ABSTRACT
The main aim of this book is to articulate and defend Inference to the Best Explanation as a partial answer to the descriptive problem of induction, the problem of giving a principled account of the way we actually go about making non-demonstrative inferences. As I argued in chapter 1, this is different from the justificatory problem, the problem of showing that our inductive practices are reliable or truth-tropic. After all, we might actually be unreliable reasoners. It is perhaps difficult to take this possibility seriously in the case of mundane inferences we have made hitherto about observed medium sized dry goods, but it is of course a possibility that at least epistemologists find it very easy to take seriously as regards the unobserved and the unobservable. And it may well seem that Inference to the Best Explanation makes the justificatory problem particularly recalcitrant, since it may seem particularly implausible that explanatory considerations should be a reliable guide to the truth. Indeed some may find this so implausible that, finding they cannot shake the belief that our actual methods are pretty reliable, they refuse to accept even the descriptive pretensions of Inference to the Best Explanation. That is one way that the descriptive and justificatory problems are related, and it is one reason for me to discuss the bearing of Inference to the Best Explanation on the justificatory problem of induction, to argue that it does not raise any special problem of justification (though the ordinary Humean problem is more than trouble enough). Another is simply that it is such a natural question to ask, even if you are by now irredeemably committed to the idea that explanatory considerations guide our inferences. Suppose this is in fact the way we think: what bearing does this have on the question of whether we in general and scientists in particular are in a successful truth business?
