Against metaphysics by fiat

Ruth Chang believes that one can intrinsically resolve the question of whether a vague predicate F applies to its borderline case a through arbitrary stipulation—one can resolve it by fiat. First, I clarify what it is to intrinsically resolve a question by arbitrary stipulation. Next, I argue that Chang’s view is wrong. Cases that involve vagueness are cases of (what I’ll call) competing similarity. For this reason, for any borderline case a of a vague predicate F, we have intrinsic reason to classify a as F and intrinsic reason to classify a as not-F. If we have such reasons, then we cannot intrinsically resolve the question of whether a is F by arbitrary stipulation. My conclusion has significant upshots: it undermines Chang’s case for parity as a fourth relation of comparability, and it suggests that supervaluationists ought to distance their view from Chang’s. Moreover, this paper makes progress on the so-called characterization problem of vagueness.