29 October 2009, Logic and Cognition Seminar, Alistair Isaac

Peirce first identified abduction as a type of inference distinct from induction and deduction. He claimed that abduction permeates every aspect of human thought, from low-level perception to scientific theory choice. In contemporary philosophy, abduction is frequently considered only in the latter capacity, as a form of high-level scientific reasoning. In A.I., abduction is closely associated with the frame problem, the problem of how to determine relevance. Again, however, philosophers tend to identify this as a problem only for high-level reasoning. Fodor, for example, argues that low-level perceptual processes are encapsulated, and thus immune to the challenges of holistic reasoning, like abduction. In this talk I provide an overview of several different perspectives on abduction. I argue that the fundamental stumbling block for formal models of abduction is its inherently creative character. I conclude with a defence of Peirce’s original insight, arguing that empirical research on perception is a field from which models of abduction in other domains (such as scientific reasoning) can and should draw inspiration.

For more information contact Nina Gierasimczuk at N.Gierasimczuk at uva.nl.