Why Both Popper and Watkins Fail to Solve the Problem of Induction

Abstract

Accepted science not only does, but should, inform our technological practice. If someone wants to build a bridge that will stand up tomorrow or a plane that will fly tomorrow she should assume in particular that currently accepted low-level generalisations will continue to hold tomorrow. Someone who claimed (without evidence) that falling bodies will soon start to fall with an acceleration which increases as the cube of the time of fall would be regarded as downright irrational. Someone who encouraged passengers to fly on an aeroplane built on that supposition about future falling bodies would be regarded as criminally irresponsible. But we know, following Hume, that, since all the observational evidence we have for the generalisations accepted by science is of necessity evidence about the past, and since deductive logic is not content-increasing, we certainly cannot deductively infer that accepted generalisations will continue to hold in the future from any amount of evidence we may have. But what then is the basis for these very firm judgments about rationality and responsibility? This is, of course, the notorious ‘pragmatic problem of induction’. John Watkins has recently joined the long list of philosophers who have attempted to solve the problem.