Approaches to Agrippan scepticism

Abstract

I begin from the intuitive requirement that a proper explanation of the possibility of empirical knowledge cannot assume any empirical knowledge in order to explain its possibility. But in conjunction with a radical scepticism I call 'the Agrippan problematic' this requirement leads to the ostensibly paradoxical observation that we require empirical knowledge to explain the possibility of empirical knowledge, and yet we cannot assume any empirical knowledge in order explain its possibility. This ostensibly paradoxical observation is at the heart of that problematic.

That observation is paradoxical, and not a straight contradiction, in virtue of how the Agrippan sceptic argues. But we should say that it is ostensibly paradoxical because the challenge to the epistemologist is to show why it is not paradoxical. To that end I delineate four applications of the general argumentative methodology of the Agrippan sceptic in regards to four epistemic loci I identify within the traditional notion of a justification or warrant for a belief. The most prominent of those epistemic loci is that of a justification itself, and in application to that locus the argumentative methodology of the Agrippan sceptic is commonly known as 'the epistemic regress argument'. I suggest that a formally sufficient answer to that argument should also enable an answer to the other three applications of the Agrippan sceptic's argumentative methodology.

There are various responses in the literature to the epistemic regress argument, however I focus on three. They are: 'foundationalism', 'coherentism', and finally the epistemology of Wilfrid Sellars. Each in a different way questions the premise of the epistemic regress argument. That premise is that all justification must be inferential. I indentify two assumptions which underlie that premise: firstly, that the only way in which a belief can justify another belief is inferentially; and secondly, that a justification for some proposition must always be epistemically distinct from a belief in that proposition.

The foundationalist questions that second assumption, essentially claiming that certain beliefs can justify themselves. However, I argue that we simply cannot make intelligible sense of the idea that a belief can justify itself except in certain specific cases. The coherentist questions neither assumption, instead suggesting that the Agrippan sceptic assumes that inferential justification is linear. However, I argue that the coherentist's position is a holistic foundationalism, and suffers from the same defects as that other position. Sellars, however, questions that first assumption, arguing that a belief can justify another belief without being inferentially connected to it. I argue that Sellars's epistemology may be taken to have the resources to illustrate that ostensibly paradoxical observation to be just that - ostensibly paradoxical.