Epistemicism about vagueness is commonly thought implausible for its commitment to absolutely precise boundaries separating, e.g., the bald heads from the non-bald. I argue that attempts to escape such precise semantic boundaries fail; but I also argue that precision is anyway a red herring. The deeper problem is to explain how a vague term manages to have any particular semantic content at all – sharply bounded or not. Epistemicism's precision calls attention to difficulty, but does not generate it; the problem attaches equally to any view of vague language as meaningful. I argue that the best approach to this problem is to understand any interpretation of a vague language as relative to a given precisification of core semantic concepts. Though motivated independently of epistemicism, such an approach both (i) is compatible with epistemicism and (ii) captures indeterminist intuitions about vagueness better than do indeterminist semantic theories.