Infinite Epistemic Regress--Seminar Class Notes

Phil. 2208 Contemporary Epistemology Meeting #15 Class Notes Dr Anne Newstead The Problem of Infinite Epistemic Regress Suppose that knowing that p requires having a justification for one’s belief that p. Will the requirement of justification ever come to an end [1] In order for S to know that p, it must be the case that S possesses an independent justification of p , i.e. some grounds for thinking that p other than just the belief that p is true. Let’s call the independent justification for p, the proposition that q. Since it is not enough for knowledge that one’s support be flimsy—a mere guess—S’s awareness of q has to be—it seems—knowledge as well. [For if S’s awareness of q is merely a hunch, then q could turn out to be false, and S would be relying on false evidence for her belief that p. As we’ve seen significant reliance on false evidence, does seem to preclude knowledge.] [2] In order for S to know that p, S must also know that q. But this gives us quite a general schema for generating a regress. Suppose that the independent justification for q is called the proposition that r, and that for r, s and so on. [3] In order for S to know that q, S must also know that r. …. More formally, we have a recursive definition, using the following symbols: Ksp means S knows that p. Js(x,y) means x justifies y for S. Then: (1) Ksp  Ks (q) & J (q,p) (2) Ksq  Ks (r) & J (r,q) … Ksn  Ks (n+1) & J (n+1,n) This formula means that to know any proposition n, one has to know some further distinct proposition n+1, where this distinct proposition n+1 justifies n. It follows that one can never know a single truth without knowing infinitely many truths.

2

Infinite Regress Scepticism (1) To know a single truth, one has to know infinitely many truths. (2) But one can never know infinitely many truths at once. (3) Therefore, one does not know even a single truth.

Another version (due to Hetherington in KP, ch. 22): (1) In order to know that p, you cannot be merely guessing that p. (2) But you are always merely guessing that p. (3) So you never know that p. The argument for (2) is an infinite regress argument. Since there will always be in any finite chain of justifying propositions, some last proposition that is itself not further justified (by another proposition), then one might conclude that one’s justification for a belief always has an aspect of being a guess—because it lacks complete justification.

Responses to such scepticism Most responses try to block the infinite regress, by denying that one does need an endless chain of justifications in order to know a single belief. One assumption needed to get the regress going is the assumption that only another belief or proposition can justify another belief or proposition. If one rejects this assumption, then one might hold with foundationalists that some beliefs are basic: they rest on no other beliefs for justification, but are justified nonetheless. Descartes’ cogito is supposed to be an example of a basic belief, one is that is self-evident and provides, as it were, its own justification. Hume makes knowledge of sensations basic and is an empiricist foundationalist. Another response is to deny that everything must have a justification and to allow that justification comes to an end. Cartesian foundationalism does provide a justification for every belief, since the cogito is self-justifying. Contextualists hold that in some contexts some beliefs are basic in a modified sense: they are used to justify other beliefs, but do not have a further justification for themselves in terms of yet another belief. Instead such basic framework or hinge propositions are entrenched in a practice or way of life (Wittgenstein). (On some interpretations, one can view Wittgenstein’s contextualism as a heterodox form of foundationalism.) Finally, some theorists reject the atomistic model of justification suggested by the regress argument, and opt instead for the coherentist position, on which what justifies a belief is

3 not one other beliefs, but its coherence with one’s entire set of beliefs. On this model, beliefs are mutually supportive of one another, and justification is holistic.

We have already encountered a version of infinite regress scepticism in Hume’s Enquiry—p.46: ‘If I ask why you believe any particular matter of fact, which you relate, you must tell me some reason; and this reason will be some other fact, connected with it. But as you cannot proceed after this manner in infinitu, you must terminate in some fact, which is present to your memory or senses; or must allow that your belief is entirely without foundation’. (Enquiry V.I)