Here’s a synopsis I wrote for my epistemology class on BonJour’s paper, Can Empirical Knowledge Have a Foundation?
According to the epistemic regress problem, a belief A is justified, if at all, on the basis of another belief B, which must itself be justified by another belief C. This goes on ad infinitum. The question this brings up is how any of our beliefs can be justified if we need an infinite string of justified beliefs. The four ways to solve this are 1) the regress terminates with beliefs that are not justified, 2) the regress goes on ad infinitum and we are left with skepticism, 3) the regress circles back upon itself so that some beliefs are justified by beliefs that already appeared on the chain, and 4) the regress terminates with beliefs that are justified, but are not themselves justified on the basis of inference from other beliefs. Foundationalists argue that (1)-(3) are untenable, which leaves only (4), so Foundationalism should be accepted as an answer to the regress problem. This article argues that Foundationalism does not solve the regress problem.
1) If Foundationalism is true, then there are basic empirical beliefs.
2) There are no basic empirical beliefs
3) Therefore, Foundationalism is not true
(2) is the main premise BonJour wants to defend. He shows that the arguments for basic empirical beliefs fail. First, the feature that makes a belief justified in a basic way must itself provide us with a good reason to accept that belief. Let’s say F is the quality that makes belief B basic
4) B has basic-making property F
5) Beliefs with F are likely to be true.
6) Therefore, B is likely to be true.
The problem is that (2) itself is an empirical belief and needs to be inferentially justified by further beliefs. So B is not justified since it relies on other beliefs in order to be justified.
Another Foundationalist view that attempts to solve the problem is the doctrine of cognitive givenness, according to which basic beliefs are justified in an immediate or intuitive way by states of affairs rather than further beliefs. But if this intuition or apprehension that justifies a belief is a cognitive state, then it is difficult to see what justifies the intuition or apprehension itself. If it is not a cognitive state, then it is unclear how the intuition or apprehension justifies the belief.