Putnam, Hilary (1926–2016)

7. The incoherence of metaphysical realism

In 1976 Putnam’s philosophy underwent what appeared to be a major shift: rejecting what he referred to as ‘metaphysical’ realism, he adopted ’internal’ realism in its stead. The attack on metaphysical realism, first presented in the final chapter of Meaning and the Moral Sciences (1978) is elaborated on in "Models and Reality" (1980; reprinted in 1983) and in Reason, Truth and History (1981).

Whereas establishing the objectivity of reference, manifest in the possibility of referring to the same entities by means of different theories and descriptions, was central to Putnam’s earlier realism, the dispersion of reference into a plurality of possible relations is at the heart of his criticism of metaphysical realism. The argument draws on model-theoretic considerations. The Löwenheim–Skolem theorem entails that a first-order theory, rich enough to contain arithmetic, does not determine its models up to isomorphism (see Löwenheim–Skolem theorems and nonstandard models). Putnam extrapolates: even an ideal theory of the world, complying with all empirical and theoretical constraints, will not define a unique model, that is, a unique reference relation. In particular, causality, previously seen as anchoring language in reality, now becomes just another relation, and hence open to interpretation. The internal realist meets the challenge of the Löwenheim–Skolem theorem by questioning the platonic notion of model on which it is based. Influenced by Michael Dummett's intuitionist philosophy of mathematics, Putnam takes models to be human constructions rather than abstract entities situated in a platonic world. As such, they are given to us by our methods of construction and do not suffer from the relativity of reference that follows from the theorem. Intuitionism is based on a verificationist theory of meaning and truth that replaces the realist concept of truth with provability or (outside of mathematics), warranted assertability. Adopting such a verificationist position, a position he had criticized in earlier writings, was the high price Putnam was willing to pay for the solution to the conundrum of reference. Moreover, during the period of transition in the early 1980s, he felt that functionalism, being completely 'internalist' supported, perhaps even required, a verificationist theory of meaning. He therefore took internal realism to be in harmony with his functionalist philosophy of mind. Within a decade, however, Putnam was beginning to reconsider the verificationist solution to the Löwenheim–Skolem problem and eventually abandoned internal realism. Many of the arguments advanced during the 'internal' period remain valid, however, despite this reconsideration. Such are, for example some of the arguments against scepticism.

Putnam emphasized all along that internal realism does not lead to scepticism. To the contrary, his argument was that both the metaphysical realist who believes there is a unique theory, true of everything, including the supposedly unique reference relation, and the sceptic who denies truth and knowledge altogether, are making the same mistake – they are assuming a non-existent vantage point external to any language or description-scheme. From the internal perspective, questions about reference cannot arise, ’chair’ refers to chairs, ’cherry’ to cherries. 'To speak as if this were my problem, "I know how to use my language, but, now, how shall I single out an interpretation?" is to speak nonsense. Either the use already fixes the "interpretation" or nothing can’ (1983: 24). This response to the problem of reference did not lead Putnam to revise his externalist conception of meaning, but, as Putnam saw later, it deflated it, preventing it from providing genuine support for realism. The repudiation of scepticism is a recurrent theme uniting Putnam's earlier and later work. The strategy, however, changes; Putnam’s responses to Quine, who has also invoked the Löwenheim–Skolem theorem to argue for the relativity of reference, illustrate this difference. Whereas Putnam initially tried to reduce indeterminacy by increasing the number of constraints on an adequate translation, he later came to see the problem itself as a sceptical variation on a misguided metaphysics.

Another argument against scepticism is found in Reason, Truth and History (1981), where considerations of reference and intentionality lead Putnam to conclude that the sceptic’s favorite fantasy – that we are all brains in a vat – is self-refuting (see Scepticism). Whereas the previous argument against scepticism employs the internal perspective, the 'brains in a vat' argument is based on Putnam's externalist account of meaning and remains intact by his later rejection of internal realism. On the externalist account, the words of an isolated brain that lacks contact with reality cannot have their standard reference and meaning. In the brain-in the-vat language 'tree' does not refer to trees and 'vat' does not refer to vats but rather (if it refers) to some image produced by a computer or an evil scientist. Putnam concluded: 'If we are brains in a vat, then "We are brains in a vat" is false. So it is (necessarily) false' (1981: 15).

While a long philosophical tradition sees realism and antirealism as exhaustive alternatives, Kant, the American pragmatists, and Wittgenstein, all strove to overcome this dichotomy. Putnam saw the internal realist as belonging to the latter camp. In his earlier writings, he represented realism as an explanatory hypothesis, on par with scientific theories. Internal realism, however, was not intended to be a theory, let alone a scientific theory. Rather, it exposed sceptical attitudes as incoherent or meaningless. Wittgenstein’s influence, and particularly, his suspicion of philosophical theories, is perceptible here, not only in the subtle similarity between the extended Löwenheim– Skolem theorem and the rule-following paradox, but also in the change of perspective that constitutes the response to these related problems. With time, this affinity gave way to a more critical attitude towards Wittgenstein's philosophy of mathematics and his therapeutic vision of philosophy. Putnam's gradual withdrawal from internal realism and the growing distance from Wittgenstein go hand in hand. It was, however, when Putnam came to the conclusion that verificationism (see also Meaning and verification) is tantamount to solipsism and is in fact incoherent, that he finally gave up internal realism. His reasons are spelled out in "Corresponding with Reality" (2011; reprinted in 2012), where he also explains the connection between his realist semantics and the need to modify functionalism. He now had to readdress the problem of reference, an endeavor leading to the above mentioned work on perception, where the departure from Wittgenstein is even more conspicuous.

Is the return of realism also a return to metaphysical realism? As of 2007 Putnam's answer to this question is positive, but has a caveat: If "metaphysical realism" stands for the conviction that there is a uniquely correct and mind-independent representation of objects and their properties, then Putnam, early and late, is not a metaphysical realist. For one thing, (as mentioned in section 2), Putnam sees the possibility of different though equivalent descriptions as fundamental to science. For another, language is a human creation that has more than a single way of fitting the world. Nonetheless, truth under a description is objective according to Putnam and upholding the possibility of such truth is what makes him a realist.