Volume 32, 2007
John N. Williams
Pages 67-94
The Surprise Exam Paradox
Disentangling Two Reductios
One tradition of solving the surprise exam paradox, started by Robert Binkley and continued by Doris Olin, Roy Sorensen and Jelle Gerbrandy, construes surprise
epistemically and relies upon the oddity of propositions akin to G. E. Moore’s paradoxical ‘p and I don’t believe that p.’ Here I argue for an analysis that evolves from Olin’s. My analysis is different from hers or indeed any of those in the tradition because it explicitly recognizes that there are two distinct reductios at work in the student’s paradoxical argument against the teacher. The weak reductio is easy to fault. Its invalidity determines the structure of the strong reductio, so-called
because it is more difficult to refute, but ultimately unsound because of reasons associated with Moore-paradoxicality. Previous commentators have not always appreciated this difference, with the result that the strong reductio is not addressed, or the response to the weak reductio is superfl uous. This is one reason why other analyses in the tradition are vulnerable to objections to which mine is not.