral concerns.
On the other hand, some of the early essays in this volume show that Putnam also shares one of Russell’s minor vices: an attachment to the idea that ‘logic is the essence of philosophy’ and thus to the hope that results within logic are going to have decisive philosophical importance. He makes much play with the Loewenheim-Skolem theorem – ‘a satisfiable first-order theory (in a countable language) has a countable model’ – as if this somehow underwrote his pro-Wittgenstein and anti-Kripke claim that ‘either the use [of a language] already fixes the “interpretation” ’ – i.e. decides what objects in the world are being referred to – ‘or nothing can.’ But nobody is going to be convinced of that sort of claim by being referred to a theorem in mathematical logic. One becomes convinced of it by seeing how well attempts to give sense to notions like ‘interpretation’ and ‘reference’, attempts which try to get beyond ‘use’, actually work.
Fortunately, Putnam often drops the appeal to logic and gives us just such an exhibition of the difficulties with ‘realist’ views of reference. The difficulty to which he recurs most often is that if use is not enough, if there is some other perspective from which to decide how and where language ties onto the world, then we shall need an Archimedean point – a ‘God’s-eye view’. But the vocabulary which says how things are from that point of view will itself be subject to the same questions about reference as the vocabulary about which such questions were originally raised. The realist philosopher who advocates a causal theory of reference will, as Putnam says, ‘ignore his own epistemological position’. When the realist protests that we can only avoid pragmatism and relativism by treating reference as a non-intentional relation, a matter of non-psychological, physical fact, Putnam replies that
the idea that ‘the non-psychological’ fixes reference – i.e. that nature itself determines what our words stand for – is totally unintelligible. At bottom, to think that a sign-relation is built into nature is to revert to the idea that there are ‘self-identifying objects’ and ‘species’ there ... Such an idea made sense in the context of a Medieval world view ... In the context of a 20th-century world view, by contrast, to say in one’s most intimidating tone of voice ‘I believe that causal connections determine what our words correspond to’ is only to say that one believes in a one-knows-not-what which solves our problem one-knows-not-how.
One might be tempted at this point to ask: What problem? Who has got a problem about reference? Putnam sometimes seems to be saying, dismissively and therapeutically, that there really isn’t any problem: nothing is more certain than that ‘cat’ refers to cats, and that sort of banality is all we know about reference and all we need to know. But most of the time he insists that we really do need a semantics, and that ‘ “realist semantics” v. “non-realist semantics” ’ is a live and momentous issue. This insistence is a product of his fear that if we trivialise reference (the tie between the better sort of words and the world) then we shall also trivialise truth (the tie between the better sort of sentences and the world).
For many contemporary philosophers (e.g. Heideggerians and Davidsonians) this sort of trivialisation – taking Tarski’s so-called ‘disquotational view’ as all the theory of truth we need – looks like a good idea, because the notion of truth as the junction of two ontological realms (words and the world, the world inside us and the world outside us, subject and object) looks like a bad one. Putnam agrees that such splits (e.g. that between the ‘factual’ and the ‘conventional’ elements in knowledge) are a bad thing, but he nevertheless thinks that ‘as thinkers we are committed to there being some kind of truth, some kind of correctness which is substantial and not merely “disquotational”.’ He fears a domino effect: he thinks that if we don’t have something like a theory of truth we shan’t be able to have a ‘theory of trans-cultural rationality’ and that this inability will lead to ‘relativism’. The last essays in this volume are polemics against relativism, and some of the ambiguities of the earlier essays are clarified if one rereads them in the light of these polemics.
Putnam thinks of relativism and historicism as ‘attempts to reduce epistemic notions to non-epistemic ones: syntactical ones, in the case of positivism; anthropological ones (e.g. “structuralist” ideas) in the case of historicism or relativism’. He is certainly right that a lot of what is usually classified under these two headings springs from such a reductionist line of thought. Since Hegel, we have had a series of philosophical movements which tried to historicise epistemology, seeing in history the same sort of sovereign arbiter that earlier philosophers found in God or Nature. Putnam rightly points out that lots of little anthropological or historical structures are no better than one big ‘logical’ structure when it comes to epistemology and ethics – to discovering or creating norms for inquiry and thus for action. He brilliantly diagnoses the way in which attempts to substitute lots of little structures are a product of the same urge to reduce norms to facts which led to Platonic metaphysics.
On the other hand, many of the people whom Putnam criticises as ‘relativists’ (e.g. the late Michel Foucault) would not think of themselves as giving a reductive account of anything, and would eschew the whole idea of theorising about ‘epistemic notions’. They take the moral of reductionism’s failure to analyse norms into facts to be that we don’t want a ‘theory of norms’ at all; we don’t want a non-reductionist ‘account’ of truth or goodness or beauty any more than a reductionist account. Accounts of truth, or of other normative notions, only look interesting, on this view, as long as they are reductionist: a non-reductionist account of a norm can only be a description of the uses of terms, and of the other behaviour of people who obey the norm in question. It would no longer be recognisable as a philosophical account. For people like Foucault, the whole philosophical project looks like an attempt to reduce norms to non-norms, so if this project were given up there would be no more ‘eternal questions of philosophy’. Foucauldians, like Heideggerians, think of themselves as giving reasons for ceasing to ask these questions, not as giving relativist answers to them.
Putnam hopes for an account which will be philosophical without being reductive.
If there is no eliminating the normative, and no possibility of reducing the normative to our favourite science, be it biology, anthropology, neurology, physics, or whatever, then where are we?
His answer is
We might try for a grand theory of the normative in its own terms, a formal epistemology, but that project seems decidedly over-ambitious. In the meantime, there is a great deal of philosophical work to be done ... If philosophy is both transcendent and immanent, then philosophy as culture-bound reflection and argument about eternal questions, is both in time and eternity. We don’t have an Archimedean point; we always speak the language of time and place; but the rightness and wrongness of what we say is not just for a time and a place.
Whether one finds these last two sentences inspiring or vacuous, and whether one agrees that there is a ‘great deal of philosophical work to be done’, will depend upon whether one agrees with Putnam’s most controversial claim: that ‘if there is no room for a theory of truth’, if we do not identify truth with something, we shall have to deny ‘that our thoughts and assertions are thoughts and assertions’ and thus be caught in self-referential paradox. This amounts to saying that the old philosophical urge to say something about the nature of truth is not just one more bad reductionist impulse, one more bit of what Heideggerians call ‘the metaphysics of presence’, but is inseparable from any possible coherent image which human beings can form of themselves. Whether one agrees with this claim will depend largely on one’s view of Putnam’s own neo-Peircean and quasi-Habermasian account of truth, his attempt to identify truth with ‘idealised justification, as opposed to justification-on-present-evidence’. My own reaction to this attempt is that of a disenchanted ex-believer. I spent many years thinking that Peirce’s notion of ‘what would be believed at the end of inquiry’ provided the only account of the nature of truth which made sense, and drawing comfort and inspiration from it. But, thanks to the influence of Davidson’s disquotationalism and of various accounts (e.g. Foucault’s) of historical diversity in practices of justification, such an identification now seems to me as vacuous and unhelpful as the Kantian identification of ‘morally right’ with ‘justifiable to any rational agent’. I find it as hard to imagine what life at the end of inquiry would be like, what sort of person would count as an ‘ideal inquirer’, as to tell whether to count Attila, Savonarola, Suslov or Charles Manson as rational agents. I cannot imagine my descendants saying: ‘At last! Inquiry is finally over!’ So I suspect that ‘ideal justification’ is just one more reductionist attempt to turn a norm into a thing – something which, like the Mind of God, is so august and remote that it seems in bad taste to press for further details about it. This is why I am inclined to agree with Foucault, against Putnam, that ‘What is truth?’ and ‘What is rationality?’ are not eternal questions, but misguided ones – which we can stop asking without incoherence. This suspicion leads me to conclude that Putnam has not yet overcome the tension between the positivistic and reductionist aspirations of his teachers and the holistic and pragmatic debunking of those aspirations which has been the principal initiative of his own generation. On the other hand, neither has anybody else, and everybody would like to find more of a middle ground than has yet been discovered. Putnam’s repeated attempts to do so, and thereby to find the right combination of Russellian rigour and Wittgensteinian quietism, make his work central to contemporary philosophy. He is one of the few contemporary philosophers who has never got stuck defending his old ideas, adding qualifications and epicycles to avoid an admission of past error. Instead, he is constantly trying out new ideas, enlarging his own horizons, and breaking through the crust of philosophical convention. I wish there were more like him.