Patrick Chandler

In this paper I raise a problem for the applicability of revision theory to absolutely general languages. First, I show that, on a number of important revision theories of definitions (reformulated in a higher-order setting compatible with unrestricted quantification), some definitions trivialize truth in the intended interpretation of the language of set theory with urelements. That is, in the presence of these definitions, every sentence in this interpretation becomes true. Second, by means of a correspondence between the theory of definitions and the theory of satisfaction, I show that one of these trivializing definitions is that which the standard revision-theoretic treatment assigns to the satisfaction predicate for the language of set theory with urelements. I conclude by considering an alternative, and somewhat less natural, reformulation of revision theory in a higher-order setting which avoids the problem I raise here, describing the higher-order principle needed to implement this alternative, and considering how it might be motivated. For those willing to countenance superpluralities, I suggest that the principle is best motivated on a plural reading.