Induction and explanation
Part of : Philosophical inquiry ; Vol.IV, No.1, 1982, pages 1-16
Issue:
Pages:
1-16
Author:
Abstract:
I have argued that analyzing empirical inference by appeal to the best- explanation principle is valuable because it provides insight into the rich, central cases of evidence evaluaton using one of our best, most well- developed skills. As a first attempt to flesh out the analysis, I have distinguished two different kinds of empirical inference and two different kinds of supporting statement. Both distinctions seem intrinsically interesting even if they do not eventually provide they key to understanding evidence evaluation. The systematically focal role of tracedata is almost always worthy of note. And so many of our usual inference are simply diagnostic that this category is worthy of attention for this reason alone. Our evaluation of causal inference, testimony, enumerative induction, sample-to-population inference, and most of what is called ‘scientific’ inference may all be cast in this form.The greatest problems facing a best-explanation analysis of empirical inference concern non-diagnostic arguments. The circumstances in which we adduce evidence for non-diagnostic conclusions are more limited than sometimes thought, but there clearly are cases in which this is just what we do. On the surface, appeal to the inferential consequences of diagnostic conclusions looks promising. But it is not clear how plausibly this will reconstruct actual cases, nor just what latitude we have in choosing forms of inference.
Subject:
Subject (LC):
