Solomon Feferman

		Logic, Logics, and Logicism


This is the lecture I gave a couple of weeks ago at the Boolos
Symposium in Notre Dame.  The first half is close to a lecture I gave
to the CSLI meeting on Logic, Language and Information last June, on
Tarski's proposal to characterize logic according to what is invariant
under arbitrary permutations of the underlying domain.  I explain what
is the main problem with that proposal, and then propose a new and
improved characterization to overcome that problem.  This can be
improved still further, I'm sure, and suggestions in that direction
will be welcome.  In the final part of the lecture, I connect this
with the logicist program as viewed by Tarski and, latterly, Boolos.