Abstract

I argue that though a satisfactory semantics for the logic of inexact reference may assign no truth value to some statements, it should not assign truth (or falsity) of various degrees. Well-formed assertions are simply true or not. Inexactness does not “ramify.” I distinguish inexactness from other sorts of vagueness, including nonspecificity. I show that arguments from (i) use of quantifiers, (ii) the existence of properties which can be construed as a series of properties (as, e. g., red can be construed as a set of shades of red), (iii) the constructability of apparently paradoxical sorites arguments, and (iv) the presence of prototypes in the extension of a predicate do not show that there are degrees of truth.Much of the alleged evidence that inexactness ramifies is, in fact, a misreading of the undeniable evidence that there may be uncertainty about the truth value of a claim. In support of my claims, I discuss how cases of deeming that a predicate applies relate to its actually applying. A distinction between predicates of “pure” and “impure” function is essential to this.