We cannot be certain that science provides better accounts of nature since anti-foundationalism shows that nature does not give us anything directly. But I am far from convinced that the realisation that we do not have direct access to the world precludes the possibility of better or worse accounts of nature. Indeed, is anybody unconvinced that science gives us relatively better accounts of nature (compared even with what we knew 100 years ago)? This rhetorical question is certainly weak argumentation, but it contains a kernel of an idea that we would all do well to dwell on for a moment. Sceptical arguments can always be levelled against any knowledge claim and can be pushed further and further depending on one’s purposes. But the fact that we in general do not push scepticism to its extreme is revealing. Why not? Is it because we have steadfast arguments against Berkeleyan idealism, or Cartesian scepticism? No, it is because we find such positions utterly unpalatable. Such is the reason that many of us reject radically relativistic or sceptical doctrines. This is not to say that there are not many cogent arguments against such positions, but in the end, scepticism always wins.
Instrumentalism is a response to these conflicts. Instrumentalists try to take anti-foundationalist critiques of “truth” seriously while resisting the pull of scepticism and relativism. It is an attempt to provide criteria for epistemic justification – rules for demarcating better or worse knowledge – while avoiding a troublesome realism. But many realists find instrumentalism unnecessarily timid, unwilling to make the epistemological and ontological leaps that instrumentalism seems to call for. Realists do not deny that true knowledge will tend to be instrumentally useful, but they question instrumentalists’ reluctance to make any inferences from the basic observation that some knowledge is more instrumentally useful than others. What could possibly account for this variability in instrumental success other than it being the case that some knowledge really (and truly) accounts for the way nature is? While the instrumentalist wants to say that we call true that knowledge which is instrumentally useful, the realist argues that this is really a mealy-mouthed way of suggesting that knowledge is instrumentally useful because it is true (in the most ordinary sense of the word).
Depending on one’s philosophical standards, to “prove” realism, or show beyond all doubt that science is capable of yielding accurate knowledge about are likely unachievable. But are there alternative reasons to take scientific knowledge seriously from a realist perspective? Perhaps we might consider the consequences of failing to do so. Realists differ from instrumentalists not merely in the ontological leaps they are will to make to explain useful knowledge. Crucially, they hold that there will be knowledge that will apparently have no immediate utility. There will also be knowledge that appears completely adverse to the realisation of expressed goals. Without such assumptions, there is a danger that the resistance that nature puts up against the desires of humans, especially those involved in immense technological projects, will fail to be recognised. Thus, against thinkers like Arthur Fine and Richard Rorty, I maintain that whether one approaches science, and knowledge in general, from an instrumentalist or realist perspective has profound consequences, especially as to the way science is conceived to be best integrated into broader rational-ethical systems.
I have long found ethics to be the source of a blaring inconsistency with professed anti-foundationalists; in the same breath anti-foundationalists would denounce all attempts at normative epistemology and propound ethical dogmas. It seems that after truth, ethics should be the first thing to fall prey to the scepticism that underlies anti-foundationalism, unless one believes like Berkeley that such things can be guaranteed through divine revelation (but such a claim is at least as equally specious as those of epistemological foundationalists). However, in the face of this conflict – between the desire to promote ethical principles, and recognise the fundamental problems of epistemology – one might demand consistency, but not so ethics are abandoned. Perhaps epistemology should be embraced as ethics. Just as one might feel justified in making an ethical leap of faith, one might likewise make an epistemological leap of faith.