ConcludingRemarks

It is not uncommon for proponents of semantic indecision to gloss the phenomenon as follows: an expression is semantically undecided if and only if there are several candidate precise meanings that it could express, but it is indeterminate which of those meanings it expresses.¹¹ This characterization, however, invokes an indeterminacy operator that is not available without embracing a theory of propositional indeter­minacy.

Some Supervaluationists will no doubt maintain that the use of an indeterminacy operator in this formulation is just a manner of speaking. However, it is far from clear that the linguistic theorist can reformulate this slogan in more appropriate linguistic terminology. Flat-footedly translating the gloss using linguistic idioms delivers the following: an expression E is semantically undecided if and only if there are several meanings that are candidates for satisfying the predicate ‘is expressed by E, and the application of ‘is expressed by E^, to any of them is semantically undecided. (Due to issues relating to ‘quantifying in’ we have formulated this in terms of a linguistic predicative undecidedness instead of sentential borderlineness; see section 4.4.) On this picture, the semantic undecidedness of an expression E is explained in terms of the semantic undecidedness of another metalinguistic expression, ‘is expressed by E’. Regress looms.

Ironically, the non-linguistic theory of vagueness adopted here allows us to take this account of semantic indecision at face value. On this picture, even semantic indecision cannot purely be explained by appealing solely to linguistic practices of those using the language: one must also invoke a borderlineness operator that is not explained in terms of language.

18