Abstract

It may be argued that the language of mathematics is about the category\nof sets, although the definite article requires some justification.\nAs possible worlds of mathematics we may admit all models of type\ntheory, by which we mean all local toposes. For an intuitionist,\nthere is a distinguished local topos, namely the so-called free topos,\nwhich may be constructed as the Tarski–Lindenbaum category of intuitionistic\ntype theory. However, for a classical mathematician, to pick a distinguished\nmodel may be as difficult as to define the notion of truth in classical\ntype theory, which Tarski has shown to be impossible. Author Keywords:\nAuthor Keywords: Type theory; Categorical logic; Category of sets;\nTopos theory *1 I wish to thank Elaine Landry for organizing a symposium\nat Brock University in 1996, at which an earlier version of this\npaper was presented, and Phil Scott for helpful discussions. Support\nfrom the Social Sciences and Humanities Research Council of Canada\nis herewith acknowledged