Loading field map
How do words hook onto the world, and what makes a sentence true? These two questions have driven the philosophy of language for over a century. The answers have shaped everything from logic to metaphysics. At stake is whether meaning is determined by mental content, by causal relations to objects, or by the conditions under which sentences are true. The frameworks that have emerged often compete, sometimes complement each other, and occasionally transform the very terms of the debate.
Modern theorizing about reference begins with Gottlob Frege's 1892 distinction between sense and reference. Frege argued that a term like "the morning star" refers to Venus not directly, but through a sense—a mode of presentation that determines the referent. This view, later called Descriptivism, held that every referring expression is associated with a description that uniquely identifies its object. Bertrand Russell and later Ludwig Wittgenstein (in his early work) developed this into a full theory: the meaning of a name is equivalent to a definite description. Descriptivism dominated from the 1890s through the 1960s. It treated reference as a matter of descriptive content in the speaker's mind, making it a natural ally of Internalist Semantics, which held that meaning is determined by mental states internal to the speaker.
While Descriptivism addressed reference, philosophers also debated the nature of truth. Correspondence Theory of Truth—traceable to Aristotle and revived in the early twentieth century—holds that a proposition is true when it matches an objective state of affairs in the world. Bertrand Russell and G. E. Moore championed this view. In contrast, Coherence Theory of Truth, defended by idealists like F. H. Bradley and later by some logical positivists, claims that truth consists in the mutual coherence of beliefs within a system. For coherence theorists, there is no independent reality to match; truth is a property of a well-ordered set of propositions.
Coherence Theory lost influence after the 1960s, largely because it struggled to explain how a coherent system could be completely disconnected from the world—a charge pressed by Russell and others. Correspondence Theory did not absorb it so much as outlast it, remaining the default position in metaphysics and semantics. Yet the tension between the two views never fully disappeared; it resurfaced in debates about scientific realism and anti-realism.
Internalist Semantics (1900–1970) was not a single doctrine but a broad assumption shared by Descriptivism and early truth theories: that meaning and reference are fixed by factors inside the speaker's head—mental representations, sensory experiences, or internal descriptions. This view fit naturally with the idea that understanding a sentence requires grasping its truth conditions, which were themselves determined by internal mental content. Internalism was the background against which later externalist challenges would define themselves.
The 1970s brought a decisive break. Saul Kripke's Naming and Necessity (1970/1980) and Hilary Putnam's "The Meaning of 'Meaning'" (1975) argued that reference cannot be captured by descriptions. Kripke proposed the Causal Theory of Reference: a name refers to an object through a causal-historical chain linking the speaker back to an initial baptism, not through any description the speaker associates with the name. Putnam's Twin Earth thought experiment showed that two speakers could be mentally identical yet refer to different substances (water vs. XYZ) because their environments differ. This argument directly undermined Internalist Semantics and Descriptivism.
Externalist Semantics (1970–present) generalizes the insight: meaning and reference depend on factors external to the speaker, such as the natural environment, social linguistic practices, and causal history. Externalism does not replace Truth-Conditional Semantics; rather, it provides a theory of how reference is fixed, which Truth-Conditional Semantics can then use to assign truth conditions. The two frameworks operate at different levels: Externalism explains the grounding of reference, while Truth-Conditional Semantics explains the compositional structure of meaning.
Truth-Conditional Semantics (1967–present), developed by Donald Davidson and inspired by Alfred Tarski's theory of truth, holds that the meaning of a sentence is given by its truth conditions. To know the meaning of a sentence is to know under what conditions it would be true. This program provided a rigorous, compositional account of meaning: the truth conditions of complex sentences are built from the semantic values of their parts. Davidson's project was philosophical, not technical.
Formal Semantics (1970–present), pioneered by Richard Montague, turned Truth-Conditional Semantics into a precise mathematical framework using model theory and possible worlds. Formal Semantics is the methodological implementation of the truth-conditional program. It treats natural language as a formal system amenable to logical analysis. While Truth-Conditional Semantics provides the theoretical commitment—meaning is truth conditions—Formal Semantics provides the tools to compute those conditions systematically. The two are deeply intertwined; most work in formal semantics today assumes a truth-conditional core.
Deflationary Theory of Truth (1990–present) challenges the idea that truth is a substantive property requiring explanation. Inspired by Frank Ramsey and later developed by Paul Horwich and others, deflationism holds that the predicate "true" serves only a logical or expressive function—for example, to endorse a proposition without repeating it ("What she said is true"). Truth is not a deep feature of reality; the concept is exhausted by the equivalence schema: "p" is true if and only if p. Deflationism does not directly conflict with Truth-Conditional Semantics, but it undermines the idea that truth conditions are explanatory. If truth is deflated, then saying that meaning is truth conditions may be trivial rather than illuminating. This tension remains unresolved: many formal semanticists continue to use truth conditions as a theoretical tool while deflationists argue that no substantive theory of truth is needed.
Contextualism about Truth (2000–present) argues that the truth conditions of sentences vary with the context of utterance in ways that go beyond standard indexicals like "I" or "now." For example, whether "It is raining" is true depends not only on the time of utterance but on the location supplied by context. Contextualists extend this to knowledge claims, moral statements, and aesthetic judgments: the truth of "She knows that p" may depend on the speaker's contextually determined standards.
Relativism about Truth (2000–present) goes further. It holds that the truth of a proposition can vary across perspectives or assessment contexts, not just utterance contexts. For instance, a relativist about taste might say that "Liquorice is tasty" is true relative to one person's standard and false relative to another's, with no context-independent fact of the matter. Relativism differs from Contextualism in locating the variability in the assessment of truth rather than in the content of the utterance. Both frameworks challenge the idea that truth conditions are fixed and universal, but they offer different diagnoses: Contextualism enriches the content, while Relativism relativizes truth itself.
Today, several frameworks remain active, each with a distinct role. Correspondence Theory of Truth is the default in metaphysics and scientific realism, though it faces competition from deflationism. Truth-Conditional Semantics and Formal Semantics dominate the study of meaning in linguistics and philosophy of language, providing the standard toolkit for compositional analysis. Causal Theory of Reference and Externalist Semantics are widely accepted for proper names and natural kind terms, though debates continue about how to extend them to abstract objects and fictional entities. Deflationary Theory of Truth is a major alternative to correspondence, especially among philosophers who want to avoid metaphysical commitments. Contextualism and Relativism are active research programs in epistemology, ethics, and aesthetics, but they remain minority positions in core semantics.
The leading frameworks agree that meaning involves reference and truth conditions, and that a compositional theory is desirable. They disagree on whether truth is a substantive property, whether reference is fixed internally or externally, and whether context can shift truth conditions in ways that require relativizing truth itself. These disagreements are not signs of crisis but of a healthy field where different frameworks address different aspects of the same deep puzzles.