Loading field map
Philosophy of language is the branch of philosophy that investigates the nature of meaning, reference, and communication. It asks what it is for a sound or mark to mean something, how words connect to the world, how speakers manage to convey thoughts to one another, and how language relates to thought, truth, and reality. While it shares concerns with linguistics—especially semantics and pragmatics—it differs in aiming not primarily at describing how languages work, but at understanding the philosophical puzzles that language raises: what meaning itself is, whether translation captures something determinate, or how words can be about things that do not exist.
The field is not a single doctrine but a cluster of interlocking research programmes. Its central questions have remained strikingly stable since the late nineteenth century, even as the methods and answers have shifted. The organizing fault lines run between theories that locate meaning in the mind, theories that locate it in public conventions, and theories that locate it in the way language connects to the world. Most philosophers of language combine elements of more than one, but the tensions between these orientations structure the field.
Three questions anchor the discipline. First, what is meaning? A sentence like "Snow is white" has meaning, but it is notoriously difficult to say what that meaning consists in. Is it a mental image in the speaker's head? A set of conditions under which the sentence is true? A pattern of use in a community? A mental representation in a language of thought? Each answer has consequences for how we understand communication, translation, and the possibility of shared understanding.
Second, how do words refer? Proper names like "Aristotle" seem to pick out a specific individual; natural kind terms like "gold" seem to pick out a kind of stuff; definite descriptions like "the current king of France" seem to pick out nothing at all, yet we understand them. Explaining how reference works—whether through descriptions speakers associate with terms, or through causal chains linking a name to its bearer—has been one of the field's most productive debates.
Third, what is the relation between language and thought? Does language merely express pre-existing thoughts, or does it shape or even constitute them? Can there be thought without language? These questions connect philosophy of language to philosophy of mind, but the angle here is distinct: the focus is on what linguistic meaning and communication reveal about the mind, rather than on the mind generally.
Underlying these questions is a deeper methodological issue: whether philosophical progress on language is best made by analyzing ordinary language carefully, by building formal models, or by looking to empirical linguistics and cognitive science. Disagreement about this methodological question has often been as sharp as disagreement about the substantive answers.
Philosophy of language as a self-conscious discipline emerged in the late nineteenth and early twentieth centuries, largely but not exclusively in the German- and English-speaking world. Its immediate stimulus was the rise of modern logic. Gottlob Frege, a German mathematician and logician, confronted the question of how mathematical statements can be both meaningful and about something. His distinction between sense and reference (Sinn and Bedeutung) became foundational. A term like "the morning star" and "the evening star" refer to the same object (the planet Venus) but have different senses—different modes of presentation of that object. This distinction explained how identity statements can be informative: "The morning star is the evening star" teaches us something because it connects two senses, even though the reference is the same.
Frege also insisted on a principle of compositionality: the meaning of a complex expression is determined by the meanings of its parts and the way they are combined. This may seem obvious, but it carries a strong commitment: meaning must be a systematic, rule-governed affair, not a collection of arbitrary associations. The principle remains central to most work in the field, whether formal or not.
Around the same time, Bertrand Russell in Britain developed a theory of definite descriptions—phrases of the form "the so-and-so." His theory analyzed such phrases as complex claims that do not themselves refer to an object. "The present king of France is bald" is not a statement about a nonexistent king, but a conjunction of three claims: there is someone who is king of France; there is only one such person; and that person is bald. Because the first claim is false, the whole sentence is false, not meaningless or neither true nor false. This analysis aimed to solve ancient puzzles about how we can meaningfully speak about things that do not exist. It also illustrated a general strategy: paraphrase problematic sentences into a logical form that makes their truth conditions explicit.
The early analytic movement was shaped by a further conviction, associated with the logical empiricists of the Vienna Circle in the 1920s and 1930s. They held that many traditional philosophical problems—about God, ethics, or ultimate reality—were not false but literally meaningless, because their sentences could not be verified by experience. This verificationist program made the theory of meaning central to philosophy itself: to show that a sentence was meaningful was to show how it could be verified. While the verificationist criterion of meaningfulness later collapsed under its own weight—it could not verify its own claim about verification—the idea that philosophy should begin with an account of meaning proved durable.
By mid-century, a distinct reaction had formed, centered at Oxford. Philosophers such as J. L. Austin and the later Ludwig Wittgenstein argued that the logical-form approach distorted language by treating it as if its only serious function were to describe facts. Wittgenstein, whose early Tractatus Logico-Philosophicus (1921) had itself advanced a theory of meaning as picturing facts, later repudiated that view. In his posthumous Philosophical Investigations (1953), he argued that meaning is use: to understand a word is to understand its role in the countless "language games" of ordinary life—promising, joking, asking, commanding, greeting. The search for a single underlying essence of meaning, he suggested, was a philosophical illusion.
Austin developed this orientation into a systematic account of speech acts. He observed that many utterances do not describe anything at all but perform actions. "I promise to come" does not report a promise; it makes one. "I name this ship the Queen Elizabeth" does not describe a naming; it performs it. Austin distinguished three levels within any utterance: the locutionary act (producing the words with their conventional meaning), the illocutionary act (what is done in saying them, such as promising or warning), and the perlocutionary act (what is achieved by saying them, such as persuading or alarming). His student John Searle developed this into a detailed theory of illocutionary acts and, later, of language as a form of rule-governed social behavior.
The ordinary language school did not reject the idea that sentences have truth conditions, but it insisted that truth is only one function among many, and that analyzing all language on the model of descriptive statements distorts what speakers are doing. Its legacy survives in a permanent caution about the gap between formal models and actual speech, and in the now-universal recognition that pragmatics—the study of what speakers do with words beyond their literal content—is an essential part of the philosophy of language.
The formal programme did not disappear with the ordinary language challenge. In the 1960s and 1970s it returned with renewed sophistication, led by Saul Kripke, Hilary Putnam, and David Kaplan in the United States. Their work, often called the New Theory of Reference, challenged assumptions that both Frege and Russell had shared, and that the ordinary language tradition had not directly questioned.
Kripke's Naming and Necessity (1970, delivered as lectures) attacked the descriptivist view of proper names: the idea that a name's meaning is given by a cluster of descriptions users associate with it. Kripke argued that we can use a name to refer to something even if every description we associate with it is false of that thing. "Aristotle" would still refer to Aristotle even if, as it happens, he was not the teacher of Alexander, not the author of the Nicomachean Ethics, and so on. Instead, Kripke proposed a causal-historical theory: a name is introduced in a "baptism" and passed from speaker to speaker through a chain of intentions to use the name as one's teacher did. What fixes the reference is the chain, not any body of belief. This view is now widely held, though some philosophers maintain a hybrid position in which descriptions play a subsidiary role.
Putnam extended the argument to natural kind terms like "water," "gold," and "tiger." He argued that the reference of such terms is fixed not by the descriptions speakers associate with them (e.g., "the clear drinkable liquid in rivers") but by the underlying nature of the stuff itself—whatever scientists discover it to be. His famous thought experiment of Twin Earth imagines a planet identical to Earth except that the liquid called "water" there has the same observable properties but the chemical formula XYZ, not H₂O. Putnam argued that "water" on Twin Earth would mean something different, because the reference is fixed by the actual substance, not by what speakers think about it. The slogan he derived—"meaning just ain't in the head"—captures the view that meaning depends on the world, not merely on mental states.
This externalist turn had profound consequences. It suggested that the content of our thoughts depends partly on facts about the environment we inhabit, a thesis that philosophers of mind later absorbed as content externalism. It also reopened questions about necessary and a posteriori truths. Kripke argued that "Water is H₂O" is necessary—true in all possible worlds, because water is H₂O in any world where it exists—but also a posteriori, discoverable only through empirical investigation. This combination had been assumed impossible since Kant.
David Kaplan meanwhile developed a sophisticated account of indexicals—words like "I," "here," "now," and "that." His distinction between character (a rule that fixes the referent relative to a context, such as "I" always referring to the speaker) and content (what is said on a particular occasion) showed how a single word could have a stable meaning while contributing different referents on different occasions. This work bridged the gap between formal semantics and the pragmatic insights of the ordinary language tradition.
By the late twentieth century, it had become impossible to treat meaning as solely a matter of truth conditions. The work of Paul Grice, much of it written in the 1950s but only fully published in 1989, provided the framework for integrating pragmatics into the philosophy of language. Grice distinguished between sentence meaning (what the words literally mean) and speaker meaning (what the speaker intends to communicate). He argued that communication works because hearers can infer speaker meaning from sentence meaning together with assumptions about the speaker's rationality and cooperativeness.
Grice's principle of cooperation and his maxims of conversation (be informative, be truthful, be relevant, be clear) were not rules that speakers always obey, but default assumptions that hearers use to interpret violations. If someone says "It's cold in here" in a room with an open window, the literal sentence states a fact; but the speaker likely means it as a request to close the window. The hearer arrives at this interpretation by assuming the speaker is being cooperative and relevant. Grice called such inferences conversational implicatures—meanings that are communicated without being literally said.
This framework had a double effect. On the one hand, it preserved the idea that literal sentence meaning can be studied formally and compositionally, as the truth-conditional programme held. On the other hand, it showed that most actual communication involves far more than literal meaning. Later philosophers and linguists, including Stephen Neale and Dan Sperber with Deirdre Wilson, developed Grice's insights into full-scale theories of communication, sometimes arguing that the gap between literal meaning and what is communicated is so large that a purely formal semantics could never by itself explain human communication.
A further shift in the late twentieth century brought philosophy of language into close contact with the cognitive sciences. Chomsky's generative linguistics, with its theory of universal grammar and an innate language faculty, raised the question of whether philosophical accounts of meaning should be constrained by empirical findings about how language is actually processed. Some philosophers, such as Jerry Fodor, argued that thought itself is a kind of language—a "language of thought" or Mentalese—and that spoken language expresses that inner symbolism. On this view, the study of meaning in natural language is downstream of the study of mental representation.
Others, including Donald Davidson, took a different naturalistic route. Davidson argued that we can attribute beliefs, desires, and meanings to a speaker only by applying a principle of charity: interpreting the speaker as largely rational and largely in agreement with us about the facts. Meaning, on this view, is not something we can study apart from interpreting speakers as rational agents. Davidson also developed a program of truth-conditional semantics inspired by Alfred Tarski's work on truth, claiming that a theory of meaning for a language could be given by a recursive theory that specifies, for every sentence, the conditions under which it is true. This approach returned to Frege's idea of compositionality but grounded it in formal logic and in Davidson's holistic account of interpretation.
The naturalistic turn also produced the deflationary theory of truth, associated with W. V. O. Quine and later developed by Paul Horwich. On this view, truth is not a substantive property that sentences or propositions possess; to say that "Snow is white" is true is just to endorse the sentence "Snow is white." The entire usefulness of the predicate "true" lies in its enabling us to talk about sentences we cannot or will not explicitly utter (e.g., "Everything the professor said is true"). If this theory is right, the connection between meaning and truth is far less exalted than Frege and Russell supposed.
Present-day philosophy of language is pluralistic, and the field's boundaries have blurred. A number of overlapping research programmes coexist, each with its own journals, conferences, and practitioner networks.
Formal semantics, conducted jointly by philosophers and linguists, continues the Fregean program of giving compositional, truth-conditional theories of natural language. It has become highly technical, using tools from model theory and type theory. It now extends beyond declarative sentences to conditionals, modals ("must," "may"), questions, and gradable adjectives ("tall," "fast"). This tradition has been remarkably successful at explaining systematic patterns in how speakers evaluate sentences for truth and appropriateness.
Semantic externalism, in the tradition of Kripke and Putnam, remains the dominant view about proper names and natural kind terms, though it has been refined and qualified. Philosophers now debate whether externalism applies to all thought content, whether it is compatible with privileged access to one's own thoughts, and whether the Twin Earth argument really proves what it seems to prove.
The expressive and non-descriptive traditions have grown. Contemporary philosophers such as Paul Grice's successors and the semanticists of "expressivism" in ethics and modality argue that at least some sentences do not primarily function to describe how the world is but to express attitudes, stances, or plans. Expressivism about moral language, developed by Simon Blackburn and Allan Gibbard, holds that "Murder is wrong" expresses a disapproval or a norm, not a fact. This view has been extended to conditionals and probability statements. The recognition that language has many functions, which the ordinary language school insisted on, has become mainstream, though the specific theories differ.
The cognitive and experimental turn has made philosophy of language a partially empirical discipline. Philosophers now draw on psycholinguistics to test claims about compositionality, reference, and ambiguity. Some conduct experiments themselves—this is the field of experimental philosophy—testing whether ordinary speakers' intuitions about reference or truth conditions match the theories philosophers have proposed. Results have been mixed and sometimes contested, but the practice of treating these questions as open to empirical evidence is now established.
Contextualism vs. invariantism is one of the liveliest current debates. Contextualists argue that the truth conditions of sentences containing words like "know," "flat," or "every" shift with the context of utterance. A sentence like "John knows that the bank is open" may be true in an ordinary context but false in a skeptical context where the standards for knowledge are raised. Invariantists—many of them inspired by the New Theory of Reference—argue that the content of such words is fixed, and that apparent variation is best explained by pragmatics rather than semantics. This debate connects directly to epistemology and ethics, and shows how deeply philosophy of language cuts across other regions of philosophy.
Metasemantics—the study of what grounds meaning facts—has emerged as a central topic. If "water" means H₂O, what makes that true? In virtue of what does the word have the meaning it does? Candidates include conventions, dispositions of speakers, facts about the environment, and normative practices. This question pushes philosophy of language toward questions in the philosophy of mind, social philosophy, and metaphysics, and it is now common for philosophers to argue that an adequate theory of meaning must be integrated with a theory of mental representation and of social norms.
A final notable feature of the current landscape is the entry of philosophers working outside the Anglo-American analytic tradition into the debates. The philosophy of language has historically been dominated by English-speaking philosophers, and the canon of Frege, Russell, Wittgenstein, and their successors reflects that history. But there is growing attention to questions about multilingualism, linguistic injustice, and the relation between colonial languages and indigenous ones. Philosophers such as those working in African, Latin American, and South Asian traditions have asked how the standard theories of reference and meaning fare when applied to languages with very different structures and social institutions than English. These developments are recent and not yet fully integrated into the central debates, but they represent a genuine broadening of the field's self-conception.
Philosophy of language may seem esoteric, but its stakes are concrete. If meaning is determined by use, then there is no single correct way to describe reality, only better and worse ways relative to purposes. If externalism is true, then what we think depends on what the world is like in ways we may not be able to discern from the armchair. If context modulates truth conditions, then disputes about knowledge, morality, or vagueness may be less about the facts and more about how we choose to speak. The field does not merely analyze words; it determines what philosophical questions amount to in the first place. That is why philosophers from every other subfield—ethics, metaphysics, epistemology, philosophy of science—repeatedly find themselves arguing about language. The philosophy of language remains what Quine once called the "first philosophy" of the analytic tradition, not because it is the only important discipline, but because it is where the tools of analysis are formed.