Browsing by Subject "Vagueness"

This thesis grows out of a fascination with the vagueness of natural language, its manifestation in the ancient Sorites paradox, and the way in which the paradox is dealt with in fuzzy logic. It is an attempt to resolve the tension between two versions of the paradox, and the related problem of whether identity can be fuzzy. If it can be fuzzy, then the most popular argument against vague objects is mistaken, which would be great news for those who hold that there can be vagueness in the world independently of our representation or knowledge of it. The standard Sorites is made up of conditionals about an ordinary predicate (e.g. “heap”) by the rule of modus ponens. It is typically solved in fuzzy logic by interpreting the predicate as a fuzzy relation and showing that the argument fails as a result. There is another, less known version of the paradox, based on the identity predicate and the rule of substitutivity of identicals. The strong analogy between the two versions suggests that their solutions might be analogical as well, which would make identity just as vague as any relation. Yet the idea of vague identity has traditionally been rejected on both formal and philosophical grounds. Even Nicholas J. J. Smith, who is known for his positive attitude toward fuzzy relations in general, denies that identity could be fuzzy. The opposite position is taken by Graham Priest, who argues for a fuzzy interpretation of identity as a similarity relation. Following Priest, I aim to show that there is a perfectly sensible logic of fuzzy identity and that a fuzzy theoretician of vagueness therefore cannot rule out fuzzy identity on logical grounds alone. I compare two fuzzy solutions to the identity Sorites: Priest’s solution, based on the notion of local validity, and B. Jack Copeland’s solution, based on the failure of contraction in sequent calculus. I provide a synthesis of the two solutions, suggesting that Priest’s local validity counts as a genuine kind of validity even if he might not think so himself. The substitutivity of identicals is not locally valid in Priest’s logic, however; his solution only applies to a special case with the rule of transitivity. Applying L. Valverde’s representation theorem and other mathematical results, I lay the foundation for a stronger logic where the substitutivity rule is locally valid and the two Sorites merge into one paradox with one solution. Finally, I defend fuzzy identity against Gareth Evans’ argument that vague identity leads to contradiction, and Smith’s argument that vague identity is not really identity. The former relies on a fallacious application of the substitutivity rule; to the latter, my principal response is to question Smith’s understanding of identity and argue for a broader one. I conclude that not only is fuzzy identity logically possible, but it also has potential applicability in metaphysics and elsewhere.