ABSTRACT

If classical foundationalism turns out to be a bankrupt epistemology, what alternative does RE offer? Perhaps surprisingly, the replacement has some close parallels to what it replaces: it offers what we might call Reformed foundationalism. It accepts that our justified beliefs can all be divided into the two great classes, basic and derived; that our derived beliefs are justified only if they are supported by our properly basic beliefs, directly or indirectly; and that all the beliefs described by classical foundationalism as properly basic are indeed properly basic. Its central difference from classical foundationalism is that it insists that the class of properly basic beliefs, those which we are entitled to hold without having any reasons or evidence to back them up, is very much wider than classical foundationalism allows. For as we have just seen, Reformed foundationalism classifies both memory beliefs and a belief in God as properly basic. From this basic difference flow a number of other differences. First, RE does not require that our properly basic beliefs should be indubitable or incorrigible or infallible – it does not even require that they should be true. It says only that it must be permissible to hold such beliefs even when you have no supporting reasons or evidence. Second, it allows that there can be good arguments against any given person’s properly basic belief; and that when a person comes to know of these objections to what she believes, she is entitled to continue with this belief only if she can show where the objection goes wrong. For example, suppose I remember (as it seems to me) that last week there was a forest fire in the local woods. RE allows that memory beliefs can be properly basic, so I am fully entitled to believe that I remember there was a forest fire, even though I do not have any reasons for this belief. Suppose someone then produces good evidence that my belief is false – perhaps eyewitness reports of people who were in the local woods but who deny that there was a fire. I am then no longer entitled to hold on to my memory belief. If I retain it as a basic belief, it will not be properly basic for me. But then suppose that I investigate the matter further, and discover that all the eyewitness reports relate to the day before the day on which I remember the fire occurring. I then have evidence to undermine the initial reports which seem to disprove my initial memory belief, and I can revert to my original memory belief as a properly basic belief. In a terminology which has been increasingly common, the eyewitness reports are a defeater for my original memory belief; and my discovery of the discrepancy in dates is a defeater of that defeater. So, although there is a defeater to my original memory belief, I have a defeater for that defeater, and for this reason I can retain the original belief not merely as basic, but as properly basic.