Particularism about Truth and the Significance of Revenge

Many have suggested that all responses to the semantic paradoxes that aim to preserve classical logic face revenge paradoxes: they allow for the construction of further Liar-like objects regarding which they must either be silent or refute themselves. I consider this suggestion for the Particularist response to the paradoxes, the most sophisticated development of whichis found in Gaifman (1988, 1992, 2000). The hallmark of the Particularism is that it treats truth as a property of concrete particular representations, e.g. sentence tokens, and treats semantic pathology as arising from a token’s presence in a problematic dependency structure consisting of a network of tokens. Contra Bacon (2015, 2018) and others, I argue that Revenge arguments raise no significant challenge for the Particularist response. The Revenge sentence for the Particularist (roughly, u = “No token of u is true”) is only guaranteed to be pathological if it is understood as talking about tokens that are produced as elements of some specific interpreted language. I arguethat this should be understood in terms of a token’s being produced in a stable semantic state—a state of a linguistic agent that ensures the stable inheritance of meanings across tokens of the same linguistic type. The upshot of the Revenge Argument, I argue, is that no linguistic agent can articulate a complete truth theory for tokens produced in their own stable semantic state. This is an expressive limitation, but not an especially surprising one. I argue that the existence of this expressive limitation does not undermine the truth-theoretic virtues claimed for the Particularist response or Particularism’s claim to offer a fully general account of semantic pathology.

Open Texture and Capacious Pluralism about Content

The traditional view of propositional content presupposes that there is a privileged classification of token representations that partitions them with respect to sameness of content. I argue that there is no such thing: we utilize multiple distinct standards of sameness of content, none of which is privileged. My argument is based on the existence of semantic underdetermination. Plausibly, we use expressions (e.g. “sandwich”) that do not have determinate extensions and that are open to precisification in multiple directions. This gives rise to indefinitely many possible cases of content fission, where a semantically underdetermined token is identified as having the same content as two tokens that are precisified in opposite ways. Content fission requires recognizing a huge variety of legitimate notions of sameness of
content.

Thinking in Symbols and Leibniz's Theory of Demonstration

Torsten Odland