Current Philosophy

0026 Timothy Williamson. "Realism and the Burden of Proof," Irish Philosophical Journal 3:1 (Spring 1986), pp. 42-57.

Williamson's direct aim is to uncover the inadequacies of Hilary Putnam's recent attempt to refute realism. But in the course of doing so, he makes some significant remarks about the nature of philosophical argument in general.

There are two kinds of arguments against a theory, he says. The 'weak' kind of argument shows that the theory need not be true, that we are not obliged to believe it. The 'strong' kind shows that the theory cannot be true, that we are obliged not to believe it. The ■weak argument may be effective on a neutral party, but strong argument is required to shake someone already committed to a theory. Or so it would seem. But Williamson observes that a 'neutral' party is not really neutral. If the weak argument works, it is only because it brings to light some latent disposition against the theory in question. So only strong argument is really valid.

And even here, only the strongest of strong argument is really satisfying: the r*(iiir:tinn prf gjjjarirW^rp Anything less requires the use of premises from outside the theory, and is therefore both logically and psychologically open to dispute. Of course, such a stringent requirement for arguing against philosophical theories is rather discouraging. It makes it very difficult to knock contending theories from the field.

But if we apply these considerations to the realism/anti-realism debate, the prognosis is more hopeful. For by its own lights, anti-real ism requires that theories be verifiable. Therefore, if anti-realism is not able to positively verify itself, then it must withdraw. A tie goes to realism. Williamson makes short work of some attempts to get around this conclusion, showing that any loosening of the ideas of refutation or verification that would be required opens up too much room for other, unacceptable forms of argument. He draws the conclusion, therefore, that while the failure of anti-realism does not necessarily assure the validity of realism (for one could assert that all philosophical positions are meaningless), any plausible form of anti-realism requires a strong refutation of realism. Realism must be shown to be incoherent on its own terms.

Williamson then examines whether Putnam has, in Reason, Truth and History (1981), shown this. He looks at two broad arguments.

First, Putnam has stipulated that realists need to show that words can refer to entities uniquely. Williamson admits that the realist cannot do so without in the very act using words that he assumes refer uniquely. At worst this would be begging the question, although Williamson believes that any demand to avoid it would be unreasonable in this instance. But in any case, the realists' position is not inconsistent.

Second, Putnam has questioned the meaning of 'reference'. What can we mean when we say that a ■word refers to entities? In particular, what does the word 'reference' itself refer to? If it is a certain causal relation, what makes it so? Williamson analyzes Putnam's handling of these issues. He seems to agree that reference is a non-reducible concept: we cannot account for it in terms of anything else. But it is not an inconsistent concept, nor one that seems to be threatened by any imaginable future development.

In short, Putnam has not shown realism to be incoherent. No disproof of realism exists, and therefore anti-realism fails. Yet, Williamson adds, neither is there any proof that such a disproof is forever impossible.