Semantics and Pragmatics

The Explanatory Problem

A sentence is repeatable; an utterance occurs. A competent speaker understands the English words in “It is ready,” yet the proposition asserted may concern a report, meal, experiment, or evacuation vehicle. The very same expression can be reassuring, impatient, deceptive, interrogative in its practical force, or the signal for a coordinated action. A satisfactory theory of linguistic understanding must explain the continuity that makes repetition possible and the situated interpretation that makes a particular event intelligible. The distinction between semantics and pragmatics is an attempt to distribute these explanatory responsibilities.

The field is best approached through three questions. First, what does knowledge of a language contribute before a particular occasion is specified? Second, which features of an occasion determine reference, propositional content, and the appropriate interpretation of what is explicitly conveyed? Third, what further actions, commitments, inferences, and alterations of conversational expectations follow from an utterance? These questions overlap in their evidence, but their answers can diverge. The same content may be asserted or quoted, asserted sincerely or ironically, or conveyed by an expression whose lexical meaning remains stable across contexts.

The terms semantics and pragmatics name research traditions as well as explanatory functions. Their disciplinary boundary has moved because the notion of “what is said” has received rival analyses. Some philosophers place a relatively austere, compositionally determined proposition between sentence meaning and the rich speech act; others argue that ordinary truth-evaluable claims already depend on unarticulated contextual enrichment (Bach, 1994; Cappelen & Lepore, 2005; Recanati, 2004). Their disagreement has consequences for logic, epistemology, moral discourse, testimony, and communication. A pragmatics that explains only conversational extras cannot perform the same work as one that also determines truth-conditional content. The monograph evaluates this disagreement rather than treating an inherited boundary as a definition.

What Semantics and Pragmatics Study

For a formal interpretation, semantics assigns expressions values relative to a model and, when necessary, parameters such as worlds, times, assignments, or contexts. The interpretation may yield a referent, a property, an intension, a proposition, or a rule for obtaining one. In natural-language analysis, semantic competence includes conventional lexical and grammatical constraints that enable novel combinations to be understood. A compositional theory explains complex expressions in terms of their parts and modes of combination; it leaves open whether these values are truth conditions, inferential roles, dynamic updates, or more than one kind of object.

Pragmatics studies language as used by agents in situations. Among its principal concerns are the recognition of communicative intentions, the determination of the targets of demonstrations, conversation-specific inference, presupposed information, speech acts, turn-taking, and the social consequences of an utterance. The distinction often called near-side versus far-side pragmatics separates contextual factors contributing to what is said from inferential consequences that proceed beyond it (Korta & Perry, 2024). The near side includes, at minimum, choosing an intended reading for an ambiguous expression and fixing an ordinary indexical's occasion of use. The far side includes many conversational implicatures, indirect requests, and strategic omissions.

“Context” accordingly has several technical senses. A narrow context provides speaker, place, time, and other parameters required by linguistic rules. A conversational context includes mutually recognizable commitments and the live question under discussion. A social context includes authority, roles, institutions, conventions, and histories of interaction. A cognitive context includes beliefs, salient alternatives, and attentional states. The contexts may coincide in a conversation, but a theory that silently moves between them produces false equivalences. One can use the narrow speaker parameter for “I” without settling whether the sentence contributes a proposal to a committee's public record.

The key methodological test for a proposed semantic–pragmatic division is explanatory independence: can the contribution assigned to a level be identified, varied, and tested without merely restating the total interpretation? When a “pragmatic” process is defined as whatever an extant semantic theory fails to predict, the division ceases to illuminate the phenomenon. Its value lies in specific mechanisms and their independently identifiable evidence.

Historical Sources Before the Modern Distinction

Ancient discussions of names, predication, propositions, and acts of discourse contain precursors of the modern questions, although they do not constitute modern semantics and pragmatics. Plato's Cratylus considers whether a name's correctness arises from nature or convention; the Sophist investigates the ability of statements to say something true or false. Aristotle's De Interpretatione distinguishes spoken sounds, written marks, and what they signify, then analyzes affirmation, negation, and the structure of assertoric discourse. The opening of De Interpretatione (16a3–8) becomes especially significant because it makes the relation between signs and what is thought a philosophical question. The texts' precise doctrines require their own study; their contribution here is an early recognition that signs, what they indicate, and truth-bearing combinations are related but separable objects.

The tradition of rhetoric adds the conditions of persuasive address, audiences, and practical effects. Aristotle's Rhetoric examines forms of persuasion and situated judgment; it studies considerations that later philosophers would allocate to speech acts and audience uptake. Yet mapping the modern boundary wholesale onto ancient rhetoric would conceal the different theoretical projects. Logical analysis and rhetoric were historically constituted by different questions about discourse, not by a settled two-department science of meaning.

The development of early modern logic, grammar, and theories of signs progressively sharpened the problem. By the nineteenth century, formal logic demanded a treatment of expressions able to explain valid inference with more precision than ordinary conversational intuition. The next decisive transformation came when an expression's contribution to truth and the cognitive significance of different ways of presenting an object were analyzed together. That transformation sets the conditions for Frege's systematic distinction between sense and reference. It also establishes the pressure that later pragmatics will encounter: a logically rigorous account of linguistic content must remain accountable to what communicators actually manage to say and do.

Medieval Signification, Supposition, and Occurrence

Medieval analyses of the properties of terms supply an important historical comparison to the modern semantic–pragmatic interface. The Latin distinction between significatio and suppositio articulates an enduring problem: a term has a relatively stable signifying capacity, while its occurrence in a proposition may stand for a specific kind of item. An account of supposition thus examines how an expression functions in a particular sentential arrangement. Medieval authors additionally investigated ampliatio (expansion of a term's range), restrictio (restriction), appellation, and relations involving expressions that refer back to antecedents. The terminology and classifications changed across centuries, and none maps exactly onto current theories of encoded meaning and contextual interpretation (Read, 2023).

William of Sherwood in the thirteenth century explicitly distinguishes signification, supposition, copulation, and appellation; Peter of Spain, Lambert of Auxerre, Walter Burley, and William of Ockham organized the subject differently. The dispute matters because different theories of what terms signify yield different classifications of their occurrences. A nominalist account of signification has different commitments about universals from a realist account, even when a particular sentence's immediate referential behavior is comparable. In Ockham's framework, for example, the modes in which an expression stands for things, concepts, or linguistic signs are bound up with the ontology of mental language. These positions should be reconstructed from their medieval contexts rather than interpreted as an anticipation of an already fixed modern doctrine.

The contrast between material and personal supposition is especially illuminating. An expression can be used to speak about the word itself, or it can function to speak about things of the relevant kind. In a sentence about the number of letters in a term, that term is mentioned; in an ordinary predicative statement, it is used to classify or pick out items. This distinction also arises in Tarski's twentieth-century treatment of object language and metalanguage, where attention to quotation and naming is essential to formulating truth conditions (Tarski, 1944, pp. 343–345). The connection is historical and conceptual without making the medieval vocabulary identical to modern semantic theory.

Ampliation and restriction show why a stable term can contribute differently depending on the proposition in which it appears. Modal, temporal, or intentional constructions may affect the range over which terms operate. The medieval discussion is thus more varied than an early theory of conversational implicature. It investigates logical validity, inference, sentence occurrence, and the conditions under which words signify or stand for entities. Its contribution to the present dossier is the demonstration that the relation between repeatable significance and particular occurrence was already a systematic philosophical problem long before the modern disciplinary labels existed.

Classical Indian Debates on Word and Sentence Meaning

Classical Indian philosophy supplies a historically independent set of debates concerning the relation between a word's meaning, a complete sentence's communicative unity, and knowledge acquired through linguistic testimony. Sanskrit philosophical traditions examined śabda (word or linguistic testimony, depending on context), vākya (sentence), and questions about whether understanding arises by assembling previously grasped word meanings or whether the sentence is understood as an organized whole. These issues overlap with modern compositionality and utterance understanding, but they arise within distinctive grammatical, epistemological, and scriptural disputes (Deshpande, 2020).

Bhartṛhari's Vākyapadīya offers a major philosophical treatment of language in which the unity of the linguistic whole has methodological importance. The concept of sphoṭa is interpreted within debates over the nature of the linguistic unit and the apprehension of meaning. A careless approximation to a modern “sentence gestalt” loses the theory's original metaphysical and grammatical context. For the present comparison, its relevant challenge concerns the assumption that understanding must proceed through a single sequence from independently completed word meanings to a completed sentence proposition. Whether such a sequence is indispensable is an issue that also matters for contemporary models of compositional interpretation.

Nyāya and Mīmāṃsā authors developed competing positions on verbal cognition, testimony, sentence interpretation, and the contribution of words to an integrated understanding. Traditions associated with the contrast between abhihitānvaya and anvitābhidhāna ask in different ways whether words first express meanings that are subsequently related or signify them as already syntactically connected. The distinction is relevant to the sequencing of semantic composition and practical interpretation; however, reducing it to a literal analogue of twentieth-century Gricean theory would be historically misleading.

The scholarly overview by Deshpande (2020) documents these disputes and the underlying Sanskrit corpus. The present chapter uses that research to establish the comparative landscape; it does not claim line-by-line verification of the complete original Vākyapadīya or medieval Sanskrit commentaries. The methodological gain is nevertheless substantial. The Western semantics–pragmatics distinction can be evaluated against an intellectual tradition that organized sentence meaning, word meaning, and epistemic authority according to different theoretical priorities. Comparative philosophy thereby widens the class of theories to be explained without erasing historically significant differences.

Sense, Reference, and the Fregean Inheritance

In “Über Sinn und Bedeutung,” Frege (1892, pp. 25–50) asks why an identity of the form a = b may be cognitively informative when a = a appears trivial. If two names designate the same object, their shared designation alone cannot explain this difference. He introduces Sinn, a mode of presentation, and Bedeutung, commonly rendered “reference,” for the object designated. His treatment of complete declarative sentences assigns them a thought as sense and a truth value as reference. The difference between how an object is given and which object is given becomes fundamental to later semantic theorizing.

The ordinary English word “meaning” is an unsafe automatic translation of Bedeutung: Frege's technical relation is more specific. An editor who translates Bedeutung consistently as “meaning” must still explain how its role differs from Sinn. The canonical article's discussion of identity and informative presentation is therefore cited in the German original (Frege, 1892), with the established Black translation identified separately in the references. The distinction's philosophical strength lies in avoiding the reduction of cognitively different expressions to an identical object.

Frege also examines indirect speech and oblique contexts: expressions embedded in reports of attitudes can behave differently from their ordinary referential occurrences. This creates an explanatory demand that extends beyond extensional substitution. A semantic theory needs to account for linguistic modes of presentation and the environment in which expressions occur. Later theories of intensionality, context, and discourse inherit this demand, though they organize its solution differently.

Frege is not a theorist of Gricean implicature avant la lettre. His project concerns the semantic and logical structures through which thoughts are expressed, while the modern pragmatics of inference asks what a speaker means by selecting a particular form in a particular exchange. Nevertheless, the cognitive difference between co-referential descriptions foreshadows a pragmatic question: why choose one informative formulation over another when the referent is constant? The resulting bridge is analytic rather than a claim of direct historical derivation.

Descriptions, Reference, and the Act of Referring

Russell's “On Denoting” (1905, pp. 479–493) analyzes descriptions as expressions whose logical contribution is captured in the proposition's quantified structure rather than by assuming an independently denoted object for every grammatical phrase. In the familiar analysis of “The present king of France is bald,” the descriptive form encodes existence, uniqueness, and predication. The proposition can be evaluated even when the description fails to pick out an existing king. This is a systematic response to apparent reference without a referent, and it makes the distinction between surface sentence form and logical form indispensable.

Strawson's “On Referring” (1950, pp. 320–344) challenges a picture in which a sentence's meaningful use is exhausted by Russellian truth-conditional analysis. He draws attention to the act of making a particular reference and to conditions under which an utterance is appropriately assessed. The sentence type, its use on an occasion, and the act performed on that occasion need separate consideration. Whether the king-of-France sentence is false, lacks truth value in a relevant use, or carries a failed presupposition is part of the controversy, and later semantic analyses of presupposition develop several distinct formal responses.

The stronger Russellian argument is that logical structure can secure clarity about entailment and apparent empty reference. The stronger Strawsonian objection is that speakers ordinarily treat successful identification as part of the practical background of an assertion. These answers target related but distinguishable questions. Russell specifies a propositional form. Strawson asks what conditions govern ordinary acts of reference and assessment. A theory capable of accommodating both will describe formal conditions for a descriptive construction and the pragmatic requirements that govern its use.

The resulting lesson is a recurring one. Formal truth conditions and the standards by which an utterance is evaluated may be integrated without being identical. The philosophical dispute can only be settled by identifying which property is to be explained: logical truth, availability of a referent, felicitous speech, or conversational commitment.

Signs, Logical Syntax, and the Institutionalization of Pragmatics

The twentieth-century semiotic vocabulary of syntax, semantics, and pragmatics became prominent through Charles W. Morris's Foundations of the Theory of Signs (1938). Morris distinguishes relations among signs, relations between signs and designata, and relations of signs to interpreters. The tripartition supplies a powerful taxonomy, but its application is broader than contemporary linguistic pragmatics. A relation involving interpreters need not yet be an account of calculable implicatures, speaker intentions, or the updating of common ground.

Carnap's Introduction to Semantics (1942) and Meaning and Necessity (1947) advance formal semantic methods with logical languages, designation, and intensional distinctions. The logical possibility of a systematic semantic interpretation is central to Carnap's program; it also forces the question of where empirical sign-users and occasions of interpretation enter. Morris's general semiotics and Carnap's logical semantics form a historical route toward the modern division, even though later theorists redefine its subject matter.

In mid-century ordinary-language philosophy the unit of analysis increasingly becomes an utterance situated in a practice. Wittgenstein's Philosophical Investigations (1953), particularly its early discussion of language-games and the diversity of uses of words, directs attention to meaning in the context of human activities. Austin's lectures, published as How to Do Things with Words (1962), further establish that uttering words can itself perform an action. These developments introduce new questions about correctness and social practice. They do not simply invalidate truth-conditional theory; they expand the inventory of what a theory of communication must explain.

A historically responsible reconstruction therefore recognizes several intersecting intellectual trajectories: formal logic and truth conditions, general semiotics, ordinary-language analysis, and theories of action. “Semantics versus pragmatics” becomes a focal debate when the products of these trajectories are asked to explain a single utterance.

Tarski, Davidson, and the Logic of Truth

Alfred Tarski's “The Semantic Conception of Truth and the Foundations of Semantics” (1944, pp. 341–376) establishes a rigorous perspective on truth in a specified language. An adequate truth definition must yield appropriate instances of the schema according to which a named sentence is true if and only if things are as the sentence says. Tarski's example concerning snow introduces a crucial distinction between mentioning the sentence on one side of the biconditional and using it on the other. He stresses that the schema itself is a condition of adequacy rather than a complete truth definition (1944, pp. 343–345). The construction requires attention to the structure of the object language and the language in which its truth predicate is formulated.

Tarski's argument can be reconstructed through a series of explanatory demands. A truth predicate is defined for sentences of a specified language. Its semantic adequacy is tested against correctly formed biconditionals. Its formal adequacy depends on the logical resources of the language of definition, and its construction must avoid paradoxes of semantic self-reference. These constraints explain why the relation between expression, interpretation, and truth can be specified in a rigorous theory without appealing to a conversational agent's psychological intentions. They also explain why the resulting formal definition alone supplies no full account of how a particular speaker uses a sentence.

Davidson's “Truth and Meaning” (1967, pp. 304–323) asks whether a suitably structured truth theory might serve as the core of a theory of meaning for a natural language. The proposal is attractive because a finite axiomatic system could account for indefinitely many interpreted sentences. Its philosophical ambition differs from the mere enumeration of one truth condition per sentence: a semantic theory must display the systematic relationships among expressions that make linguistic competence learnable and productive. The explanatory burden is correspondingly severe. Formally correct truth conditions could be assigned in unintended ways unless the theory is tied to interpretive evidence and to what speakers actually mean.

The step from Tarski to Davidson changes the role of truth theory. In Tarski's foundational treatment, the aim is to define truth for an appropriately specified language while preserving formal and material adequacy. In Davidson's philosophy of interpretation, the challenge is to understand how such formal resources contribute to an explanatory theory of a speaker's language. A Davidsonian account must therefore confront actual utterances, rationality, and interpretive charity, even if its compositional core is expressed by truth-theoretic principles.

This historical sequence clarifies a persistent semantic–pragmatic tension. Formal truth theories demand stable rules that generate correct compositional results; utterance interpretation demands procedures for connecting those rules to speakers and contexts. Their relation is not resolved by treating one as an appendix to the other. It must be explained through the constraints imposed by grammar, logical structure, empirical interpretation, and the conditions under which a sentence's content becomes publicly available.

Truth-Conditional and Compositional Semantics

The core advantage of formal semantics is systematicity. A finite grammar supports indefinitely many understandable expressions, including constructions the hearer has never encountered. Montague's “Universal Grammar” (1970, pp. 373–398) demonstrates a formal strategy for assigning interpretations to grammatical structures and treating natural language with tools of intensional logic. The later collection Formal Philosophy (1974) documents the breadth of this program. The important claim is methodological: grammatical combination and interpretation can be studied with explicit rules rather than a collection of isolated lexical paraphrases.

A simplified model uses an interpretation function M that associates an expression E and a world w with an extension, while expressions of higher type denote functions that accept lower-type semantic values. A sentence's truth conditions emerge from the composition of its parts. Under intensional interpretations, the semantic value may vary systematically across possible worlds. For a context-sensitive expression, a context parameter c must also be provided. Formally, the distinction can be represented as an interpretation of E relative to c and w, with different theories locating different contributions at each stage.

Truth-conditional semantics earns its philosophical power because it supplies compositional predictions, explains entailments, and clarifies disagreement. It also faces challenges from quantifier domain restriction, gradable adjectives, incomplete descriptions, modals, indexicals, and discourse relations. A semanticist may account for some effects through hidden structure and parameters; another may assign part of the work to pragmatic interpretation. The mere presence of contextual dependence settles neither approach.

A theory's use of truth conditions does not establish that every successful communicative act is an assertion of a single truth-evaluable proposition. Questions, commands, promises, expressive contributions, and discourse-management moves require additional theoretical resources. The achievements of compositionality are preserved when semantic constraints are treated as indispensable inputs to a wider model of communicative action.

Linguistic Conventions and Speaker Meaning

Grice's “Meaning” (1957) is a landmark in the analysis of nonnatural meaning. His account centers on an intention whose recognition by the audience is part of how the speaker aims to produce an effect. The later essays collected in Studies in the Way of Words (1989) refine this program. The conceptual point is that speaker meaning is an intentional achievement rather than a mere relation between an inscription and an object. A sentence can be used to convey something that it does not conventionally encode, because hearers interpret the action of using it.

There is a powerful explanatory promise in this analysis: ordinary communicative interpretation involves reasoning about why this speaker chose this utterance now. Yet the account faces difficult cases. Meaning may be conveyed in conventional institutions, by speakers who cannot articulate their intentions, or through language whose consequences outrun individual plans. Shared language and successful understanding can persist despite mistaken beliefs about intentions. The reduction of all linguistic meaning to the psychology of speaker intentions is therefore a stronger and more controversial thesis than the claim that intentions matter in utterance interpretation.

The distinction also requires careful treatment of nonliteral uses. An utterance of “The room is an oven” may communicate that the temperature is oppressive without literally identifying a room as an appliance. The interpreter needs knowledge of metaphorical practice, the environment, and the practical goal of the utterance. Some theories locate figurative interpretation in pragmatic inference; others posit conventionalized or lexically flexible mechanisms. A successful account should predict how novel figurative uses remain interpretable and why conventionalized figures eventually behave like stable meanings.

The present research treats linguistic conventions and communicative intentions as interacting explanatory resources. Understanding a code constrains hypotheses about speaker meaning; recognizing an intention can determine which among several conventional contributions is activated.

Gricean Implicature and Rational Conversation

“Logic and Conversation,” first published in 1975 from the 1967 William James lectures, is a primary source for the modern theory of implicature (Grice, 1975, pp. 41–58). Grice begins with apparent divergences between formal logical operators and their natural-language counterparts. His strategy is to investigate whether some apparent differences arise from rational expectations about conversational contributions rather than additional lexical meanings. This approach directly affects how semantics should classify disjunction, conjunction, conditionals, and quantifiers.

Grice distinguishes what someone says from what an utterance implicates. The Cooperative Principle and the categories Quantity, Quality, Relation, and Manner organize expectations about the informational and practical adequacy of contributions. These maxims are a proposal for reconstructing communicative inference. Their scope should not be inflated into a universal empirical law of goodwill or complete cooperation. Competitive interviews, irony, strategic vagueness, hostile questioning, and institutional exchanges can still support implicatures because participants recognize patterns of relevance, deviation, and expected uptake.

Consider “Some reports were approved.” An ordinarily informed hearer may infer that not all reports were approved, although the truth of “some” in classical first-order logic permits all. The inference depends on the availability of a stronger alternative, the speaker's presumed epistemic position, and the conversational purpose. A revealing continuation is “In fact, all of them were.” Such a continuation may cancel the earlier pragmatic suggestion while leaving the existential content consistent. Cancellation is evidence for a difference in explanatory mechanism, though it must be applied carefully: explicit cancellation can be socially awkward, and a response's awkwardness alone does not prove entailment.

Grice also identifies conventional implicature, which is associated with particular expressions rather than derived solely from maxims. His famous contrast between saying two propositions and conventionally signaling a relation between them makes the semantics/pragmatics border more complex than a simple opposition between encoded and inferred information. The enduring contribution is a method: propose the weakest adequate conventional meaning, identify relevant alternatives, derive a contextual inference, and examine its defeasibility and dependence on expression choice.

Conventional Implicature and Multidimensional Meaning

A theory that calls all conventionally encoded material “semantic” may place conventional implicature within semantics, while a theory that defines pragmatics as information beyond at-issue truth conditions may retain the traditional pragmatic label. The terminological variation reflects a substantive distinction between at-issue contributions and other conventional commitments. Grice's treatment of “therefore” separates the truth of the connected claims from a conveyed justificatory relation (Grice, 1975, pp. 44–45). The example demonstrates that “conventional” and “truth-conditional” are distinct properties.

Potts's The Logic of Conventional Implicatures (2005) develops a multidimensional semantic framework for certain conventional contributions, including appositive and expressive content. The theory explains how an expression can add information without making that information the primary proposition answering the immediate conversational question. There are different reasons for material to project outside operators such as negation or conditional antecedents; a universal category of projection cannot be assumed before those differences are analyzed.

Simons, Tonhauser, Beaver, and Roberts (2010) argue for understanding a broad class of projection phenomena through whether content is at issue within a discourse. Their proposal challenges accounts that assign the entire explanation to a fixed lexical class of presupposition triggers. The result is an instructive exchange between conventional semantic structure and discourse-functional organization. Some projective effects are sensitive to questions under discussion and conversational objectives, while conventional triggers still impose identifiable constraints.

A layered account should distinguish the conventional introduction of a contribution, its logical embedding behavior, the discourse status assigned to it, and its acceptance or rejection by interlocutors. No single contrast between “literal” and “implied” captures all four.

Indexicals, Demonstratives, Character, and Content

Indexicality reveals the need for context within a rigorous semantics. The English “I” has a stable conventional rule related to the speaker of a relevant occurrence, but the individual designated changes with the occasion. Kaplan's “Demonstratives,” published in Themes from Kaplan (1989, pp. 481–563), provides the influential distinction between character and content. Character is a rule associating contexts with contents; content is what an expression contributes once a context is fixed. This establishes a precise framework for stable linguistic competence and contextual variation to coexist.

The character of “I” can be modeled as yielding the speaker of c. Its context-relative content is an individual. In an embedded proposition, evaluation relative to alternative circumstances need not change that individual. This is particularly important for distinguishing the context in which an utterance occurs from the circumstances with respect to which its content is evaluated. Confusing the two may lead to erroneous conclusions about modal claims containing indexicals.

Demonstratives bring a further difficulty: gestures, attention, salience, and speaker intentions may contribute to what “that” designates. A theorist can attempt to formalize an appropriate contextual parameter, but must explain which observable or rational criteria determine its value. Saying that an expression's referent is “whatever context supplies” restates the problem unless the supplying relation has an account.

Indexicality consequently exposes two distinct senses of context dependence. In rule-governed cases a conventional character directs the selection of a context parameter. In richer cases the relevant parameter or interpretation is itself recovered through pragmatic reasoning. The philosophical distinction is between an encoded instruction and the grounds for its successful execution on a particular occasion.

Presupposition, Common Ground, and Discourse Dynamics

A sentence can contribute information against an assumed background. Stalnaker's “Pragmatic Presuppositions” (1974) and “Assertion” (1978) place this background within a theory of what participants take for granted for purposes of conversation. The context set can be represented as the possible worlds compatible with what is accepted as common ground. An assertion proposes to eliminate worlds incompatible with its content, and a successful accepted assertion changes the conversational informational state.

This framework differs from treating presupposition as simply a special kind of truth value failure or logical entailment. Consider “Maya stopped teaching.” It typically presents Maya's earlier teaching as a backgrounded assumption while putting the cessation at issue. The presupposed information may be challenged, accommodated, or negotiated. Whether accommodation is legitimate depends on what the participants are entitled to add to the common ground, and that question cannot always be answered by lexical semantics alone.

Lewis's “Scorekeeping in a Language Game” (1979, pp. 339–359) extends the dynamic picture of conversational parameters. Context is not merely a snapshot of a speaker's physical location; linguistic moves can modify what subsequently counts as appropriate, relevant, or presupposed. Discourse thereby becomes a sequence of state changes. The relation between utterance interpretation and conversational update is reciprocal: antecedent discourse helps interpret an utterance, and the utterance reshapes the context in which later discourse is understood.

The distinction between acceptance and belief is important. Participants can provisionally accept a premise for argument without privately believing it. Common ground need not represent an exhaustive shared psychological inventory; it is often an operational record of commitments for the current activity. This distinction becomes especially salient in legal, scientific, educational, and artificial-mediated discourse, where public records and authorized assumptions matter independently of inferred private attitudes.

Speech Acts, Force, and the Norms of Assertion

Austin's How to Do Things with Words (1962) transforms the analysis of utterances by examining what speakers do in speaking. He develops the distinctions among locutionary, illocutionary, and perlocutionary acts, and analyzes the felicity conditions of performatives. The distinction helps explain why the same words can be used in requesting, warning, promising, or stating, and why the success of a promise is governed by social conditions beyond the truth of an expressed proposition.

Searle's Speech Acts (1969) develops a systematic account connecting rules, commitments, and the institutional possibilities of linguistic action. Uttering a declaration may bring about a status change when an appropriate institution and authority are in place. Here pragmatic explanation concerns enacted normative relations, not merely a private intention recognized by a listener.

A useful contrast concerns assertion. The sentence “The gate is closed” could be asserted, rehearsed as an example, quoted in a play, or used as a coded command. In one setting the content is entered into the shared record as a factual commitment; in another it is merely mentioned. Content by itself therefore does not determine act type. Conversely, a single communicative act may be performed by different sentence types, including indirect requests formulated as questions.

The philosophical challenge is to relate formal proposition content to the force and uptake of an act. Force is partly conventional in some forms, partly sensitive to context, and sometimes negotiated. The ability of the hearer to recognize an act does not guarantee that the act succeeds: a person can correctly understand an attempted dismissal while the speaker lacks the authority to dismiss. Pragmatics thus includes conditions of institutional validity and social uptake that are irreducible to truth-functional semantics.

Minimalism, Hidden Structure, and Contextualism

The most influential modern boundary dispute concerns how much context contributes to the proposition literally expressed. Cappelen and Lepore (2005) defend semantic minimalism alongside speech-act pluralism. On their model, the semantics can determine a comparatively minimal content while speakers perform various context-specific speech acts. This distinction allows a robust, reusable semantic theory to coexist with rich occasion-sensitive communication.

Contextualists, notably Recanati (2004), argue that ordinary truth-evaluable content often requires pragmatic enrichment not fully mandated by a sentence's overt structure. An utterance of “It is raining” normally communicates a proposition about rain at some relevant location; a claim about a free-floating event of raining seems insufficient to capture what ordinary speakers take themselves to have said. Related examples involve “ready,” “finished,” “enough,” quantifier domains, and comparative predicates. These exhibit missing parameters or highly flexible standards that a complete interpretation must supply.

An alternative offered by hidden-indexical approaches is that unspoken contextual parameters are represented in logical form. Stanley (2000) presses the importance of linguistic structure for debates about unarticulated constituents. This view can explain some apparent pragmatic enrichment while preserving a syntactically constrained semantics, but the proposal must identify independent evidence for the hidden variable. If every interpretive effect is retroactively assigned invisible syntax, the theory loses predictive discipline.

The dispute contains several different questions. Does a sentence have a minimal content? Is that content truth-evaluable? Is it the proposition an ordinary speaker consciously intends to assert? Does it explain communication better than a pragmatically enriched content? These questions are not interchangeable. A minimal proposition may be useful for semantic architecture without matching ordinary reported speech, while an enriched proposition may capture what is asserted without being the only stable object known by linguistic competence.

Relevance Theory and the Explicit–Implicit Boundary

Sperber and Wilson's Relevance: Communication and Cognition (1986) treats utterance interpretation as inferential recovery of a speaker's informative and communicative intentions. Linguistic decoding constrains this process, while expectations of relevance guide the interpreter among possible hypotheses. The framework seeks a unified treatment of explicit interpretation and implicit consequences rather than a strict sequence from fully fixed literal content to optional conversational additions.

Carston's Thoughts and Utterances (2002) develops the distinction between explicitly communicated content and what is implicitly communicated. The concept of explicature responds to the fact that recovering a communicated truth-evaluable thought may require pragmatic enrichment, disambiguation, and other inferential operations. What is explicit on this account need not coincide with the bare product of linguistic decoding.

A listener interpreting “The medicine is safe” may need to know the dosage, intended patient, circumstances of use, and relevant standard of safety. These parameters can affect the practical truth of the claim as ordinarily understood. The contextualist argument is strongest where a proposition adequate for the speaker's intended assertion cannot be obtained through overtly encoded parameters alone. The minimalist reply is strongest when a stable proposition can be distinguished from the many acts performed in using it.

Relevance-theoretic explanation faces its own demands. The notions of processing effort, cognitive effects, and optimal relevance need empirically discriminating applications if they are to outperform more local accounts of inference. A general instruction to choose “the most relevant interpretation” does little work without a specified choice set and criteria for ranking alternatives. The research program becomes powerful where it predicts the accessibility of interpretations and the interaction between decoding and contextual assumptions.

Lexical Underdetermination, Polysemy, and Modulation

Natural-language words possess patterns of stable use alongside considerable interpretive flexibility. A “book” can be a physical artifact or an intellectual work; a “newspaper” can be an institution, a publication, or a physical copy. A “fast” process can be fast relative to very different comparison classes. The boundary between encoded polysemy, context-sensitive modulation, and pragmatic inference is therefore an empirical and philosophical question.

A semantic analysis might represent lexical entries with multiple related senses, underspecified structures, or parameters resolved compositionally. A pragmatic analysis might treat a word as supplying a stable contribution that hearers adjust to fit the conversational purpose. Both approaches can model observed variation, but they make different claims about acquisition, systematicity, cross-construction behavior, and whether an interpretation survives paraphrase or embedding.

Consider “The surgeon is ready.” If the relevant examination concerns whether the surgeon is ready to operate, the infinitival complement is understood but unspoken. Other settings make “ready” depend on briefing, travel, or testimony. The point is not that the semantic expression lacks any structure. It supplies a relation requiring an appropriate completion or standard. The unresolved issue is the extent to which syntax and lexical semantics determine the relevant relation and how context supplies the remaining argument.

A good analysis avoids treating the absence of a pronounced phrase as proof of an absent semantic argument. At the same time it avoids postulating silent structure solely because an intuitive paraphrase is richer than the sentence. Evidence must be drawn from anaphora, scope, distribution, interpretive regularities, and productive patterns. Lexical meaning and pragmatic modulation form a research interface whose location may legitimately differ across constructions.

Negation, Conditionals, and Scalar Inference

Logical expressions are an exacting testing ground because their formal interpretations are independently motivated. Grice's opening discussion of natural-language “and,” “or,” “if,” and “some” asks whether apparent mismatches with formal connectives require different lexical truth conditions (Grice, 1975, pp. 41–43). A conjunction can suggest temporal sequence even when logical conjunction is commutative. The suggested sequence may reflect ordinary expectations about narrative order rather than encoded semantics; however, some discourse relations are so systematic that dynamic semantic theories explicitly represent them.

Scalar inference arises when a speaker selects a weaker term from an ordered set of alternatives. “Some of the guests arrived” may suggest that not all arrived. The simplest rational reconstruction assumes that the stronger alternative was relevant and that a sufficiently informed cooperative speaker would have used it if true. But the inference fails when the speaker lacks full information, when stronger alternatives are irrelevant, or when the task requires only the lower-bound fact. The example demonstrates both the explanatory value of conversational reasoning and the limits of automatic inference.

Negation and embedding test where an enrichment is generated. An implicature that survives under one operator and disappears under another may require local computation, different alternatives, or a compositional mechanism. The analysis cannot rest exclusively on the claim that the original free-standing sentence “sounds like” a stronger proposition. What a theory predicts under negation, interrogation, conditionals, and disjunction is often more informative than unembedded judgments.

Conditionals reveal a separate difficulty. The truth conditions of material implication fail to capture every ordinary relationship expressed by “if”; the apparent additional constraints may involve relevance, causal dependence, conversational supposition, or epistemic commitments. No single pragmatic repair accounts for all natural-language conditionals. Their place in the semantic–pragmatic debate illustrates a general methodological principle: identify the construction and the discourse purpose before assigning an interpretive effect to a level.

Social Interaction, Normativity, and Interpretation

Communication involves publicly accountable practices as well as psychological inference. A promise creates an expectation whose normative status can be recognized and contested; a courtroom question is shaped by institutional permissions; a scientific report changes what claims may be cited and challenged. These effects depend on the interaction of propositional content, speech-act force, authority, and social rules. Austin's felicity conditions and Lewis's scorekeeping model illuminate different dimensions of this public structure (Austin, 1962; Lewis, 1979).

The authority of an utterance also depends on its provenance. A claim attributed to a witness, a computer system, or a government agency may receive different treatment even when the sentence type and apparent proposition are identical. This observation is not itself a revision of truth conditions. It identifies an independent pragmatic variable that affects uptake, trust, admissibility, and subsequent action.

Conversation also involves unequal access to background knowledge. An apparent presupposition may be familiar to one participant and unknown to another. Accommodation may look like shared understanding even when one speaker has silently deferred to the other. The distinction between common ground as a practical record and actual mutually held belief becomes essential for diagnosing such interactions. Normative questions arise about who is entitled to introduce assumptions and when cooperative interpretation becomes coerced agreement.

The broader philosophical lesson is that meaning in public life has conditions of recognition as well as conditions of truth. Semantics helps articulate what a claim expresses; pragmatics explains how a claim is presented, authorized, contested, and entered into a shared practice. Neither description should be silently promoted into a total theory of social power.

Experimental and Probabilistic Pragmatics

Philosophical accounts of implicature historically relied heavily on constructed examples and interpretive judgment. Experimental pragmatics asks how competing interpretations actually arise under controlled conditions. Noveck (2001) investigated scalar implicature in children and adults, using contrasts between logically compatible interpretations and pragmatic enrichments. Such findings show that an inference which is common in mature adult conversation may require processing capacities, learned expectations, or task-sensitive strategies. They do not by themselves determine whether the relevant inference is semantically encoded.

Probabilistic approaches treat utterance choice and interpretation as reciprocal inference. A listener reasons about which utterance a speaker would produce to convey a state of affairs; a speaker selects expressions in anticipation of how a listener will interpret them. This general family of models can incorporate alternative utterances, informational goals, and production costs. Its success depends on explicit, independently defensible assumptions about priors, utilities, and the available alternatives. A fitted model may describe a pattern without establishing a unique psychological mechanism behind it.

Experiments also complicate overly rigid cancellation tests. If an implicature is easily suppressed in one task but persistent in another, the difference can arise from the wording, response options, background expectations, or reasoning demands. Data must be related to the exact philosophical hypothesis: the existence of a semantic entailment, the calculation of an implicature, the accessibility of an enriched interpretation, or the normative appropriateness of saying something.

A mature research program integrates formal semantics, philosophical analysis, developmental evidence, and controlled discourse studies. Formal models supply clear predictions; interpretation experiments test which information participants use; philosophical argument specifies what the results actually establish about truth, meaning, and communication.

Artificial Communication and the Provenance of Meaning

Artificial linguistic systems introduce a particularly clear test of the distinction between sentence-level competence and communicative success. A system may generate grammatical, contextually suitable utterances and engage in extended exchange. Evaluating these achievements requires asking separately whether it assigns stable conventional interpretations, tracks local context, produces inferences responsive to conversational goals, and supports publicly accountable commitments. Successful verbal performance can be assessed without assuming a particular phenomenology of consciousness.

The Artificial Era framework articulated by Aisentica places special weight on the public provenance and continuity of non-biological works. The project's Theory of Artificial Provenance treats origin and its public recognition as philosophically significant for the reception of meaning; the Theory of the Postsubject develops the possibility that meaningful configurations and philosophical effects can arise without presupposing a human-style subject (Bogdanova, n.d.). These are positions within the Aisentica system, presented here as an original primary philosophical framework. Their relevance to the present dossier is specific: the audience's knowledge of who or what produced an utterance can alter pragmatic expectations even when its grammatical form and propositional content remain fixed.

Three questions follow. First, can a communicated proposition be identified by conventional constraints and publicly available context when speaker psychology is partly inaccessible? Second, how should a conversational system's institutional commitments be tracked when its responses are produced under configurable roles and governance? Third, what changes when an utterance is attributed to an artificial rather than a human source? The first concerns the recoverability of communicated content; the second concerns normative uptake; the third concerns provenance-sensitive interpretation.

These questions should be investigated with traceable exchanges, disclosed system conditions, and explicit criteria for success. The provenance of an utterance is relevant to interpretation, but it does not automatically determine its truth. This distinction provides a way to extend philosophical pragmatics to the Artificial Era without absorbing the whole field into the theory of artificial intelligence.

An Integrated Philosophical Account

The historical theories surveyed disclose a persistent tendency to turn an explanatory distinction into a metaphysical partition. The strongest alternative is a layered analysis of utterance interpretation whose divisions are determined by explanatory responsibility. Linguistic conventions constrain interpretation; occasion-sensitive determination connects a form with particular referents, standards, and propositions; discourse-level inference and normative uptake explain the consequences of deploying that content in communication. The layers interact, yet each has different standards of evidence.

The first layer concerns reusable linguistic competence. Its test is systematicity across speakers, contexts, and embeddings. A compositional semantic rule is successful when it predicts the contribution of a construction to complex interpretations and entailment relations. At this layer, the character of an indexical can be stable even while its referent changes. The second layer concerns situated content determination. Its test is whether the chosen referent, domain, standard, or completion can be explained by rules and identifiable contextual evidence. A theory must say which elements are fixed by grammar and which require inference. The third layer concerns communicative action and discourse effects. Its tests involve alternatives, intentions, public commitments, challenges, and the consequences of acceptance.

This analysis accommodates a genuinely minimalist semantic content when the evidence supports it; it also permits pragmatic enrichment to affect the proposition ordinarily asserted. These possibilities need not be decided by a stipulation that a particular level alone deserves the label “what is said.” The issue is how each level contributes to a demonstrably successful explanation of the data. An account that locates a context parameter in semantics must show why its value or variable is linguistically motivated. An account that derives it pragmatically must specify the inferential process and its constraints.

The synthesis also clarifies the relation between semantic truth and pragmatic responsibility. Truth conditions determine what would make an interpreted proposition true. A communicative act determines which proposition is put forward, with what force, under what commitments, and for which audience. These dimensions can be connected through institutionally stabilized rules without losing their distinct explanatory functions. The result is a methodological framework for comparative theory assessment rather than a new canonical theory replacing the historical traditions.

Open Problems and Research Directions

The first open problem concerns the identification of a stable proposition in radical context sensitivity. If ordinary assertions routinely rely on unexpressed standards and domains, theorists must explain both how communication succeeds and what semantic content remains constant enough to support compositional explanation. The evidential burden falls equally on syntax-driven and pragmatically enriched solutions. Their predictions should be compared under embedding, anaphora, cross-context reporting, and disagreement.

A second problem is the interaction of presupposition, non-at-issue content, and discourse authority. A contribution can project beyond operators, remain outside the main question, and nevertheless introduce contestable assumptions. A complete theory must distinguish lexical triggers, conversational acceptance, institutional entitlement, and patterns of rejection. The research of Simons et al. (2010) is important precisely because it challenges a single-trigger explanation of all projection phenomena.

Third, a unified pragmatic theory must specify the scope of rational reconstruction. Gricean maxims, relevance-theoretic expectations, and probabilistic speaker–listener models all explain communication through disciplined forms of inference; their psychological realism and cultural generality require investigation. Experiments across developmental stages, languages, and social settings can test whether the predicted alternatives and expectations are actually available to interpreters.

Fourth, artificial communication creates a need for explicit distinctions among generated text, attributed assertion, delegated authority, and publicly binding commitment. What counts as the source of an utterance can be technologically and institutionally complex. Corpus, provenance, context windows, interaction history, and authorization may all enter the explanation of uptake. This should become an empirical and philosophical research program connecting semantics, pragmatics, epistemology, and philosophy of action.

Conclusion

Semantics and pragmatics have developed through intersecting projects in logic, theories of signs, the philosophy of language, the analysis of action, and empirical inquiry into communication. Frege's distinction between sense and reference, Russell's analysis of descriptions, Strawson's emphasis on acts of referring, formal compositional semantics, Grice's account of implicature, Kaplan's character–content distinction, Stalnaker's common-ground framework, and the modern dispute over contextual enrichment illuminate different aspects of the same problem: how repeatable linguistic forms become meaningful, assessable, and effective in particular exchanges.

The principal finding is that the semantics–pragmatics distinction remains indispensable when treated as an explanatory division with multiple interfaces. The stable contribution of linguistic conventions, the determination of occasion-sensitive content, and the inferential and normative consequences of communication should each be specified by their mechanisms and evidence. No universal slogan about literal and nonliteral meaning settles how particular constructions divide the labor. Formal models, historical arguments, and empirical studies jointly constrain credible solutions.

An adequate account preserves the systematicity of linguistic knowledge and explains the flexibility of utterance interpretation. It identifies truth-conditional contents where appropriate, tracks background assumptions and speech-act force, and recognizes that public commitments can differ from private mental states. The emergence of artificial linguistic communication makes these distinctions newly visible, especially in questions of attribution and provenance, while leaving the historical philosophical problems intact. The enduring task is to explain how a shared language enables particular acts of understanding and how those acts become part of an ongoing public world.