abduction canons of reasoning for the discovery, as opposed to the justification, of scientific hypotheses or theories.
Reichenbach distinguished the context of justification and the context of discovery, arguing that philosophy legitimately is concerned only with the former, which concerns verification and confirmation, whereas the latter is a matter for psychology. Thus he and other logical positivists claimed there are inductive logics of justification but not logics for discovery. Both hypotheticodeductive and Bayesian or other probabilistic inductive logics of justification have been proposed. Close examination of actual scientific practice increasingly reveals justificatory arguments and procedures that call into question the adequacy of such logics.
Norwood Russell Hanson distinguished the reasons for accepting a specific hypothesis from the reasons for suggesting that the correct hypothesis will be of a particular kind. For the latter he attempted to develop logics of retroductive or abductive reasoning that stressed analogical reasoning, but did not succeed in convincing many that these logics were different in kind from logics of justification. Today few regard the search for rigorous formal logics of discovery as promising. Rather, the search has turned to looking for ‘logics’ in some weaker sense. Heuristic procedures, strategies for discovery, and the like are explored. Others have focused on investigating rationality in the growth of scientific knowledge, say, by exploring conditions under which research traditions or programs are progressive or degenerating. Some have explored recourse to techniques from cognitive science or artificial intelligence. Claims of success generally are controversial. See also CONFIRMATION, INDUCTION , REICHENBAC. F.S.