On Classical Determinate Truth

TL;DR

This paper introduces new classical, compositional theories of truth and determinateness that unify classical logic with a natural class of determinate sentences, extending Kripke-Feferman truth and comparing favorably to recent theories.

Contribution

Theories are fully compositional, classical, and include a defined determinateness predicate, providing a natural and principled approach to truth and determinateness in logic.

Findings

Theories capture a natural class of determinate sentences.

They are proof-theoretically equivalent to Fujimoto and Halbach's $ ext{C}^+$.

Negative results show primitive determinateness predicates limit semantic rule formulation.

Abstract

The paper proposes and studies new classical, type-free theories of truth and determinateness with unprecedented features. The theories are fully compositional, strongly classical (namely, their internal and external logics are both classical), and feature a \emph{defined} determinateness predicate satisfying desirable and widely agreed principles. The theories capture a conception of truth and determinateness according to which the generalizing power associated with the classicality and full compositionality of truth is combined with the identification of a natural class of sentences -- the determinate ones -- for which clear-cut semantic rules are available. Our theories can also be seen as the \emph{classical closures} of Kripke-Feferman truth: their $\omega$-models, which we precisely pinned down, result from including in the extension of the truth predicate the sentences that are satisfied by a Kripkean closed-off fixed point model. The theories compare to recent theories proposed by Fujimoto and Halbach, featuring a primitive determinateness predicate. In the paper we show that our theories entail all principles of Fujimoto and Halbach's theories, and are proof-theoretically equivalent to Fujimoto and Halbach's $\cdplus$. {We also show establish some negative results on Fujimoto and Halbach's theories: such results show that, unlike what happens in our theories, the primitive determinateness predicate prevents one from establishing clear and unrestricted semantic rules for the language with type-free truth.