Jeff B. Paris
(University of Manchester)
|Polyadic Inductive Logic and Spectrum Exchangeability
|Πέμπτη, 02 Οκτ 2014
Εθνικό και Καποδιστριακό Πανεπιστήμιο Αθηνών, Mithe, G. Gkountarouli
Inductive Logic a la Carnap & Johnson was founded on the idea that
the probabilities we assign in our everyday lives may be logical in the sense that
they are determined by our knowledge via general rationality considerations.
[So this is the ‘Inductive Logic’ of Philosophy and is not the same as the ‘Inductive
Logic’ of Computer Science.] As a potentially applicable topic Inductive Logic
is nowadays generally considered within Philosophy to have failed, largely on
account of Nelson Goodman’s GRUE Paradox.
However as a ‘pure’ mathematical investigation into rationality, with its relevance
to the aspirations of AI, it survives and has recently seen some interesting de-
velopments. In particular the serious investigation into Polyadic Inductive Logic
(Carnap et al only ever considered unary properties) has uncovered some new
rational principles which appear to be remarkably powerful, even when reflected
down to the unary level.
In my talk I shall try to give a fairly informal introduction to one such principle.