Arguments for relativism

Relativism is the radical offspring of non-realism, which is itself descended from the idealism of Berkeley and Kant. Non-realism holds that we cannot meaningfully talk about they way things are independent of our experience of them: to use Michael Dummett's formulation, what makes a statement true is not independent of our procedures for deciding it is true. The main argument in favour of non-realism is essentially negative: it avoids the difficulties endemic to metaphysical realism (a.k.a. "objectivism" or "absolutism").
Realists hold that our judgments are true when they accurately describe or correspond to a reality that exists independently of our perceptions, conceptions, theories or desires. On this view, a true statement such as "water contains oxygen" describes a fact about this independent reality. It rests on a scientific model that may be said to "carve nature at the joints". But an obvious question arises: how can we determine that our judgments are true in this sense? The obvious answer is that we test them by making experiments and observations. I say it will snow today, and I test this by watching the sky. I say water contains oxygen and I confirm this by showing that one of the elements separated out by electrolysis supports combustion. When our assertions are decisively confuted by experience, we conclude that they are false—i.e. they describe a state of affairs that does not obtain.
Relativists accept that this is how we normally conceive of truth and falsity—in ordinary usage, the word "true" means something like "corresponds to the facts"--and as an account of our everyday epistemic procedures it is unobjectionable. But they argue that it loses coherence if it is elevated to the metaphysical level. For what is really happening, even when we are confirming the most mundane belief about the empirical world, is that we are satisfying ourselves that this belief coheres with our other beliefs. We confirm that the sea is salty by tasting it or by conducting a chemical analysis of seawater. But these procedures only confirm our belief about sea water in the sense of showing it to be compatible with or even entailed by a host of other beliefs: for instance, that the sample we are examining is typical; that nothing else tastes quite like salt; that our sensory faculties are trustworthy on this occasion; that salt tastes roughly the same at different times. What we can never do, argue relativists and other non-realists, is check the degree of correspondence between our judgments and reality as it is independent of our experience of it. To do this we would have to take a "sideways on" view of the cognitive relation between subject and object. But this is impossible since any vantage point we adopt will necessarily be that of the subject. For the same reason, we cannot compare our overall conceptual scheme or theoretical model of reality with reality as it is "in itself."
The driving idea behind empiricism and the upshot of Kant's critique of speculative metaphysics is thus that concepts must be tied to experience if they are to have legitimate employment in science or philosophy. Relativists argue that the metaphysical realist's concept of truth fails this test, for it takes the notion of "correspondence with reality" out of its everyday employment, where it is genuinely useful, and tries to press it into metaphysical service, where it is neither useful nor legitimate. So even if, in its normal usage, "truth" means something like correspondence with reality, the ultimate criterion of truth turns out to be coherence with other beliefs. To put it another way: our philosophical conception of truth cannot simply be an expanded version of our commonsense notion of truth as correspondence. And this implies that truth must always be relative to some belief system, to some particular epistemic standpoint. This is the first of the two theses identified above as constituting the doctrinal kernel of relativism. Numerous philosophers have affirmed it. Yet many of these have sought to avoid relativism by rejecting the second thesis—that no standpoint is metaphysically privileged over all others.
This second thesis is what gives relativism its bad name. Critics commonly reduce it to the claim that any point of view is as good as any other and then attack it with some variation of Plato's arguments against Protagoras. But virtually no well-known philosophers actually hold that all standpoints are of equal worth. Richard Rorty, for instance, who is widely regarded as a relativist, dismisses that position as "silly." (Richard Rorty, Objectivism, Relativism, and Truth, p. 89). Rorty, Kuhn and most other relativists accept that one can have cogent reasons for preferring one standpoint to another; the preferred point of view may, for instance, exhibit greater logical consistency or greater predictive power than other available perspectives. But they argue that such reasons cannot confer any special metaphysical status on the standpoint in question. They cannot, for instance, show it to be the one favoured by God, or dictated by Reason, or most in accord with human nature.
Relativists typically justify this conclusion along the following lines. Any proof of a standpoint's superiority must rest on premises that express fundamental assumptions and basic values. For instance, arguments for the superiority of the standpoint of modern science over that of religion will presuppose the value of consistency, of solving theoretical puzzles, and of being able to manipulate one's environment. A person who defends the literal truth of the bible but shares these values is likely to be persuaded fairly quickly by these arguments. But a person who holds that truth appears to humans as paradoxical, and who values tradition and religious faith over experimental evidence and predictive power will not be persuaded. An argument can only be convincing to one who accepts its premises. Some premises, though, like those just mentioned, are so fundamental that they are not usually argued for at all. Rather, they are constitutive of a particular outlook.
The relativists' thesis is not that one cannot support standpoints with arguments; it is that in the end all such arguments must be circular since they inevitably rest on premises that are themselves part of the standpoint. Critics will here point out that there is a difference between denying that the superiority of one standpoint over all others can be proved and denying that such a standpoint exists. In reply, relativists are likely to claim that this distinction is an abstract one that no consistent empiricist or pragmatist would make. To insist that one standpoint is objectively superior to all others, they argue, even though there is no way of proving this, is dogmatic and pointless; to claim that one's own standpoint enjoys this unique but undemonstrable superiority is dogmatic and implausible.
A critic might also object that what relativists call "cogent" reasons for preferring one standpoint to another are not epistemically relevant: that is, they do provide grounds for thinking that the standpoint generates or ensures beliefs that are objectively true. But this is clearly a point most relativists would be willing to concede. The notion of objective truth referred to here is not a concept for which they have a use, preferring instead something like William James' conception of truth as "what is good in the way of belief.