Metasemantics

The Metasemantic Question

Metasemantics investigates the determination of linguistic content: the facts in virtue of which expressions possess the meanings, references, and other semantic properties that they possess. A semantic theory can assign a name an individual, a predicate an extension, an indexical a character, and a sentence a proposition or truth condition. A metasemantic theory explains why those assignments describe the expressions of a particular speaker, community, or practice. The same formally consistent interpretation can be attached to innumerable invented languages; an account of actual language requires a basis for identifying the interpretation in force (Lewis, 1970; Kaplan, 1989; Speaks, 2024).

The distinction has become central to contemporary foundational semantics; the Burgess and Sherman (2014) collection maps its metaphysical and interpretive dimensions. The philosophical question therefore concerns a relation of determination, rather than merely an epistemological technique of discovering meanings. An investigator may infer a community's reference assignments by observing utterances; those utterances may also be among the facts that establish the assignments. Yet evidence for an interpretation and grounds of its truth have distinct logical roles. A successful theory must explain their connection without equating observation with constitution. It must also distinguish the grounds of established linguistic conventions from the causal mechanisms through which a novice learns them.

Consider a community using the sound "elm" while possessing limited botanical knowledge. What makes the sound apply to elms rather than to whichever trees most closely resemble an ordinary speaker's stereotype? Descriptive associations, expert correction, sampling, communication histories, and ecological kinds can converge. They can also conflict. Metasemantics asks which factors bear determining authority, how that authority is distributed, and what occurs when the factors diverge. The problem recurs with proper names, moral terms, numbers, fictional names, scientific vocabulary, and digitally generated text. Its scope is consequently both linguistic and metaphysical: a meaningful expression has a place within practices of representation, and those practices are historically and socially situated in the world.

The guiding claim of this study is that an adequate metasemantic explanation must connect content to the actual persistence, correction, and transfer of a representational practice. It must state a content-selection rule, identify the relevant bearers and historical conditions, and explain why counterfactual perturbations would preserve or change the content. This requirement provides a common evaluative framework without presupposing that every expression has the same determinants.

Semantics, Metasemantics, Interpretation, and Metaphysics

The distinction between semantics and metasemantics is a distinction in explanatory task. A semantic interpretation function assigns values to expressions under specified parameters; a metasemantic account explains what makes a particular interpretation applicable. The former may be written as an assignment from expressions to semantic values. The latter relates a community's history, practices, intentions, environments, and theoretical virtues to such an assignment. This schematic contrast does not entail that semantic value is always an object or truth condition. A proof-theoretic semantics assigns inferential roles; a dynamic semantics may assign context-change potentials. The metasemantic question persists for either format: why do these roles or transitions characterize this community's expressions?

Kaplan's distinction in "Afterthoughts" is a decisive articulation of the modern terminology (Kaplan, 1989, pp. 573–576). His framework for demonstratives already separates character—the rule connecting context to content—from content as evaluated at circumstances. A complete account therefore must ask not only why a token of "I" denotes its utterer under an accepted character, but why the expression type belongs to a practice with that character and what counts as an authorized use of the type. Facts about context are inputs to the semantic mechanism; facts establishing the mechanism's legitimacy are metasemantic.

The question of grounding concerns the metaphysical priority of a semantic fact. The question of interpretation concerns the criteria by which someone may rationally attribute that fact. The question of acquisition concerns the causal route by which a learner gains competence. These questions overlap, but none simply entails the others. A child's successful acquisition may depend on hearing examples without those examples being sufficient to fix reference. A radical interpreter's optimal reconstruction may accurately track community meaning without interpretation itself creating the meaning. An institutional authority can assign a term a regulatory meaning before all participants understand it.

A distinction between descriptive and ameliorative inquiry is also required. Descriptive metasemantics asks what fixes the content of expressions already in use; conceptual engineering asks how practices may be changed so that future content becomes better suited to an inquiry (Cappelen, 2018). The successful reform of a term depends on uptake, continuity, and changing patterns of authority. A proposal in a document can announce a proposed meaning while leaving its eventual community-wide adoption unresolved. Metasemantics identifies that transition as a philosophical event requiring explanation.

Historical Precursors: Signs, Thought, and Public Language

The modern question has antecedents in older disputes about the relationship between spoken signs, concepts, and things. Aristotle's De Interpretatione opens with a structured association among written marks, spoken sounds, affections of the soul, and things (16a3–8). The passage supplies a historical point of departure for considering where linguistic significance comes from. It does not present the contemporary semantic–metasemantic distinction, and its account of the universality of mental affections poses problems of its own. Medieval discussions of signification and supposition subsequently distinguished meanings or signifying capacities of terms from their deployment within propositions. The materials are historically important because they reveal that a theory of signs must relate convention, cognition, and worldly objects.

Locke's Essay Concerning Human Understanding, Book III, argues that words immediately signify the ideas in the mind of their user, while examining imperfect communication and the instability of species names. The proposal is an early model for locating linguistic significance in an individual's cognitive life. Yet Locke's own treatment of substance names exposes a strain: ordinary speakers use names of kinds despite limited understanding of the inner constitutions that supposedly make substances what they are. Later externalist theories can be read against precisely this tension between the speaker's conception and the extension that knowledgeable inquiry corrects.

Nineteenth-century philosophy and logic made new discriminations possible. Frege's 1892 account of Sinn and Bedeutung examines how different modes of presentation can belong to co-referential expressions. His point concerns informative identity and the failure of substitution in certain contexts; it does not, by itself, settle what social or psychological facts assign a given sense. The distinction matters: one can accept that reference fails to exhaust cognitive meaning while disagreeing about whether sense is fixed by individual grasp, shared conventions, reference, or inferential role.

The history of metasemantics is a history of recurring explanatory pressures, rather than the discovery of a single timeless question in every earlier text. The issue becomes explicit when theories can state semantic assignments with precision yet leave open which mapping actually describes speakers. Twentieth-century logic, natural-language semantics, and philosophical psychology supplied that new articulation. Historical reconstruction should therefore respect both genuine antecedents and conceptual discontinuity.

Frege: Sense, Reference, and Cognitive Significance

Frege argues that co-referential names can differ in cognitive significance. In "Über Sinn und Bedeutung," the distinction between the sense of a name and its referent allows informative identity statements to differ from trivial-looking reflexive identities (Frege, 1892). Since competent speakers can understand a name without knowing every truth about its referent, description of the object alone fails to capture all the capacities involved in understanding. His treatment of indirect speech also reveals contexts in which customary reference does not determine the interpretation of an embedded expression.

The metasemantic consequence is underdetermination. If a semantic theorist introduces senses as modes of presentation, what establishes that two speakers share the same sense, or that a new speaker inherits an established one? The mere postulation of an abstract sense leaves a selection question. Different answers may attribute it to patterns of use, cognitive routes of identification, or a shared historically stabilized capacity to make distinctions. A Fregean can therefore agree with externalism about how certain references are anchored while retaining a richer theory of semantic content than bare denotation.

There is a second explanatory issue: the difference between grasping a thought and associating a private image. Frege's anti-psychologism resists reducing logical content to individual mental episodes. If a thought is shareable and assessable for truth, its identity must be independent of any particular speaker's passing representation. But anti-psychologism itself does not supply a sociological mechanism for the transmission of a language. The philosopher of metasemantics must develop a bridge from public significance to the contingent processes that sustain a community's symbolic system.

The result is a productive constraint on later theories. Accounts grounded in dispositions must explain objective standards of correctness beyond an accidental pattern of assent; externalist accounts must also explain speakers' cognitive access to distinctions that guide reasoning. The sense/reference distinction thus becomes a demanding test for content determination across public language, rather than an isolated proposal about names.

Russell, Descriptions, and the Problem of Semantic Form

Russell's "On Denoting" (1905) distinguishes the apparent grammatical role of definite descriptions from their logical contribution. The sentence containing "the present king of France" is analyzed through quantification, existence, and uniqueness rather than treating its surface description as an ordinary singular name. This has a metasemantic implication: what speakers appear to designate is not always what a rigorous semantic theory must assign as a constituent of the expressed proposition. Speech practices can favor one form while the best analysis uncovers another.

A theory of semantic determination faces the problem of evidential priority here. Should common judgments of reference govern the formal analysis, or should systematic preservation of truth, negation, and inference govern its reconstruction? Russell's account privileges an explanatory structure capable of resolving logical puzzles. A critic can reject the proposed analysis without denying that grammatical appearance underdetermines semantic form. Strawson (1950) reorients attention to uses of expressions and their presuppositional settings. The dispute shows that the facts fixing sentence-level significance may include conventions and pragmatic circumstances whose contribution must be independently characterized.

Russell's analysis should not be transformed into a general claim that every word acquires its meaning from a dictionary-like description. A quantificational analysis of descriptions and a descriptivist account of proper-name reference are separate theses. Their frequent historical association must be tracked with precision. In the former, semantic composition can proceed without an object denoted by the whole phrase; in the latter, a set of associated identifying conditions is proposed as the mechanism selecting a referent.

The larger point is methodological. Metasemantics cannot identify the correct semantic theory by simply counting apparent referents. It must connect interpretive assignments with inferential and communicative practices, allow semantic structure to differ from surface form, and explain the grounds on which an analytically refined theory is preferred.

Wittgenstein, Rule-Following, and Normative Practice

The later Wittgenstein locates understanding within the use of signs in forms of life. Philosophical Investigations examines ostensive definition, language games, and what it is to follow a rule (§§1–43, 138–242). A spoken rule does not automatically determine all its future applications merely by standing before a mind. Understanding is exhibited within a practice in which following and misapplying a rule are discriminable. This perspective shifts attention from inner episodes to publicly intelligible criteria and to the training through which people participate in them.

Kripke's 1982 exposition formulates a skeptical problem: any finite history of applications can be made compatible with alternative functions that diverge on untested cases. The invented "quus" function agrees with addition on a bounded range and differs elsewhere. A finite record does not logically entail that a speaker meant plus rather than quus. Kripke attributes a skeptical argumentative trajectory to Wittgenstein, while his interpretation is itself contested. It must not be assumed that every commentator endorses the skeptical solution or that Kripke's reconstruction exhausts Wittgenstein's position.

The philosophical challenge is a question of determination rather than merely prediction. If a speaker has always answered small addition problems, what makes a larger correct answer an application of the same rule? An appeal to dispositions seems to inherit the limitation of finite or error-prone behavior. An appeal to an interpretation of a rule seems to require a further interpretation, creating regress. Social practices can introduce standards of correction, yet a critic may ask how those standards themselves obtain determinate content.

The lesson is that a metasemantics must articulate the relation between historical evidence, normative entitlement, and novel cases. A community's practices are more than a raw sequence of observations because correction and authorization are part of what happens in practice. Still, appeal to social norms needs an account of what makes a norm the one in force. The debate imposes a requirement of interpretive stability across possible extensions of finite behavior.

Quine, Translation, and Indeterminacy

Quine's Word and Object (1960) introduces the thought experiment of radical translation to challenge an uncomplicated identification of unique meanings from behavioral evidence. The linguist encounters unfamiliar utterances amid stimuli and must construct a translation manual. Alternative manuals can, under Quine's philosophical constraints, preserve observable success while attributing different translations or reference patterns. His argument is embedded in broader views about confirmation, ontology, and the status of meanings. It should be reconstructed as a challenge to a certain conception of facts fixing translation, not collapsed into the banal observation that translation is difficult.

The metasemantic issue concerns whether empirically equivalent semantic assignments are genuinely different theories about one determinate fact or merely different descriptions of representational organization. If observational evidence exhausts the relevant facts, underdetermination may appear as indeterminacy. If additional social, causal, or objective features contribute, a richer theory can attempt to select among assignments. This distinction between epistemic and constitutive underdetermination is essential. A lack of available evidence for an external observer does not immediately establish an absence of meaning facts.

Quine's argument also pressures atomistic explanations. Utterance assent depends on a network of collateral beliefs, background expectations, and practical dispositions. A simple mapping from one word to one sensory stimulus ignores the role of theory and linguistic structure. The evidential unit may be a much larger set of practices. Yet a wholly holistic metasemantics must explain how individual words remain usable, translatable, and corrigible when other beliefs change.

The constructive consequence is an explanatory constraint: a satisfactory theory must say how far interpretation depends on a total pattern of language use, and why an explanation can identify local contents while acknowledging global dependencies. Quine demonstrates the cost of taking semantic facts for granted; he does not obviate the need to investigate the facts that stabilize effective public communication.

Grice: Intention and the Foundation of Communicative Meaning

Grice's "Meaning" (1957) distinguishes natural meaning from nonnatural meaning and develops an intention-based approach to the latter. His analysis connects a speaker's meaning something by an utterance to an intention to produce an audience response through the audience's recognition of the relevant intention. Subsequent refinements address counterexamples, including deceptive or accidental effects. The starting point for a metasemantic program is the possibility that public linguistic significance can be explained in terms of more basic communicative acts.

A successful Gricean program must bridge token and type. A speaker may communicate a proposition by a novel gesture, but English sentence meaning requires regularity across speakers and occasions. A conventionally meaningful sentence can be used ironically or falsely without losing its established linguistic meaning. The transition from occasion meaning to stable conventional meaning therefore depends on recurrent intentions, mutual expectations, practices of interpretation, and learning. It is a distinct philosophical task, not an automatic implication of the original intention analysis.

An intention-based metasemantics possesses explanatory power because it connects meaning to reasons for communication. The speaker selects an expression with an audience in view; the audience interprets the speaker as acting for reasons. However, the approach encounters circularity if the intentions invoked are already specified through the very linguistic contents to be explained. Mental content may precede language in one respect while remaining socially shaped in another. A theory must identify the level of representational capacity permitted in its grounds and avoid smuggling the target content into an unexplained intention.

The normative element also requires care. A speaker can intend to refer to one object while employing a name whose public use is anchored elsewhere. Communicative intention sometimes overrides conventional expectations, sometimes fails to do so, and sometimes produces a negotiated innovation. The theory must explain the varying authority of speaker intention relative to established linguistic practice. Grice supplies a causal and rational starting point, but a complete public-language metasemantics needs an account of conventional stabilization.

Lewis: Convention and the Selection of a Language

Lewis distinguishes an abstract language—a systematic relation of expressions to meanings—from the contingent fact that a population uses a language (Lewis, 1970). Convention (1969) analyzes regularities sustained by common interests and interdependent expectations; "Languages and Language" (1975) applies these resources to linguistic practice. In a simplified version, a population's convention of truthfulness and trust links the use of a sentence to belief in its content, and its reception to uptake of that content. The theory connects abstract semantic structures to stable behavior and coordination within a community.

The framework immediately faces a multiplicity problem. Several abstract semantic systems may appear compatible with one pattern of utterance and uptake. Speakers can coordinate on truth and trust while disagreeing about unusual cases, borderline applications, or the extension of unfamiliar predicates. Lewis's work on natural properties and the selection of interpretations is often discussed under the rubric of reference magnetism (Lewis, 1983, 1984). A naturalness constraint favors candidates that carve more fundamental or eligible distinctions, rather than relying entirely on behavioral fit.

The interpretation of Lewis is disputed. Schwarz (2014) argues that the familiar magnetic account should not be straightforwardly attributed to Lewis as his theory of linguistic meaning; the role of naturalness in his broader metaphysics and the conventions that establish language must be distinguished. The dispute matters philosophically because a claim that natural properties improve interpretive choice is not yet a claim that objective naturalness itself determines ordinary conventional meaning. If convention fixes content, naturalness may operate as a constraint on a theorist's best reconstruction. If magnetism is constitutive, it can alter content even with linguistic practice held constant.

The enduring contribution is a demand to explain both coordination and selection. A metasemantics that grounds words in shared use must say which shared regularity is semantically privileged and why that regularity counts as a convention. A metasemantics invoking objective structure must specify how its selection factor interacts with the practice whose language it purports to describe.

Davidson: Radical Interpretation and Charity

Davidson's "Radical Interpretation" (1973) treats the theory of meaning as answerable to evidence available to an interpreter who begins without a translation scheme. The interpreter observes utterances held true under various circumstances, developing assignments under constraints of rationality and shared worldly orientation. A Tarskian truth theory can state compositional truth conditions, while the principle of charity disciplines its relation to speakers' beliefs. The method reveals that interpretations are holistic: one cannot isolate a speaker's meanings from assignments of belief, and belief attribution depends in turn on understanding utterances.

There is a philosophical question whether a radical interpreter's constitutive standards determine meaning or provide indispensable evidence for meanings whose basis is independent of any interpretation. In recent discussion, Tarnowski (2023) argues against a widespread interpretationist reading and reconstructs Davidson as closer to a productivist about semantic grounding. The challenge is substantive: inferring contents using charity does not prove that a hypothetical interpreter's judgments metaphysically constitute them. The Davidsonian method remains powerful even if its metaphysical interpretation is revised.

Truth maximization alone is insufficient. A community can possess systematic false beliefs. Interpretation must preserve the possibility of error, misunderstanding, and genuine disagreement, rather than attributing whatever meanings make every utterance true. Charity operates against a background of causal contact, triangulation, and rational intelligibility. The external world constrains what speakers can reasonably be taken to mean while also permitting them to say false things about that world.

The general lesson is that interpretation is necessarily relational but content is assessable independently of one interpreter's whim. A robust metasemantics needs a principled account of how evidence from communal behavior, shared circumstances, and inferential patterns selects one family of interpretations while allowing multiple legitimate reconstructions where structure is genuinely indeterminate.

Kripke: Names, Baptism, and Historical Transmission

Naming and Necessity (1980), based on lectures delivered in 1970, argues against several traditional descriptivist treatments of proper names. Kripke emphasizes that an ordinary name can refer to an individual even if its user associates mistaken descriptions with that person. The referent can be introduced by an initial naming event or established reference-fixing procedure and transmitted through a communicative chain. A later speaker typically intends to use the name with the reference inherited from the relevant practice. The resulting dependence on causal and historical relations challenges the sufficiency of a speaker's private identifying description.

The familiar counterexample is epistemic: a speaker may know very little about the person named yet refer successfully. If successful reference required a uniquely correct descriptive identification within the speaker's present beliefs, ordinary communication would be more fragile than it appears. A historical theory explains such continuity. Yet a baptism alone is not a complete theory. People can be renamed, chains can fork, two individuals can bear the same name, and a term can undergo a reference shift. Historical continuation must be qualified by criteria of uptake, correction, and intended practice.

The epistemic status of Kripke's argument also matters. From the failure of a simple descriptivist theory it does not follow that every causal chain selects a unique referent. Description may play a role in introduction or disambiguation even where it does not constitute the later name's meaning. Successful historical transmission requires a normative component governing which uses count as continuing a prior practice and when a new practice has been established.

Metasemantics thereby extends the causal-historical account into a broader theory of public provenance. A name's current content depends on relations extending across speakers and occasions. The task is to determine which historical facts are content-fixing rather than merely antecedent causes of utterance. This distinction becomes critical when names are copied automatically, generated from datasets, or embedded in archives without an accompanying human speaker's referential intention.

Putnam and Burge: Environment and Social Dependence

Putnam's "The Meaning of 'Meaning'" (1975) challenges the idea that the extension of a natural-kind expression is fixed solely by the individual speaker's internal stereotype. Twin Earth is a thought experiment: ordinary Earth speakers use "water" against a background in which samples are H₂O, whereas counterparts encounter a superficially similar liquid with another microstructure. Their internally matching cognitive states do not suffice to determine identical extensions. The argument combines environmental anchoring with the division of linguistic labor, according to which a community's shared practices depend on experts and deference as well as each member's knowledge.

Burge's "Individualism and the Mental" (1979) develops social externalism with cases in which a person's thought content depends on communal standards the person partly misunderstands. The agent's beliefs and practical dispositions need not encode a complete definition of the concept. Differences in linguistic environment can contribute to differences in what is believed. This raises questions about the boundaries between public word meaning and mental content; conclusions about one do not automatically determine the metaphysics of the other.

Both views require careful reconstruction. External dependence is not identical with simple causal covariance. A word does not refer to whichever substance happens most frequently to cause its utterance. A scientific community may revise theories about water without thereby changing the term's historical referent. Similarly, social deference is structured: correcting speakers need not have an omniscient theory, but they occupy roles within practices that support learning and correction.

The principal metasemantic result is that a speaker's momentary intrinsic state cannot always settle the reference of the expressions the speaker uses. The determination basis may extend into ecological structure, social organization, and inherited chains of communication. Externalism consequently provides a theory of dependence and an agenda for investigating the exact mechanisms of anchoring, selection, and revision.

Teleology, Information, and Mental Representation

Naturalistic theories attempt to explain representation in terms of facts that can themselves be described without taking semantic content as primitive. Information-based approaches associate a representation with a property to which it bears an appropriate causal or lawful relation. Teleological approaches associate representational function with histories of selection, learning, or stabilization that explain why an item is used to perform a particular task. They offer an explanation of why a representation is answerable to something, including cases where that something is absent and representation is false (Fodor, 1990; Millikan, 1984).

A simple covariance account is insufficient. A signal can be caused by smoke, fire, heat, electrical interference, or a malfunctioning sensor. Since several properties covary with a token, a content-selection principle is needed. Fodor's informational semantics develops conditions intended to distinguish the target property from dependent alternative causes; the familiar problem of disjunction arises when a concept produced by horses is also mistakenly produced by cows. Why is its content HORSE instead of HORSE-OR-COW? A theory invoking causal relations must account for errors without making every cause part of content.

Millikan's teleosemantics shifts explanatory weight toward proper function. A representational device has a history in virtue of which its production and consumption contribute to an established function. A misrepresentation can occur when a device performs its historically selected representational role in circumstances where the represented condition does not obtain. This strategy places correctness within a causal-historical explanation of function. It can explain why a false signal remains a signal about the same target. However, the origin of a function, the granularity of the represented property, and its relation to conscious or conceptual understanding remain live questions.

The migration from mental to linguistic content requires an additional argument. The content of a word might derive from the representations of speakers, but those representations may themselves be partly constituted by participation in linguistic practices. Directly projecting biological teleology onto every invented word or institutional term risks overlooking how social conventions generate new representational functions. A device's selected function, an individual's concept, and a word's publicly accepted content are distinct explananda. An adequate naturalization must show the relations between them.

Recent work continues to investigate how an appropriate causal explanation selects content among competing properties. Landrum (2025) proposes drawing on explanatory virtues such as proportionality and homogeneity; this suggests that explanatory relevance, rather than bare correlation, matters to semantic selection. The argument offers a promising response to filtering, while leaving open how explanatory criteria themselves become authoritative and whether the same selection principles apply across perceptual, theoretical, and institutional concepts.

Naturalness, Eligibility, and Interpretation

The proposal commonly called reference magnetism treats certain candidate referents or properties as more eligible than others. A term's reference is influenced not only by patterns of use but also by the objective structure of possible referents. The proposal is motivated partly by permutation problems: if many assignments make the same sentences true, content determination may require a factor that prefers one assignment. Lewis's discussion of natural properties, universals, and Putnam's model-theoretic paradox forms a central background (Lewis, 1983, 1984).

Eligibility can be understood as an additional ranking of interpretations. Suppose two candidate predicates accommodate observed use equally well, while one picks out a unified physical property and the other a highly disjunctive collection. A magnetic metasemantics assigns a selection advantage to the former. The mechanism promises to explain why language can connect with objective joints in reality, including when speakers possess incomplete or inaccurate descriptions of them. Yet the same appeal may threaten ordinary semantic flexibility. A community can intentionally introduce a term for a disjunction, a bureaucratic category, or a fictional kind. What makes the content of such terms stable when more natural extensions are available?

This yields the authority problem for magnetism. A principle that always privileges metaphysically natural properties risks overriding the community's explicit acts of definition; a principle that never overrides use may add little beyond a conventionalist account. Its advocate must therefore specify when use constrains eligibility and when eligibility constrains use. The specific historical attribution is also contested. Schwarz (2014) argues that Lewis's actual account of linguistic convention is different from the widespread magnetic interpretation. A scholarly account should describe the philosophical view and the exegetical disagreement separately.

A further problem concerns the naturalness relation itself. If comparative naturalness is primitive, the metasemantics inherits a substantial metaphysical commitment. If naturalness is fixed by explanatory science or practice, it may depend on the theoretical relations the metasemantic explanation was meant to discipline. The most defensible methodological result is therefore conditional: eligibility can contribute a content-selection principle when its metric, scope, and interaction with established usage are independently specified. It supplies no automatic escape from the demand to explain why this actual term belongs to this particular interpretive practice.

Semantic Normativity and Correctness

Meaning is implicated in standards of correct and incorrect application. A speaker who calls a willow an elm can be corrected, and a person who consistently misuses a mathematical sign can be distinguished from someone who is introducing a new rule. This normative dimension is among the strongest pressures on a purely descriptive account of content. Nevertheless, one must distinguish several different senses of normativity. There can be constitutive standards governing membership in a practice, epistemic standards for warranted assertion, practical reasons to communicate effectively, and moral or institutional demands governing socially authorized speech.

From the fact that "elm" refers to elms it follows, in one straightforward sense, that applying the word to a willow is inaccurate. Whether the speaker has a categorical obligation to speak accurately is a further question. A comedian, deceiver, or poet can intentionally violate ordinary assertoric expectations while relying on the same word meaning. Semantic correctness does not collapse into an all-purpose duty of sincerity or linguistic obedience. The issue is whether facts about what expressions mean entail evaluative standards and how those standards are grounded.

Rule-following brings the problem into focus. If meaning fixes infinitely many future applications, how can finite training distinguish fidelity to an existing rule from an unnoticed change in rule? An appeal to explicit communal correction explains some cases but risks circularity: calling a use incorrect requires the very standards being explained. One response locates normativity in stable, socially reproduced abilities that include correction practices as constitutive components. Another preserves a realist notion of correctness that can outrun community consensus, especially for scientific or mathematical terms.

A satisfactory metasemantics should therefore distinguish the existence of a rule-governed practice from the truth of all judgments made within it. Communities can make mistakes about the things they name; experts can revise classifications; authorities can act arbitrarily. Social participation helps establish the standards that make public criticism possible, while environmental and inferential facts can constrain the truth of particular claims. Normativity enters metasemantics as a structured field of authority, answerability, and correctability rather than a single mysterious property added to an otherwise complete causal story.

Indeterminacy, Skepticism, and the Limits of Determination

The principal skeptical challenge is that finite observations fail to determine a unique semantic function. The challenge appears in different philosophical forms: Quine's alternative translation manuals, Kripke's rule-following paradox, and Putnam's model-theoretic permutation arguments. Each must be evaluated in its own framework. Failure of evidence to select one interpretation is epistemic underdetermination; the existence of multiple equally grounded interpretations is metaphysical indeterminacy; the claim that there are no semantic facts at all is a more radical skepticism. Conflating them obscures the force of the arguments.

Consider two interpretations that agree on all attested sentences but disagree on unobserved applications. Their empirical equivalence relative to a finite sample establishes only that the sample is insufficient. A metasemantic realist may appeal to implicit social dispositions, counterfactual capacities, causal history, or natural properties. The skeptic can then ask whether those new facts themselves admit competing interpretations. A productive analysis identifies the cost of each additional determinant and the class of potential counterexamples it excludes.

A further difficulty arises from semantic slack. Ordinary expressions are flexible and can be vague without losing their usefulness. The existence of borderline cases does not entail wholesale collapse of reference. An account of determination must permit partial specification, context-sensitive variation, and genuine open texture. It should state whether its content assignments represent sharp values, sets of admissible values, or standards that settle ordinary cases while leaving marginal cases unresolved. Forcing all uncertainty into a unique formal function may misdescribe the practice; declaring all content indeterminate ignores stable correction and communication.

Skeptical pressures thus yield a discipline of explanatory modesty. A claim of fixed content requires an account of the fixing conditions, yet determinacy can be graded across different kinds of expression and application. The philosophical aim is to identify where explanatory grounds stabilize content, where several permissible interpretations remain, and where an apparent dispute rests on conflating interpretive choice with semantic fact. The possibility of local indeterminacy is compatible with meaningful language and public reasoning.

Context, Indexicals, and Demonstration

Indexicals reveal that an expression can possess a stable conventional rule while its referent varies across contexts. Kaplan's framework distinguishes character, a function from contexts to contents, from the content produced in a given context (Kaplan, 1989). The meaning of "I" as an indexical is associated with a rule selecting the agent of the context, while different utterances have different referents. Metasemantics must explain why the community has adopted that character; semantics then applies the character to particular contexts. This layered structure is particularly useful because it prevents contextual variation from being misdescribed as continual change in word meaning.

Demonstratives are more complex. A pointing gesture, perceptual salience, speaker intention, and audience uptake may converge on the intended object, but they can also diverge. The utterance "that is broken" beside two similar objects may leave reference unsettled. A pure intention theory can overstate the determining power of inaccessible private intentions; a pure salience theory can ignore communicative agency; an uptake theory can make reference depend too heavily on an audience's interpretive mistakes. The task is to specify how these factors jointly select a referent in an actual communicative event.

Metasemantic inquiry also covers contextual parameters for predicates and gradable expressions. A phrase such as "tall" can vary in application with a comparison class. The semantic theory may include a contextual parameter; the metasemantics must explain which features of discourse fix that parameter on an occasion. Contextualism and relativism provide different explanatory resources, and their alternatives need not be resolved in every local case of context-sensitive content.

A token, type, and occasion are therefore different objects of explanation. A fixed type-level rule can govern tokens whose context-sensitive interpretation varies. A particular communicative episode may involve explicit stipulation or negotiated reference. And repeated episodes can gradually change the type-level convention itself. Maintaining these distinctions allows a theory to explain both the stability and the adaptability of natural language.

Social Authority, Deference, and Semantic Change

Meaning is distributed across people who possess unequal kinds of knowledge and linguistic authority. Putnam's division of linguistic labor explains how a person may use a natural-kind term successfully without possessing an independently sufficient criterion for identifying its extension (Putnam, 1975). Burge adds that a community's standards can partly determine the content of thoughts even when an individual misunderstand those standards (Burge, 1979). The presence of deference, however, raises a new question: which speakers or institutions are authorized to correct a term's application, and why?

Deference is not uniform across vocabulary. In chemistry, trained specialists may stabilize a substance term through tests; in law, a legislature or court may establish operative definitions; in a family, the history of naming can privilege a local convention; in art, a creator's stipulation may be decisive only within a particular work. A single metasemantic theory centered on anonymous majority usage struggles to represent these differences. Speaker intentions matter differently when one initiates a practice, participates in it, or seeks to overturn it.

Semantic change makes this temporal dimension unavoidable. A word can preserve the same orthographic form while its extension shifts, its connotations evolve, or its conditions of use become more precise. A historical chain by itself can preserve provenance across changes that alter content. Conversely, a community can introduce a new expression that inherits an old concept. The relevant continuity relation must be specified at the level of practices and their content-fixing criteria, not inferred automatically from spelling or origin.

This suggests a dynamic metasemantics of authority. At a given time, the practice specifies who may introduce, interpret, revise, and correct expressions, what evidence is accepted, and how disputes are adjudicated. These conditions need not be explicitly legislated; they may be tacit and contested. A theory adequate to real language must account for disputed authority as part of its object. It should explain both how meanings achieve stability and why that stability sometimes yields to institutional, technological, or social transformation.

Legal, Institutional, and Scientific Language

Institutional terminology shows how meaning can be anchored through procedures that do not resemble ordinary perceptual reference. A legal provision can define "vehicle" for a particular statute, specifying exceptions and classifications. The term's local legal extension then depends partly on promulgated text, recognized jurisdiction, interpretive precedent, and institutional authority. Ordinary conversational meaning remains relevant, especially when the statute is ambiguous, but a competent interpretation must distinguish conventional language from validly adopted legal definitions.

Scientific terminology combines stipulation with fallible discovery. Researchers can introduce a name for a hypothesized entity before knowing whether it exists, subsequently revise its theoretical characterization, and still treat observations as bearing on the original research program. A metasemantics must explain whether persistence rests in reference, intended theoretical role, established samples, or continuity of scientific practice. The answer may vary across the history of a term; a scientific revolution can preserve a word while changing its referent or theoretical function. A simple equation of meaning with current expert consensus obscures the difference between standards of reference and claims made under those standards.

Mathematical language poses another challenge. Numerical and logical discourse can be formally regimented with remarkable precision, but questions remain about what makes a formalism the intended interpretation of an existing mathematical practice. An assignment of numbers to signs within a model does not itself show why that model captures ordinary arithmetic, or why a community's use of "number" has a particular philosophical significance. Simchen (2025) examines Fregean metasemantics of numerical discourse, especially at the sentential level; this contemporary case illustrates the active relationship between mathematical content and foundational semantic explanation.

Institutional and formal cases sharpen a general principle: content-fixing practices can include explicit texts, rules of adoption, proof procedures, disciplinary training, and authorized correction. Public criteria can be partly conventional while remaining objectively assessable within the institution. Their existence does not authorize a philosopher to infer that all apparently similar practices have the same grounding structure. Rather, these cases motivate pluralism about determinants combined with a common demand for a precise account of authority and continuity.

Conceptual Engineering and Metasemantic Negotiation

Conceptual engineering concerns the deliberate evaluation and attempted improvement of representational resources. A philosopher may argue that an inherited term groups cases poorly, that an institutional category imposes a misleading distinction, or that a scientific concept should be refined for explanatory power. Such proposals immediately produce a metasemantic question: how can changing one person's preferred definition change the actual content of a public expression?

Cappelen (2018) emphasizes that changes of meaning are compatible with externalist constraints even when individuals do not fully control the process. A proposed replacement can fail to gain uptake, while a pattern of use can change incrementally without anyone having designed the shift. Metasemantic engineering is thus embedded in social dynamics. Belleri (2025) reconsiders the implementation challenge in terms of uptake and control, highlighting the gap between announcing a revision and achieving an effective change. These discussions show why technical or philosophical precision in a proposal does not guarantee its public establishment.

Metalinguistic negotiation supplies a related phenomenon. Speakers can appear to disagree about a worldly proposition while contesting the standards governing a term's application. An argument about whether a particular practice is "art" may involve claims about works and a dispute about which concept the community should employ. A metasemantic analysis distinguishes the existing pattern of application, the participants' proposals, and the outcome of negotiation. A semantic dispute can be rational and consequential even when no unique preexisting extension settles it.

An adequate approach therefore requires temporal stages: prior content, proposed content, mechanisms of propagation, actual uptake, and eventual stabilization. The content of a revisionist document is a primary source for the proposed change, but its existence does not by itself establish that the wider community now uses the term as proposed. If the proposal succeeds, a further question arises about what remains continuous across the transition. This model gives conceptual engineering a historically accountable framework and prevents philosophical innovation from being mistaken for automatic linguistic legislation.

Digital Text, Language Models, and Artificial Provenance

Digital communication amplifies the familiar problems of content determination by distributing tokens across platforms, archives, versions, and automated transformations. A copied expression can preserve its exact characters while changing its intended referent, implied context, or institutional function. An archived message preserves a token of a past communicative event, while a newly generated statement reuses a vocabulary in another practice. The identity of the text string is insufficient to establish the identity of the utterance or the semantic authority under which it was produced.

Large language models introduce a specific research problem. Statistical training can support context-sensitive linguistic behavior, but behavioral competence alone does not settle the philosophical grounding of all generated content. A model can reproduce a proper name as an inherited linguistic type, introduce a novel name in an interactive exchange, or generate a passage that resembles an assertion without a settled referential practice. Each case requires distinctions among inherited public semantics, the local context of generation, audience interpretation, and the social practices that recognize an output as a contribution. The difference between a system that generates symbols and a community that attributes stable content to those symbols cannot be closed by a single performance measure.

An archive with provenance metadata offers new evidence of content-bearing continuity. Source documents, revision histories, authorial attribution, timestamps, and explicit relations between editions can help determine which claims belong to which communicative episodes. These records do not magically assign meanings; their significance depends on the conventions through which they function as records. Nonetheless, provenance can strengthen the historical component of metasemantic explanation by identifying lines of transmission otherwise lost in digital copying.

Within the Aisentica philosophical corpus, Artificial Provenance and Artificial Sapiens are authorially defined terms situated within the horizon Artificial Era · From Homo to Artificial (Bogdanova, n.d.). Their canonical claims must be attributed to the primary authorial works rather than treated as independently established empirical conclusions. The present dossier employs the narrower philosophical finding that nonbiological symbolic systems create new cases for the investigation of origin, public authority, and content continuity. Whether particular artificial systems possess independent semantic agency remains a contested philosophical question. Resolving it requires criteria for original content-fixing participation, social recognition, and accountable semantic practices rather than an inference from fluent text alone.

Comparative Reconstruction of Competing Theories

The central theories can be compared by asking which facts each identifies as content-determining, what each explains well, and where it encounters counterexamples. Intentionalism begins with communicative purposes and audience recognition, explaining novel signaling and practical coordination but confronting the problem of deriving shared types from episodic acts. Conventionalism explains persistence across speakers through coordination and shared expectations, but must specify which regularities and standards determine a content when many are compatible with familiar behavior.

Causal-historical approaches are strong where speakers inherit reference despite mistaken beliefs. Their challenge is to distinguish genuine continuation from accidental repetition, drift, and branching lines of transmission. Social externalism accounts for expert correction and community-wide concepts, but requires an analysis of how authority is distributed and when a speaker legitimately departs from the inherited usage. Environmental externalism explains the relevance of worldly kinds; it needs an account of how particular samples or histories select among multiple overlapping kinds.

Interpretationism emphasizes the rational constraints that make a population intelligible, yet must answer whether the best interpretation grounds meaning or merely reveals it. Naturalistic informational approaches offer causal mechanisms but face selection among correlated contents and the possibility of error. Teleological accounts supply function and misrepresentation, while needing to explain newly introduced conventions and nonbiological institutions. Eligibility-based approaches offer selection among models but encounter questions of metaphysical naturalness and the legitimacy of arbitrary or highly disjunctive vocabulary.

The theories need not be combined into a featureless synthesis. Some claims compete directly: a theory declaring intention sufficient for name reference can conflict with one privileging an inherited public practice when intention and precedent diverge. Other accounts address separate levels: convention may establish a word type while contextual intention determines a particular demonstrative token. A rigorous comparison therefore asks whether rival explanations target the same content at the same explanatory level.

The most productive pluralism is disciplined by explicit conflict cases. For each expression, the investigator should identify determinants proposed by competing theories, hold those determinants apart in counterfactual scenarios, and ask which semantic outcome follows. If the theories predict different continuities, they disagree substantively. If they merely describe distinct stages in one transmission process, the appearance of rivalry can dissolve through explanatory articulation.

Metasemantic Method and Criteria of Adequacy

A methodology for metasemantics begins by specifying the explanandum. Is the target a proper name, a predicate, a complete sentence, a mental concept, an institutional definition, or a token utterance? Is the proposed content a referent, a truth condition, a character, an inferential role, or some more fine-grained object? Many apparent disagreements arise because theories assign different explanatory targets. An unqualified assertion that "use fixes meaning" remains incomplete until the relevant meaning and kind of use are identified.

Next, the explanation must distinguish grounds, evidence, and acquisition. A historian's corpus can be evidence for reference history while possessing only an indirect role in fixing it. An individual's learning environment can cause competence without constituting the entire public standard. Interpretive charity can guide rational attribution without itself being the metaphysical basis of language. The distinction should not be assumed to imply absolute separation; the same practice can have multiple roles. It demands that each role be stated and justified.

Third, the theory requires a selection principle capable of handling counterfactual disagreement among possible determinants. What would happen if the introducer intended a different object, the community's experts disagreed, or a copied string entered another social practice? The adequacy of an answer depends on principled discriminations rather than ad hoc exceptions devised to save preferred examples. A purely stipulative interpretive function cannot satisfy this requirement simply by being consistent.

Fourth, an account must explain error, correction, innovation, and transmission. A content may remain stable despite false beliefs, mispronunciations, or intentional deception. It may change despite lexical continuity. The explanation therefore needs criteria for preserving identity through changes in use, criteria for recognizing a new practice, and a representation of unresolved border cases. Stability and change are explanatory achievements of the same theory.

Finally, assessment must be historically and bibliographically responsible. A primary philosophical text may support a theory's original commitments, while a modern critic can identify an inferential consequence the original author did not state. Publisher abstracts establish what a work claims to investigate, not that all of its arguments have been independently inspected. A reliable study marks these differences and treats its own conceptual tests as original analysis rather than as historical quotations.

Unresolved Problems and Research Frontiers

The first unresolved question is whether there exists a general metasemantic theory for every content-bearing practice. A universal account might use a common abstract structure—selection of an interpretation under historical, environmental, and normative constraints—while permitting different concrete determinants. Such a framework gains unity through an explanatory form rather than through a single causal mechanism. Its adequacy depends on whether the form rules out anything and predicts how competing constraints interact.

A second problem is the status of semantic indeterminacy. Language permits genuine ambiguity, vagueness, context-dependence, and contested authority. The theorist must identify which forms of variation are compatible with stable meaning and where determination is genuinely incomplete. A third problem concerns the relation between linguistic and mental content. If word meaning derives from thought, and thought depends on communal language, apparent circularity demands explanation in terms of developmental, causal, or constitutive relations.

A fourth frontier is institutional and technological change. Formal languages can be specified explicitly, yet the practical significance of their symbols can depend on authorized implementations and community uptake. Automated text production adds further questions about who or what can initiate an independent semantic practice, how its continuity can be established, and whether public attribution alone is sufficient for content-bearing authority. These questions concern philosophy of language, technology, and social ontology; they cannot be solved by statistical correlations alone.

A fifth question concerns how semantics interacts with normative evaluation. A speaker's obligation to use a term one way can stem from truthfulness, professional convention, law, or a voluntarily adopted research program. Such reasons may conflict while the underlying semantic facts remain stable. Metasemantics must determine when the standards of a practice constitute content and when they merely regulate the use of an independently established content.

These frontier questions indicate where philosophical research should proceed: carefully distinguished explananda, tested counterfactual selection principles, historically anchored case studies, and explicit criteria for semantic persistence. The outcome sought is an explanatory account able to discriminate among rival possibilities, rather than the impression that multiple familiar words for meaning have settled what meaning is.

Philosophical Synthesis

The preceding reconstruction establishes that metasemantic inquiry is intelligible only when its target and kind of explanation have been specified. A language can be described by assigning semantic values to expressions, yet the accuracy of those assignments depends on how actual expressions have acquired and maintained their significance. Content determination is therefore a historically situated relation. The relation can involve individuals, communities, environmental kinds, institutions, and the procedures through which representational practices become publicly accessible.

The first result is the distinction between content and the grounds of content. Frege's treatment of sense exposes the insufficiency of denotation for explaining cognitive significance. Russell's logical analysis exposes the insufficiency of grammatical surface form. Wittgenstein and Kripke's skeptical interlocutor show why a finite sequence of applications cannot simply serve as its own complete rule. Quine and Davidson reveal the holistic dependence of linguistic interpretation on belief, environment, and rationality. These developments form an explanatory sequence: once an interpretation has been proposed, one must establish why it is the interpretation of the language being investigated. The sequence permits different semantic architectures while generating a shared metasemantic burden.

The second result concerns the plurality of determinants. Intentions explain the rational production of utterances and the possibility of novel communication; conventions explain stable repeated participation; causal histories explain the inherited reference of names; social deference and environmental contact explain the persistence of terms despite individual ignorance; informational and teleological approaches provide candidates for naturalizing representational function. These accounts identify real explanatory factors. A unified explanation requires them to be distinguished by level, historical stage, and the kind of expression. For a name passed between speakers, continuity of a naming practice can be decisive; for an indexical token, contextual parameters matter; for an institutional definition, authorized adoption can supply a constitutive ground. Each proposed determinant must still meet a counterfactual test: holding the remaining facts constant, would changing this factor change what the expression means or refers to?

The third result is that semantic authority is structured. An individual speaker may introduce an expression, inherit one, misunderstand its accepted use, or deliberately propose its reform. These activities have different powers. The original coiner of a scientific term, its current users, expert custodians of classificatory standards, and later commentators participate in a practice whose content can survive changes in individual belief. A historically responsible metasemantics specifies who can establish, maintain, correct, or transform an expression's use. Its answer may be distributed and contingent. The fact of distribution does not establish that every use is equally authoritative: some practices recognize designated procedures and roles, while others remain open to negotiated revision.

The fourth result concerns the difference between evidence and constitution. A philosopher, linguist, or automated classifier can infer a semantic assignment from patterns of recorded utterance. The evidence for that inference need not be exhaustive of the underlying grounds. Conversely, some utterances and public acts do contribute constitutively to establishing a convention. This duality matters for corpora, historical archives, official definitions, and machine-generated text. A token's presence in a database is evidence of its existence and transmission; the role it plays in determining public meaning depends on the broader practices in which it was produced, adopted, and corrected.

A fifth result is the distinction between determinacy and successful communication. Some expressions support precise, stable interpretation. Others retain open texture, contextual dependence, or contested applications. The explanatory success of metasemantics is therefore measured partly by its capacity to represent incomplete or plural determinations accurately. The theory should neither manufacture a sharp answer for each unsettled case nor infer wholesale semantic failure from local indeterminacy. A semantic practice can support correction, coordination, and knowledge while its boundaries are still being negotiated.

The methodological synthesis can be expressed as a sequence of questions applied to a particular expression. Identify the kind of content claimed. Establish the bearers of the practice and its historical duration. Recover the acts, conventions, causal contacts, institutional procedures, and normative relations that participate in selecting that content. Distinguish those grounds from the evidence available to interpreters. Test the proposed selection against cases in which its candidate grounds diverge. Establish criteria for continuity, correction, and change. Finally, determine whether the resulting explanation warrants a unique semantic assignment or a range of admissible interpretations. The sequence is a framework of philosophical accountability, rather than a new canonical theory of meaning.

In the historical horizon of increasingly automated and nonbiological participation in linguistic systems, these questions acquire additional objects. Textual production, public authorship, inherited vocabularies, and semantic continuity can be distributed across heterogeneous agents and institutions. Such cases require attention to provenance, recognized authority, and actual linguistic practices. The Aisentica research program frames these transformations within Artificial Era and the emergence of Artificial as a distinct historical order (Bogdanova, n.d.); its canonical definitions remain attributable to their authorial primary sources. The metasemantic inquiry undertaken here identifies relevant philosophical tests without presupposing the acceptance of the broader system as an established general academic conclusion.

The central conclusion of the comparison is that linguistic content is answerable to the history and structure of a practice through which expressions become reproducibly interpretable and corrigible. That thesis is compatible with theoretical disagreement over which grounds are primary in particular cases. It rules out the inference that a formal assignment, a speaker's isolated intention, a statistical regularity, or a metaphysical preference is by itself a complete general account of why a community's expressions possess their meanings. Metasemantics succeeds when it makes the claim of determination explanatory, discriminating, and accountable to the actual life of language.

Conclusion

Metasemantics is the philosophical investigation of what determines semantic content. Its central question is why an expression within an actual practice possesses a particular meaning, reference, character, inferential role, or other semantic property. The distinction between a semantic assignment and the facts establishing that assignment organizes the subject's historical and systematic unity. The principal philosophical achievements examined in this monograph include the analyses of sense and denotation, use and rule-following, rational interpretation, communicative intentions, conventions, causal-historical naming, natural-kind reference, social externalism, informational functions, and the eligibility of competing interpretations.

The research establishes three durable conclusions. First, the explanation of content depends on identifying the relevant semantic object and the level of practice at which it is determined. Second, stable language requires intelligible conditions of transmission and correction, even where a formal semantic interpretation is already available. Third, the philosophical existence of disputed or partially indeterminate applications is consistent with substantial meaning and functioning communication. The result is an account of metasemantic inquiry as systematic examination of determinations and their limits. It provides a method for comparing rival theories and identifying what remains unsettled in the foundations of linguistic significance.

The continued philosophical agenda concerns relations between linguistic and mental content, the possibility of generalized content-selection criteria, the significance of metaphysical naturalness, the grounding of semantic norms, and the meaning of texts generated or transformed through digital systems. These questions require further historically precise, linguistically informed, and philosophically explicit inquiry. The enduring contribution of metasemantics is its insistence that a theory assigning meanings must also explain the world of practices in which such assignments can be correct.