Loading field map
Is the meaning of a sentence determined by the conditions under which it would be true, or by the way speakers use it to communicate? This question has driven the philosophy of language for centuries. The first answer—that meaning is a matter of reference and truth—leads to formal, world-directed theories. The second—that meaning is a matter of practice, intention, and context—leads to use-based accounts. The twelve frameworks in the history of the subfield can be understood as different ways of developing, combining, or rejecting these two poles.
Long before the modern era, medieval philosophers developed a sophisticated semantic framework known as supposition theory. A term's "supposition" was its reference in a particular propositional context, distinguished from its ordinary signification. Scholastics also divided expressions into categorematic terms (which have independent meaning, like nouns) and syncategorematic terms (which modify other terms, like quantifiers). The framework assumed that spoken language ultimately represents mental language—a universal system of concepts shared by all humans. This mental-representation assumption gave the Scholastic framework remarkable longevity, surviving into the early modern period as the default model of meaning. It was eventually displaced not by a single rival but by a series of new approaches that abandoned the mental-language picture.
The early twentieth century saw two closely related frameworks emerge from the work of Gottlob Frege and Bertrand Russell. Ideal Language Philosophy held that natural language is logically defective and must be replaced or regimented into a logically perfect notation—one in which each expression has a unique, unambiguous sense. Frege's distinction between sense (Sinn) and reference (Bedeutung) and Russell's theory of definite descriptions were tools for this project. Descriptivism applied this approach specifically to proper names: a name like "Aristotle" is semantically equivalent to a definite description such as "the teacher of Alexander the Great." The two frameworks shared a Frege–Russell root and reinforced each other, but they differed in scope. Ideal Language Philosophy aimed to reform all of language, while Descriptivism focused on the reference of singular terms. Both treated meaning as a matter of truth conditions: to understand a sentence is to know the conditions under which it is true.
The Vienna Circle took the Ideal Language program in a more radical direction. Logical Positivism added a verificationist criterion of meaning: a sentence is cognitively meaningful only if it can be empirically verified or is analytic. Metaphysical claims were dismissed as nonsense. This framework narrowed the scope of philosophy of language to the analysis of scientific language. Its legacy was indirect but powerful: by insisting that meaning be tied to verification conditions, Logical Positivism set the stage for later truth-conditional theories. The verificationist criterion itself was eventually abandoned, but the idea that semantics should be grounded in formal methods—logic, set theory, model theory—survived and influenced the development of Truth-Conditional Semantics and Formal Semantics.
Ordinary Language Philosophy explicitly rejected the Ideal Language program. Led by the later Ludwig Wittgenstein, J. L. Austin, and Gilbert Ryle, it argued that natural language is not defective; philosophical problems arise from misusing it. Wittgenstein's concept of language-games emphasized that meaning is determined by use in specific forms of life. Austin's careful attention to the nuances of everyday speech—such as the difference between "I know" and "I believe"—showed that ordinary language contains distinctions that formal theories miss. This framework did not merely oppose Ideal Language Philosophy; it offered a positive method: examine how words are actually used to dissolve philosophical puzzles. Ordinary Language Philosophy coexisted with the formal program for a time, but its influence waned as formal methods became dominant. However, it left a lasting legacy in the next framework.
Speech Act Theory grew directly out of Ordinary Language Philosophy. Austin observed that many utterances are not statements that can be true or false; they are actions. When someone says "I promise to be there," they are not describing a promise but performing one. Austin distinguished three dimensions of a speech act: the locutionary act (uttering a sentence with a certain meaning), the illocutionary act (the action performed, such as promising or ordering), and the perlocutionary act (the effect on the hearer). John Searle later systematized the theory with a taxonomy of illocutionary acts. Speech Act Theory has since become a partly independent research tradition, branching into feminist speech act theory and institutional speech act theory. It remains active today, especially in philosophical pragmatics, where it complements Gricean approaches by focusing on the force of utterances rather than their implied content.
In the 1970s, Causal Theory of Reference directly challenged Descriptivism. Saul Kripke argued that proper names are rigid designators: they refer to the same object in all possible worlds, not via a description. A name like "Aristotle" refers to the actual person who was baptized with that name, not to whatever satisfies a description. Hilary Putnam extended the argument to natural kind terms like "water" and "gold": their reference is fixed by a causal-historical chain linking speakers to the substance itself, not by a set of defining features. This framework replaced Descriptivism as the dominant theory of reference for proper names and natural kinds. It also opened the door to Externalist Semantics by showing that meaning depends on facts about the world, not just on speakers' mental states.
Two closely linked frameworks developed the truth-conditional insight into rigorous theories. Truth-Conditional Semantics, associated with Donald Davidson, used Alfred Tarski's truth definitions to give a recursive account of meaning: a theory of meaning for a language is a theory that entails, for every sentence, a statement of its truth conditions. Formal Semantics, pioneered by Richard Montague, went further by applying the model-theoretic methods of mathematical logic to natural language. Montague treated English as a formal language that could be interpreted in set-theoretic models. The two frameworks share the core assumption that meaning is compositional—the meaning of a complex expression is a function of the meanings of its parts and their mode of combination. They differ in method: Truth-Conditional Semantics focuses on the structure of truth theories, while Formal Semantics builds explicit model-theoretic interpretations. Both were influenced by Logical Positivism's emphasis on formal rigor, though they dropped the verificationist criterion. Today, Formal Semantics is a thriving interdisciplinary field at the intersection of linguistics and philosophy.
H. P. Grice introduced a framework that carved out a distinct domain for pragmatics. Gricean Pragmatics distinguishes what a sentence literally says (its truth-conditional content) from what a speaker means by uttering it. Grice's theory of conversational implicature explains how hearers infer intended meanings that go beyond literal content, guided by the Cooperative Principle and maxims of conversation (quantity, quality, relation, manner). This framework does not compete with Truth-Conditional Semantics; it complements it. Semantics handles the conventional, truth-conditional core; pragmatics handles the context-dependent inferences. Gricean Pragmatics remains a central framework in philosophy of language and has spawned Neo-Gricean and Relevance Theory variants. It coexists with Speech Act Theory, though the two focus on different aspects of utterance meaning: Grice on implied content, Speech Act on illocutionary force.
The Causal Theory of Reference led to Externalist Semantics, the view that the meanings of many expressions are partly determined by factors outside the speaker's mind. Putnam's Twin Earth thought experiment argued that "water" refers to H₂O even if a speaker cannot distinguish it from XYZ; meaning is not in the head. Tyler Burge extended externalism to social environment: the meaning of a word like "arthritis" depends on the expert community's usage. Internalist Semantics resists this conclusion. Internalists argue that there is a level of "narrow content" that is determined solely by the speaker's internal mental states, independent of the environment. Two-factor theories attempt to combine narrow content (internal role) with wide content (external reference). This debate remains the central fault line in contemporary philosophy of language. Externalism has become the dominant view for reference and natural kinds, but Internalism continues to attract defenders who argue that mental causation and psychological explanation require narrow content.
Today, several frameworks remain active as partly independent research traditions. Formal Semantics is the dominant method in linguistic semantics, providing precise models of compositionality and quantification. Gricean Pragmatics is the standard framework for explaining how context enriches literal meaning. Speech Act Theory has evolved into a specialized subfield, especially in social and political philosophy. The Externalist–Internalist debate structures much of the metaphysics of meaning. What these frameworks agree on is that meaning is a multi-layered phenomenon: truth-conditional content, speaker meaning, illocutionary force, and reference all require separate but interacting theories. What they disagree on is which layer is fundamental. Externalists and Formal Semanticists tend to privilege truth-conditional content; Griceans and Speech Act theorists emphasize communicative practice. The original tension between truth and use has not been resolved—it has been refined into a productive division of labor.