Adriano Aguado Nocilla
University of Barcelona
The notion of proposition has been central to analytic philosophy due to its multiple explanatory roles: as bearer of truth, semantic content of utterances, object of propositional attitudes, and referent of that-clauses. However, these roles generate a well-known tension. On the one hand, propositions must be coarse-grained enough to account for what synonymous utterances share across and within languages (e.g., “Superman flies” / “Clark Kent flies”; “Superman flies” / “Superman vuela”). On the other hand, they must be fine-grained enough to individuate beliefs, since it is widely accepted that believing that Superman flies differs from believing that Clark Kent flies. These two roles appear to impose conflicting requirements on the structure of propositions.
Existing approaches attempt to resolve this tension by distributing explanatory work. Salmon (1986) treats such sentences as expressing the same proposition but differing in modes of apprehension, while Soames (1987) appeals to attitudes toward sentences. Yet these views face objections and offer only partial accounts of how propositional content relates to modes of presentation.
This paper proposes a type/token framework to address the problem. Propositions are understood as types: psychological abstractions of content structured in terms of individuals, properties, and relations. They are not tied to any particular linguistic or representational form. By contrast, propositional vehicles—utterances, beliefs, desires, and other attitudes—are tokens that instantiate these types through specific modes of presentation, such as particular languages, names, or descriptions.
This distinction allows us to reconcile the competing demands on propositions. At the type level, propositions are coarse-grained and capture the shared content of synonymous utterances. At the token level, differences in modes of presentation account for the fine-grained distinctions required to individuate beliefs and explain substitution failures. Thus, “Superman flies” and “Clark Kent flies” can express the same propositional type while corresponding to distinct tokens that explain cognitive differences.
Unlike related accounts, this proposal offers a naturalized connection between types and tokens. Proposition types are not abstract entities but psychological abstractions grounded in agents’ representational capacities, arising from similarities among tokens. This provides a unified framework in which propositions can fulfill their various explanatory roles while resolving the problem of fineness of grain.
Chair: tba
Time:
Location:
SOPhiA 2026 