Burge also presents a theory of truth and sketches how the usual Tarski biconditionals can be derived, and one of the important points of his paper is that the attempt to give a theory of truth places certain constraints on the logic of the object language.' In the process of advocating his theory, Burge makes a number of comparisons between it and my intensional free description theory (IFDT) [2]. (The theory is intensional in that formulas of the form (x) (Ax < Bx) -+ xx Ax = 7xBx are not valid.) One of the matters about which we disagree is whether identities involving non-denoting terms are ever true. I think that for many speakers of English "Pegasus is Pegasus" is true, and one of the central motives in the development of my system was to preserve this fact. Burge finds speakers divided on