Loading field map
Context and indexicality is the branch of philosophy of language that studies how the meaning of an utterance depends on the situation in which it is produced. Its central subject is the class of expressions—words like "I," "here," "now," "this," "that," "today," and "you"—whose reference shifts systematically with the circumstances of use. When two people say "I am tired," each says something true or false depending on who is speaking, even though the sentence type is identical. The subfield asks how this is possible, what kind of meaning such expressions have, and what the broader consequences are for our understanding of language, thought, and communication.
The philosophical puzzle begins with a simple observation: many sentences do not have a single, fixed truth value. The sentence "It is raining here now" is true when uttered in London on a wet Tuesday and false when uttered in Cairo on a dry Wednesday. Yet the words are the same. This variability was noticed by ancient and medieval logicians, but it became a central problem in the late nineteenth and early twentieth centuries, when philosophers and logicians sought to model language as a formal system. The German logician Gottlob Frege distinguished between the sense of an expression (its mode of presentation) and its reference (the object it picks out). For Frege, indexicals posed a special challenge: the sense of "I" seems to be the same for every speaker, yet its reference changes with the speaker. The American philosopher C. S. Peirce, working independently, introduced the term "index" for signs that point to their objects through a causal or contextual connection, such as a pointing finger or a weathervane. But the systematic philosophical treatment of indexicals did not emerge until the mid-twentieth century.
The decisive modern formulation came from the Israeli-American philosopher Yehoshua Bar-Hillel, who in the 1950s argued that indexical expressions are "pragmatic" rather than purely semantic: their contribution to truth conditions cannot be specified without reference to the context of use. This set the stage for the two major research programmes that have organized the field ever since: the formal semantic approach, which tries to build a rigorous theory of how context determines content, and the pragmatic approach, which treats indexicality as one part of a broader account of how speakers use language to communicate.
The dominant approach in the late twentieth century, associated with the American philosopher David Kaplan, treats indexicals as expressions whose meaning is a character—a rule that maps contexts to contents. In Kaplan's framework, a context is a package of parameters: the agent (speaker), the time, the place, and the possible world of the utterance. The character of "I" is the rule "the speaker of the context"; the character of "now" is "the time of the context"; the character of "here" is "the place of the context." The content of an indexical in a given context is the object picked out by that rule. Thus the content of "I" when I speak is me; when you speak, it is you. The sentence "I am hungry" has a different content in each context, but the same character—the same rule for determining content.
This framework solved several problems at once. It explained how two utterances of the same sentence type can have different truth values without any ambiguity: the sentence type has a fixed character, but the content varies with context. It also explained how we can understand an utterance we have never heard before: we know the character of the indexicals, and we can compute the content from the context. The framework was refined by the American philosopher John Perry, who distinguished between the content of an utterance (what is said) and the character (the rule). Perry also introduced the notion of "token-reflexive" rules: the character of "I" is not just "the speaker" but "the speaker of this very token," which allows the theory to handle cases where the same speaker utters the same sentence at different times.
A crucial distinction within this programme is between pure indexicals and demonstratives. Pure indexicals like "I," "now," and "today" have their reference determined entirely by the context parameters: no act of pointing or mental intention is needed. Demonstratives like "this" and "that" require an additional element: the speaker's demonstration or intention. The philosopher David Kaplan, who developed the most detailed version of the formal theory, argued that the reference of a demonstrative is determined by the speaker's directing intention, not by the physical gesture alone. This distinction matters because it shows that context is not a simple list of objective features; it includes the speaker's mental state.
The formal programme has been highly successful, but it has limits. It assumes that the relevant context parameters are fixed and enumerable, but natural language is more flexible. For example, "here" can refer to a room, a city, a country, or a planet, depending on the speaker's intention. The formal theory can handle this by treating "here" as a demonstrative-like expression whose reference is determined by the speaker's intended location, but this blurs the line between pure indexicals and demonstratives. A deeper problem is the so-called context-sensitivity of context: the parameters that matter for an utterance are not given by the physical situation alone but are themselves determined by the speaker's intentions and the conversational purposes. The philosopher John Perry later argued that the formal framework needs to be supplemented by a theory of reflexive content—the thought that the utterance itself is about the speaker, the time, and the place—to explain how indexicals work in communication.
A second major approach, rooted in the philosophy of language of the late twentieth century, treats indexicality as a special case of the general phenomenon of speaker meaning. The American philosopher Paul Grice distinguished between what a sentence means (its conventional meaning) and what a speaker means by uttering it (the speaker's communicative intention). On this view, indexicals are not a separate category of expressions with a special semantics; they are simply expressions whose conventional meaning is too thin to determine a complete content, so the hearer must use contextual clues to infer the speaker's intention.
This approach is associated with the work of the philosopher John Searle and, more recently, with the "contextualist" tradition in the philosophy of language. Contextualists argue that the formal programme overestimates the role of fixed rules and underestimates the role of general reasoning about the speaker's intentions. For example, the sentence "I am ready" does not have a complete content until we know what the speaker is ready for—ready to leave, ready to eat, ready to give a speech. The context supplies this information, but not through a rule that maps a parameter to a value. Instead, the interpreter uses background knowledge and conversational clues to infer the intended completion.
The pragmatic programme has been influential in drawing attention to the underdetermination of content by the sentence. It argues that even sentences without indexicals, such as "The table is covered with books," are not fully determinate: "the table" might refer to a specific table, but the sentence does not say which one. The hearer must use the context to pick out the referent. On this view, indexicality is not a special feature of a few words but a general feature of all language: every utterance requires the hearer to fill in gaps using context.
The two programmes are not mutually exclusive. Many philosophers accept the formal framework for the core cases of pure indexicals and the pragmatic framework for the broader phenomenon of context-sensitivity. The disagreement is about the boundary. Formalists argue that the core cases are governed by rules, while pragmatists argue that even the core cases require intention-based reasoning. The debate is partly empirical—about how speakers actually process language—and partly conceptual—about what counts as the "meaning" of an expression.
A central theme in the subfield is the special status of the first-person indexical "I." The philosopher Hector-Neri Castañeda argued that "I" has a unique feature: it cannot be replaced by any description without loss of cognitive content. If I say "I am hungry," I express a thought that is different from the thought expressed by "The person who is writing this article is hungry," even if the two refer to the same person. The first thought is self-locating: it tells me something about my own position in the world. The second thought is about a person who happens to be me, but it does not have the same cognitive significance.
This observation has led to a rich discussion of self-locating knowledge. The philosopher John Perry used the example of a shopper who sees a trail of sugar on the floor and thinks, "Someone is leaking sugar." The shopper does not know that he himself is the leaker until he realizes, "I am the one leaking sugar." The second thought is not just a new fact about the world; it is a new fact about the shopper's own place in the world. Perry argued that indexicals are essential for action: a person can know all the objective facts about the world and still not know what to do unless they know which person they are, where they are, and when it is.
This "essential indexical" thesis has been debated. Some philosophers, such as David Lewis, have tried to reduce indexical knowledge to knowledge of a proposition about a centered world—a world with a designated speaker, time, and place. On this view, the shopper's knowledge is not a special kind of knowledge but a proposition about a centered world. Others, such as Perry, have argued that this reduction fails because it does not capture the action-guiding role of the indexical. The debate remains unresolved, but it has important implications for the philosophy of mind, the theory of action, and the nature of self-knowledge.
The formal framework has been most successful with "I," "now," and "here," but it faces a special difficulty with the temporal and spatial indexicals. The character of "now" is "the time of the context," but the context of an utterance is a moment in time. However, the utterance itself takes time to produce, and the hearer may process it at a different time. The sentence "It is now 3:00" is true if uttered at 3:00, but the hearer may hear it at 3:01. Does the sentence refer to the time of the utterance or the time of the hearing? The formal framework assumes a single context, but real communication involves multiple times.
A related problem is the "here" of the utterance. The sentence "It is hot here" is true if uttered in a hot room, but the hearer may be in a different room. The formal framework treats "here" as the place of the context, but the context of the utterance is the speaker's location, not the hearer's. This is not a problem for the theory, but it shows that the context is not a neutral, objective situation but a set of parameters centered on the speaker.
A more serious challenge is the phenomenon of deferred reference. When a person says "I am parked out back," they do not mean that they are parked but that their car is parked. The indexical "I" refers to the speaker, but the predicate "parked" applies to the speaker's car. This is not a case of indexicality in the strict sense but a case of metonymy—the use of a word to refer to something related to its literal referent. The formal framework can handle this by treating the utterance as a non-literal use of the sentence, but it shows that the boundary between literal indexicality and pragmatic inference is not sharp.
The field today is characterized by a broad consensus on the core phenomena and a continuing debate about the best theoretical framework. The formal programme, in the tradition of Kaplan, remains the standard starting point for most philosophers. It has been extended to handle more complex cases, such as indexicals in reported speech ("John said that I am happy" can refer to the speaker of the report or the speaker of the original utterance) and indexicals in written texts (where the context is not a single utterance but a text with a complex history). The pragmatic programme has been developed into a general theory of context-sensitivity that includes not only indexicals but also gradable adjectives ("tall"), quantifiers ("many"), and epistemic modals ("might"). This broader debate is often called the "semantics-pragmatics" debate, and it is one of the most active areas in contemporary philosophy of language.
A notable recent development is the study of indexicality in non-assertive speech acts. Questions, commands, and promises also contain indexicals, and the formal framework has been extended to handle them. For example, the command "Come here" has a different content depending on who is being commanded and where the speaker is. The theory of speech acts has been integrated with the theory of indexicals, but the integration is not complete.
Another area of active research is the relationship between indexicality and context in the sense of common ground. The philosopher Robert Stalnaker has argued that the context of an utterance is not just a set of parameters but a set of propositions that the speaker and hearer mutually accept. Indexicals are used to update this common ground. On this view, the context is not a fixed background but a dynamic entity that changes as the conversation proceeds. This "dynamic" approach has been influential in formal semantics, but it is not a rival to the Kaplan framework; it is a complement that explains how the context is built and maintained.
The field also connects to the philosophy of mind through the problem of indexical thought. The question of how a person can think of themselves as themselves, and how they can think of a place as "here," is a problem for the philosophy of mind as much as for the philosophy of language. The philosopher Gareth Evans argued that indexical thoughts are a fundamental form of thought that cannot be reduced to descriptive thought. This view has been influential in the philosophy of mind, but it remains controversial.
Finally, the field has been enriched by cross-linguistic research. Not all languages have the same indexical system. Some languages have a distinction between "we" that includes the hearer and "we" that excludes the hearer. Some languages have a "here" that distinguishes between near the speaker, near the hearer, and far from both. These differences show that the indexical system is not a universal logical structure but a set of conventions that vary across languages. The philosophical framework must be flexible enough to accommodate this variation.
The durable present landscape is thus one of a mature field with a well-established core and a set of open frontiers. The formal framework of Kaplan is the standard reference point, but it is not the final word. The pragmatic approach has shown that the formal framework is not sufficient for all cases of context-sensitivity. The debate between the two is not a dispute about the facts but a dispute about the best way to model the facts. The field continues to be a central part of the philosophy of language, and its results have implications for logic, linguistics, cognitive science, and the philosophy of mind.