Proper Names

The Philosophical Problem of Naming

A proper name enables discourse to remain about an individual while descriptions, beliefs, circumstances, and the names available to interlocutors change. A speaker can say that a person was born in one city, acquired a new profession, and later became famous, keeping the subject of the conversation fixed while predicating changing properties. The apparent simplicity of this function has made naming a test case for theories of language and thought. In ordinary speech one asks who is named; in philosophical semantics one asks what makes an occurrence of a name refer to a particular bearer, what it contributes to a sentence, and how different speakers understand it well enough to communicate.

At least four questions are present. The semantic question concerns the contribution of a name to the truth-conditional content of a sentence. The metasemantic question concerns the facts in virtue of which that contribution is determined. The epistemic question concerns a subject's recognition, beliefs, and ability to identify a bearer. The social-historical question concerns how name-using conventions persist, spread, branch, and occasionally change. These questions interact but are not equivalent. A semantic theory saying that a name contributes its bearer still owes an explanation of how speakers pick out that bearer. A causal history of transmission still owes an account of what an audience understands when it does not know that history. An epistemology of recognition still owes an explanation of mistaken or contradictory beliefs about one and the same individual (Cumming, 2023; Recanati, 1993).

The problem becomes most visible in five constructions. An informative identity, such as the astronomical identification of Hesperus and Phosphorus, tells a competent speaker something new even if each name designates the same planet. A modal claim concerns what the named individual could have been, regardless of the descriptions currently associated with it. A negative existential seems to say something intelligible about an allegedly nonexistent bearer. A belief report may cease to preserve ordinary truth when one co-referring name is substituted for another. A naming change may allow a form inherited from one referential practice to acquire a new public bearer. A theory of names must explain the phenomena without deriving answers by treating any one example as decisive in advance.

The scope here is the philosophy of proper names as singular and socially organized linguistic expressions. The broader theory of reference, the semantics of definite descriptions, theories of indexicals, and theories of fictional entities enter where their arguments determine the structure of naming. Their independent disputes remain distinct research subjects. The study's governing criterion is explanatory articulation: a philosophical position is adequate only to the extent that it identifies which naming phenomenon it explains and which supporting assumptions make the explanation possible.

What Counts as a Proper Name

The grammatical category of a proper name needs separation from that of a proper noun. A proper noun is a lexical item; a proper name can be an entire noun phrase containing multiple lexical items. The name of a person may comprise given name and surname, while the name of a river, bridge, institution, artistic work, or corporation may contain articles and ordinary nouns. The Cambridge Grammar of the English Language and contemporary semantic discussion both use such examples to prevent the theory from equating names with single capitalized words (Huddleston & Pullum, 2002; Cumming, 2023). Capitalization is an orthographic convention, not an invariant metaphysical or semantic criterion.

Linguistic form supplies several diagnostics, none sufficient by itself. Proper names often serve as subjects without an overt determiner, occur in direct address, and resist ordinary descriptive modification. Yet many appear with articles, as in established geographical names; others admit constructions such as 'three Marys' or 'a different Shakespeare.' A city name can also be used metonymically for its government or population, and a personal name for an author's works. These extensions need not be evidence that the original name literally denotes a collection of attributes. They require inquiry into shifts between literal use, grammatical conversion, and pragmatic inference.

In philosophy the central object is often a particular use of a name rather than an unqualified word type. The phonological or written form 'Alex' is compatible with indefinitely many bearers. A specific name-use can participate in a relatively stable practice directed toward one bearer. This distinction preserves the reality of homonymy without assuming that each spelling has a single referent or that every occurrence institutes a fresh reference from nothing. It is particularly important in multilingual communities, historical records, family names, and databases that contain multiple persons with identical labels.

A broad theory should also ask what makes an expression a name when it names an organization, a mathematical object, a historical event, or a fictional character. Some of these uses invite ontology-dependent interpretations; others are intelligible as parts of documentary, literary, or institutional practices. The working definition for this monograph is therefore functional and provisional: a proper name is a linguistic expression governed by a naming practice and characteristically used to single out a particular intended bearer, whether or not every use succeeds in reference. Its exact semantic category remains an object of theoretical disagreement, especially between referential and predicative approaches (Burge, 1973; Jeshion, 2015).

Names, Individuals, and the Ontology of Reference

An apparent name-bearer may be a material organism, a geographic place, a legal organization, a fictional person, an abstract object, or an allegedly nonexistent entity. The philosophical study of names must distinguish the question whether an expression is used as a name from the question what kinds of entities there are. Otherwise ontology is covertly built into semantics: either every intelligible name is assumed to denote an existent thing, or every nonconcrete name is classified as a linguistic illusion. Both assumptions would settle disputed cases by stipulation.

A formal singular constant in first-order logic is assigned a domain member by an interpretation. This provides a controlled model of reference, but its assignment function is not yet a historical account of how people name. Natural names may enter a language long before anyone possesses a stable inventory of the named individual's properties. Moreover, the domain of a model is a mathematical parameter; interpreting it metaphysically as a list of everything that exists requires further argument. Formal reference and natural-language reference overlap as explanatory objects without sharing an identical origin.

Different ontological commitments arise when a name occurs in an existential claim, a fictional narrative, a report of belief, a legal act, or a counterfactual proposition. 'Ada exists' differs in structure from 'Ada is fictional'; the latter can be an ordinary true claim about a created literary character even when the embedded use of 'Ada' inside a story is understood by pretense rather than ordinary singular reference. Conversely, one can name a legal corporation without supposing that its legal identity is simply identical to any one physical person or building. The semantics needs a theory of context, object-kind, and reference success before strong ontological consequences can be read off from names alone (Braun, 1993; Thomasson, 1999).

The ontology of names themselves must also be distinguished from that of their bearers. A name as a repeatable expression, a particular utterance, an inscription in a registry, and a convention connecting form and bearer are different entities. Changing a label in a database is not automatically changing the individual. Conversely, transferring the same label from one organization to its successor can preserve the form while altering the institutional target. A complete theory thus needs at least the expression, its token occurrence, the relevant practice, the bearer or intended bearer, and the speaker's referential act as conceptually separable relata.

Historical Foundations Before Analytic Semantics

Reflection on naming preceded modern truth-conditional semantics. Plato's Cratylus examines whether names are correct through nature, convention, or some mixture and asks what a name can disclose about the thing named. Its dispute over the correctness of onomata encompasses words and naming more broadly than the contemporary technical category of proper names. Socrates' movement from etymological speculation to questions about knowledge of things places a lasting constraint on linguistic inquiry: the explanatory power of names cannot be established merely by multiplying derivations of their forms (Plato, Cratylus, especially 383a–390e and 436b–440e).

Aristotle treats spoken sounds as symbols of mental affections and written marks as symbols of spoken sounds, while analyzing names within the logic of predication and propositions (De Interpretatione, 16a3–8 and 16a19–b5). His use of onoma has a wider scope than modern proper-name semantics. The ancient distinction between naming and saying something true or false prepares the later question of how a term's contribution differs from that of a complete assertion. Neither Plato nor Aristotle should simply be redescribed as endorsing a modern causal-historical or Millian theory; the analytical vocabulary and problems differ.

Ancient Chinese debates about ming (names, designations) and shi (objects, affairs, or realities) provide an additional intellectual history of linguistic discrimination and practical coordination. The writings associated with the later Mohists, Gongsun Long, and other disputers examine classifications, naming adequacy, and the relation between linguistic distinctions and things. Their central questions frequently concern general names and kinds rather than the contemporary technical problem of referring proper names. The comparison is illuminating precisely when this difference is preserved: philosophical interest in naming is much older and broader than the specific analytic debate about singular terms (Fraser, 2024).

Medieval reflection on suppositio, signification, appellation, and the properties of terms deepened accounts of how terms stand for things in propositional contexts. The medieval traditions do not supply one uniform doctrine of proper names. They show why a theory of singular terms requires an account of syntactic placement, predication, inference, and reference within a proposition. The historical movement toward modern logic and nineteenth-century semantics brings these issues into a more explicit dispute over connotation, denotation, logical form, and the contents of thought.

Mill and Nonconnotative Denotation

John Stuart Mill's A System of Logic (first published 1843), Book I, Chapter II, especially section 5, distinguishes connotative terms from terms that designate without implying attributes as part of their signification. A general term such as 'white' applies to things in virtue of a property it connotes. A proper name, in his characteristic examples of persons and places, serves as a mark by which speakers identify and discourse about a bearer without describing what the bearer is. The original contrast concerns the semantic role of the name, not the absence of beliefs in the minds of its users (Mill, 1843, bk. I, ch. 2, §5).

The argument may be reconstructed as follows. A connotative term semantically specifies an attribute that determines its range of correct application. Proper names remain applicable to their bearers even when their bearers change or lack particular salient properties. Therefore their distinctive function cannot be explained simply as connoting the properties commonly associated with those bearers. The conclusion permits names to carry rich personal and cultural associations while denying that the associations constitute what the name literally predicates. The insight survives in modern referentialism: information enabling someone to find an object is not automatically part of the semantic value assigned to a name.

The position nevertheless leaves further work. If a name does not contain a descriptive criterion, what establishes its bearer? Mill's examples appeal to the social act of marking or naming; they do not constitute a fully developed account of historical transmission, multiple bearers, unsuccessful naming, or cognitive value. Nor does the classification of proper names as nonconnotative alone tell us the truth-conditional contribution of an empty name. Later developments can be understood as attempts to make explicit what Mill's basic distinction leaves open.

Mill's use of 'connotation' also differs from colloquial English, where a name's connotations may mean emotional or cultural associations. Confusing these senses obscures the central philosophical issue. A name can suggest a family, nationality, literary tradition, or era while lacking those associations as fixed conditions for naming the individual. Preserving the distinction allows the thesis that names are semantically thin to coexist with an account of their social richness.

Frege: Sense, Reference, and Informative Identity

Frege's 1892 Über Sinn und Bedeutung begins from the cognitive difference between identities that appear trivial and those that disclose new information. When two terms name one and the same object, the difference between the terms can still matter to an informed speaker. Frege articulates this by distinguishing Sinn (sense) from Bedeutung (reference, in the translation used here). A name presents its referent in a particular way; the reference alone does not determine the cognitive significance of an identity statement. The German terms matter because translating Bedeutung as 'meaning' in ordinary contemporary English risks conflating the two levels Frege was attempting to separate (Frege, 1892/1948).

The thought experiment is straightforward. Suppose one speaker knows an astronomical body under a morning designation and another under an evening designation, without recognizing that both pick out Venus. The identity statement supplies knowledge by connecting previously distinct modes of presentation. If semantic analysis supplied only the common referent as the entire explanation, it would fail to account for why competent understanding of each name can coexist with ignorance of their identity. Frege's sense is designed to explain that epistemic difference without treating the identity as a report merely about inscriptions.

Frege's theory is more sophisticated than a simple catalogue of privately held descriptions. Sense is an articulated mode of presentation capable of participating in public thought, while speakers may nevertheless associate different routes of presentation with a particular proper name. Whether every name's sense can be replaced by a uniquely identifying definite description is a further thesis; it should not be silently equated with Frege's own semantic distinction. The relation between sense, linguistic competence, and individual recognition remains among the theory's central interpretive questions (Frege, 1892/1948; Zalta, 2026).

Frege also addresses sentences containing expressions without ordinary referents. The distinction between an expression's intelligible sense and its successful reference permits meaningful discourse where reference fails, although the truth-value status of sentences involving empty expressions requires interpretation within his broader semantics. In indirect discourse, where a speaker reports what another thinks or says, customary reference shifts so that expressions can refer to customary senses. The ambition is systematic: one theory seeks to connect informativeness, compositionality, and intensional contexts. Its challenge is to explain how publicly shareable senses are individuated, transmitted, and related to variable historical naming practices.

Russell: Description, Acquaintance, and Logical Form

Bertrand Russell's 1905 paper 'On Denoting' transforms a problem about apparent referring phrases into one about logical form. A definite description such as 'the present king of France' appears grammatically to stand for a person, yet a speaker can understand the expression when nobody satisfies it. Russell's analysis replaces an alleged singular constituent with quantificational structure: for a sentence saying the F is G, there is an F, there is at most one F, and that individual is G. The explanatory achievement is to allow a meaningful proposition without postulating a mysterious object denoted by the unsuccessful description (Russell, 1905).

In a contemporary notation, the analysis of 'the F is G' takes the form ∃x[F(x) ∧ ∀y(F(y) → y = x) ∧ G(x)]. The formula expresses existential commitment and uniqueness at the level of the analyzed proposition. It does not assert that every natural-language occurrence of a descriptive phrase is exhausted by this formula regardless of context. Nor does it follow from the theory of descriptions itself that a proper name has the same logical form as a description. Russell's separate epistemology of acquaintance and his treatment of ordinary names complicate that inference.

Russell came to reserve the strongest form of logically proper naming for expressions connected to direct acquaintance, whereas many ordinary historical names were treated as descriptively mediated. A person can utter 'Bismarck' without acquaintance with the historical individual; on Russell's analysis, the thought available to that speaker may depend on descriptions, testimony, and associated knowledge. This produces a contrast between an idealized logical singular term and everyday proper-name use. Critics argue that this account makes successful reference too dependent on the speaker's stock of identifying beliefs. An ordinary speaker may know enough to use an inherited name while knowing very little about the bearer (Russell, 1911; Kripke, 1980).

Russell's enduring contribution here is methodological. A linguistic string's apparent syntactic category does not determine its semantic or logical analysis; we must determine the truth conditions and inferential behavior of complete sentences. His conclusions about descriptions helped create the precise debate about names, but analysis of descriptions and analysis of proper names must remain separate. Donnellan later complicated the picture further by contrasting referential and attributive uses of descriptions, suggesting that speaker reference and the literal descriptive mechanism can come apart (Donnellan, 1966). This shows why a name's apparent resemblance to a description cannot establish identity of semantic type.

Descriptivism and the Cluster Response

Descriptivist theories hold, in different forms, that the reference of a name is determined by identifying content associated with it. A simple version makes a name equivalent in meaning to a definite description. A speaker's 'Aristotle' might then designate the unique individual satisfying a description the speaker associates with that name. This account can explain how a hearer identifies the intended bearer and why a name participates in informative identities. Its strongest attraction is the unified treatment of meaning, recognition, and reference determination.

The initial difficulty concerns mistaken information. A student might know a writer by a familiar title yet be wrong about when the work was written, where the writer lived, or even which titles belong to that person. We often still judge the student to have been speaking about that writer. Another difficulty is informational poverty: a speaker may refer successfully to someone by relying on a community's expertise, while lacking a description uniquely discriminating that person from every other possible bearer. A third difficulty concerns contingency. If a descriptive trait fixes what the name means, then it may seem built into the identity of the bearer, although the person could have failed to possess it (Kripke, 1980, lectures II–III).

Cluster theories associated with Strawson, Searle, and Wittgenstein's examples attempt to avoid dependence on one uniquely decisive description. The bearer is connected to a family of characteristic beliefs and uses, with enough of the cluster being satisfied for reference to succeed. Searle's 1958 essay 'Proper Names' treats the relation between names and descriptions through a more flexible set of identifying conditions than a rigid synonymy with one definite description. Wittgenstein's discussion of 'Moses' in Philosophical Investigations §79 suggests that the interpretation of a name and the conditions under which a negative existential is accepted may involve a shifting network of descriptions (Wittgenstein, 1953, §79; Searle, 1958).

The cluster response faces a dilemma. If the cluster has no determinate weighting, it may fail to predict when reference is fixed. If weighting is specified so precisely that every case has an answer, the resulting theory may imitate a single complex description and inherit some of its counterexamples. A critic can imagine circumstances in which the distinctive cluster is largely false while a naming practice remains stable. A proponent can reply that cluster information is used to identify the appropriate practice rather than to supply the entire meaning of the name. That reply is important because it changes the view from a semantic synonymy thesis into a metasemantic or pragmatic account. The debate becomes more productive once those claims are distinguished (Cumming, 2023).

Modal Arguments and Rigid Designation

Kripke's Naming and Necessity, delivered as lectures in 1970, published in an article form in 1972, and issued in book form in 1980, transformed debates about names by connecting their semantic behavior to modal evaluation. A rigid designator designates the same individual with respect to each relevant possible circumstance in which that individual exists; descriptions are frequently nonrigid because their satisfier can vary with the world considered. The object designated by a name may have lacked many traits that happen to identify it here. This is a claim about evaluating counterfactual possibilities, not about whether inhabitants of possible worlds use the same spelling (Kripke, 1980; Gómez-Torrente, 2022).

Consider an individual N whom a community presently describes as the inventor of a particular instrument. If N had never invented it, the name would still concern N when used in a counterfactual inquiry. By contrast, the description 'the inventor of that instrument' could designate someone else, or no one, in the counterfactual world. If the name simply meant the description, the two expressions should behave alike under modal operators. Their divergence is evidence against straightforward descriptive synonymy. A sophisticated descriptivist can respond by proposing a rigidified description such as 'the actual inventor of the instrument'; then the question is whether such rigidity adequately explains ordinary linguistic competence and reference fixing.

The familiar astronomical identity illustrates the distinction between necessity and how something becomes known. Once it is discovered that Hesperus and Phosphorus are both Venus, their identity is necessary in the relevant metaphysical sense if both expressions rigidly designate Venus. The route of discovery is empirical. The epistemic fact that the identity was learned a posteriori therefore does not make the identity metaphysically contingent. A necessary a posteriori identity is possible because the mode of discovery and the modal status of the proposition answer different questions. Frege's problem of cognitive difference remains: necessary identity alone does not describe what a learner understood before discovery.

Rigid designation should not be conflated with direct reference. Rigidity specifies invariance of designation across worlds. Direct reference proposes a semantic contribution, often the referent itself, rather than an intermediary descriptive intension. A rigidified description can, in some accounts, be rigid without being directly referential in that latter sense. Likewise, a historical theory of how reference is fixed need not follow from modal rigidity alone. The three theses—rigidity, direct reference, and causal grounding—can be jointly defended, but their logical separation is essential to evaluating the arguments independently (Gómez-Torrente, 2022).

Kripke and Causal-Historical Reference

The causal-historical picture explains reference through the history of a name's introduction and transmission. A name may be initially fixed by ostension or by a description used to establish whom the speaker is naming. Later speakers acquire it from earlier users and intend to continue speaking of the same bearer. They can succeed in referring despite ignorance or error about the bearer, provided their use belongs to the relevant chain. On this model, descriptive knowledge can help initiate or locate a naming practice without functioning as the name's semantic content (Kripke, 1980, lecture III).

The distinction between introducing a reference and transmitting it resolves the apparent paradox of speakers who refer better than they know. A historian can inherit the name of a little-known scholar through books and testimony, while having no uniquely identifying information beyond the testimony itself. Her referential competence is partly dependent on a distributed public practice. The relevant dependency extends through editors, teachers, librarians, documents, and other participants who preserve the name's association with a bearer. The account gives semantic importance to social history rather than treating each speaker as an independent creator of reference.

Still, the causal picture as first sketched is not a complete algorithm. Which event counts as the initial naming if several groups introduce identical forms? What kind of intention suffices to continue using a name for the same bearer? How should names be treated when transmission includes misunderstanding, deliberate reassignment, institutional succession, or strategic impersonation? Kripke explicitly distinguishes intentionally borrowing a name from introducing it anew for an unrelated creature; he does not thereby supply a complete criterion for every borderline case. A theory must explain which kinds of continuity are referentially authoritative and which merely preserve an etymological chain.

The causal-historical approach is frequently combined with semantic externalism: linguistic reference can depend on historical and social facts outside an individual speaker's head. Externalism should be stated carefully. It does not imply that speakers never use descriptions, that reference is always successful, or that a naming ceremony has supernatural authority over all future practice. Its defensible claim is explanatory: public dependence on earlier users can determine what one is talking about even where personal identification is incomplete. The familiar appeal to a causal chain becomes philosophically illuminating when the relevant intentions, acts of deference, and practices of correction are specified.

Evans: Naming Practices and Reference Change

Gareth Evans's 'The Causal Theory of Names' (1973) argues that the history of a name's use cannot be reduced to an indefinitely authoritative first assignment. His discussion of 'Madagascar' supplies the now-standard reference-change case: a name alleged to have concerned a region on the African mainland became established as the name of an island. The historical etymology of the anecdote is debated, and philosophical treatments often employ a simplified narrative as a stipulated example. What matters for the argument is the possibility of a sufficiently entrenched current practice redirecting reference despite continuity in a chain of transmission (Evans, 1973; Sattig, 1998).

The challenge is structural. Suppose a present speaker inherits a name through the same communication line as earlier speakers but the relevant community has systematically applied it to another object. If initial reference invariably controlled every later use, the present speaker's true claims about the newly established object would be semantically about the former object. Yet ordinary reference judgments may favor the current referent. A rigid historical anchor becomes too rigid if it cannot distinguish reference continuity from reference shift. Invoking mere mistake is insufficient once the apparent mistake is corrected and stabilized across generations.

Evans's subsequent account in The Varieties of Reference (1982) develops the notion of name-using practices and the distinction between participants who can independently identify a bearer and those who depend on the practice. Those who directly contribute authoritative identifications can sustain or redirect the practice; dependent consumers can refer successfully without independently fixing its target. The approach makes room for public correction, changing expertise, and distinct practices employing the same form. It gives reference a social structure while preserving a substantive relation to the bearer.

The strength of the practice approach is explanatory flexibility; its vulnerability is the need to identify a practice without already assuming the reference it is meant to explain. If a researcher says that the practice directed at person X determines the referent X, a circularity threatens unless 'directed at' is analyzed through independently observable naming, recognition, transmission, and correction. An adequate account must specify what participants count as authoritative, how their authority is established, and when a new practice has emerged. Recent literature on reference change continues to debate whether historical intention, communal uptake, or locally salient uses should determine the outcome. The 'Madagascar' problem remains a productive test rather than a settled historical proof (Cumming, 2023; Sattig, 1998).

Direct Reference and Singular Propositions

A directly referential account holds that, in a standard successful occurrence, a name contributes its referent to the proposition expressed. On a simple Millian reconstruction, a sentence of the form 'N is F' expresses a proposition whose singular constituent is the bearer of N and whose predicative constituent is the property F. The name need not contribute a descriptive property that uniquely identifies the bearer. This picture aligns naturally with a direct account of modal identity and the idea that two names for the same individual contribute the same singular constituent in the relevant compositional position (Kaplan, 1989; Salmon, 1986).

The claim acquires precision when separated from two others. First, a name can contribute only its bearer in a semantic theory even though historical relations determine which bearer it contributes. Second, a sentence's truth-conditional content need not exhaust everything cognitively conveyed by using it. A speaker may know one individual under two labels without knowing their co-reference; that epistemic difference can be treated by theories of mental representation or pragmatic communication while retaining a referential compositional semantics. The strategy therefore redistributes explanatory responsibilities rather than erasing them.

An objection asks whether two co-referring names really can be interchanged in all cognitive and reportive contexts without a change in what is said. If the contents of 'Hesperus is visible' and 'Phosphorus is visible' are literally the same, what explains a rational person's acceptance of one and rejection of the other? A direct-reference theorist may distinguish the proposition expressed from the mode of entertaining it, the assertoric act, or the information communicated by choosing one form. Each response must explain the cognitive difference without covertly reintroducing descriptive senses as the name's semantic contribution.

A second challenge concerns empty names. If the referent is the only singular contribution, an empty name apparently contributes no object and may prevent a complete Russellian proposition from being expressed. Solutions include gappy propositions, special rules for negative existentials, pretense operators, and contextual reinterpretation. The theoretical economy of direct reference therefore has a cost: phenomena of failure and thought must be addressed in additional components of the overall theory. Whether this is a weakness or an appropriate division of labor depends on how systematically those components fit together (Braun, 1993; Sainsbury, 2005).

The Cognitive Significance Problem After Direct Reference

Frege's informative-identity problem retains force after the rejection of simple descriptivism. If the semantic contribution of each co-referring name is one and the same object, then an identity statement apparently asserts that the object is itself. Yet discovering that the two names co-refer can transform a person's beliefs, predictions, and practical behavior. The phenomenon cannot be explained by the bare identity relation, because the question is why a rational agent did not already know what the identity reveals (Frege, 1892/1948; Kripke, 1980).

One response distinguishes worldly content from cognitive route. An individual's information can be partitioned into distinct files associated with different encounters, documentary traditions, or linguistic forms. Learning an identity connects two files and redistributes their contents. This models the transition in knowledge without requiring the metaphysical object to divide into two. Another response treats the linguistic form or a mode of presentation as part of a secondary semantic or pragmatic level relevant to reports of understanding. The options differ over whether cognitive significance must be encoded in compositional meaning or can be generated by epistemic and communicative mechanisms (Perry, 1977; Recanati, 2012).

The mental-file strategy needs an account of file identity and file merging. A person may mistakenly fuse two individuals into one file, split one individual into two, or knowingly retain separate files because the uses occur in different social settings. A theory that simply postulates a file for every way of thinking risks restating the problem. A promising explanatory account identifies the informational practices that maintain files: perception, testimony, reference tracking, and dispositions to correct beliefs in light of recognized co-reference. Knowledge of identity then changes the organization of information accessible for inference.

This leads to a general result: semantic sameness at the level of object reference does not entail epistemic interchangeability for a thinker, while epistemic difference does not establish different objects. A theory of naming needs both distinctions. Frege locates some of the difference in senses; direct-reference approaches may relocate it in representation and communication. The issue between them is not whether informative identity exists—it plainly does—but where a philosophical theory should locate the explanatory structure that makes informative identity possible.

Belief, Opacity, and the Puzzle of Recognition

Belief reports expose the dependence of rational judgment on modes of representation. From 'N is F' and 'N = M' one may infer 'M is F' in an extensional context. But from 'Alice believes N is F' one cannot simply infer that Alice believes M is F when she is unaware of the identity. This is the familiar problem of substitution in opaque or intensional contexts. The grammatical form suggests that a name occurs in the embedded clause while the explanatory problem concerns the relation between the believer, the proposition ascribed, and the believer's available cognitive routes (Frege, 1892/1948; Cumming, 2023).

Kripke's 1979 'A Puzzle about Belief' intensifies the challenge. A French speaker Pierre acquires one set of beliefs through the name 'Londres' and later acquires apparently incompatible beliefs through 'London', without recognizing the shared city. A parallel case concerns a thinker who encounters the same individual under the two apparent personae of a politician and a musician. The puzzle emerges partly from tempting principles connecting sincere assent, translation, and belief ascription. It persists even when one does not begin by assuming that co-referring names are semantically interchangeable. Thus the puzzle challenges general theories of belief attribution and rationality as well as particular semantics of names (Kripke, 1979).

The possible responses range from Fregean senses and fine-grained propositions to two-factor explanations combining referential content with ways of thinking or contextual standards for reporting belief. A report may be evaluated as an attribution of what a subject believes about an object, as a report of the sentence or mode through which the subject represents it, or as a pragmatic communication to an audience. Distinguishing these tasks is essential: an object-level belief attribution and a characterization of the person's cognitive organization can differ while both remain useful.

The deep issue is what counts as rational coherence when the agent's information is partitioned. An agent may satisfy logical norms relative to each recognized file while violating a norm that would be enforceable only after a hidden identity is discovered. This observation neither solves every belief puzzle nor makes logic relative to arbitrary psychology. It identifies the epistemic constraint that a norm of inference must be applicable to information available under a recognized identity. The philosophy of names thereby intersects directly with the philosophy of mind, epistemology, and theories of propositional attitude.

Names in Modal Logic and Metaphysics

The distinctive work of proper names in modal reasoning becomes clearer when semantic evaluation is made explicit. In a model with a domain of individuals and a family of possible worlds, an interpreted name can be associated with one individual across the relevant worlds while ordinary predicates change their extensions. In a counterpart-theoretic framework the identity of the referent across worlds is treated differently from a strict transworld-identity framework, but the contrast between keeping the subject fixed and varying what is true of that subject remains a core target. The technical choice of model therefore affects the metaphysical interpretation of modal claims (Kripke, 1980; Gómez-Torrente, 2022).

A necessary identity is not the same thing as a necessary existence claim. If N and M are rigid designators of one individual, the identity is necessary on the relevant interpretation, even if the individual could have failed to exist. One must say how names and identity claims are evaluated at worlds lacking the bearer; this is where distinctions among weak, persistent, and obstinate rigidity become relevant. Rigidity should accordingly be defined with attention to the treatment of nonexistence and to the logical background rather than through the slogan that a name 'always refers.' The latter can misrepresent the semantics at worlds in which the object is absent.

Names also participate in debates about origin essentialism. If a particular individual could have developed differently, could the same individual have originated from entirely different parents or materials? Kripke's reflections on origin connect the modal treatment of names with questions of criteria for identity. The name alone does not establish essentialism; modal arguments and intuitions about individual identity do additional work. A philosophical investigation should state what makes a counterfactual scenario concern the same individual rather than merely a qualitatively similar replacement.

These issues intersect with scientific identities and mind–body metaphysics. If a theoretical term and an ordinary name or natural-kind expression are taken to rigidly designate the same object or kind, their true identity may be necessary despite empirical discovery. Yet a theory of proper names cannot derive every scientific or metaphysical identification without independent evidence for the claimed identity and an argument that the expressions actually behave rigidly. Naming helps clarify the logical form of the metaphysical question; it cannot substitute for the empirical or philosophical reasons by which the answer is established.

Empty Names and Negative Existentials

A proper name seems empty when no actual object satisfies the conventions that would otherwise make it refer. Examples include names introduced for mistakenly postulated planets, purportedly historical individuals whose existence is denied, and inventions in fiction. The central problem is that speakers can understand and sometimes apparently know the truth of sentences containing such names even though a straightforward singular-proposition semantics expects an object as their semantic constituent. The phenomenon is especially sharp in a negative existential: 'Vulcan does not exist' appears informative precisely because astronomical inquiry failed to establish the hypothesized planet (Braun, 1993; Cumming, 2023).

Russellian reduction offers one route. If an apparent name in the relevant occurrence abbreviates a description, the negative existential can mean that no unique object satisfies that description. The cost is that different descriptions may yield different verdicts when the speaker uses one form, and the explanation can make the semantic content too dependent on individual associations. A Fregean account can preserve sense without an ordinary referent, but must specify what follows about the proposition or its truth value. Free logics can allow singular terms that fail to denote while restricting existential inference, and different free logics disagree about how atomic sentences containing empty names should evaluate.

A gappy-proposition account allows a sentence with an empty name to lack a normal singular constituent while still supporting some higher-level treatment of its use or truth conditions. A pragmatic account may explain why an utterance communicates a true claim even if no ordinary singular proposition is expressed. A metalinguistic account may instead understand the negative existential as denying that the relevant name-using practice has an actual bearer. These are genuinely different analyses. They should be assessed for the full range of uses, including 'Vulcan was proposed by astronomers,' 'Vulcan was discovered,' and 'There is no Vulcan.' The evidential circumstances relevant to those sentences are different.

There is also a philosophical distinction between a name's failing to refer and our failing to discover its bearer. Historical names can be subject to uncertainty because records are incomplete. Calling them empty before investigation concludes would beg the question. The discipline of naming inquiry therefore includes a methodological norm: report the evidence for reference failure, and distinguish a theory's representation of uncertainty from a metaphysical judgment of nonexistence. This prevents a semantic model from turning incomplete knowledge into an unearned ontological conclusion.

Fictional Names and Literary Discourse

Fictional names are systematically introduced, reused, and understood even where the story does not claim to document an actual individual. A sentence occurring inside a narrative can be evaluated in relation to what the narrative establishes; a sentence about the narrative can assert a fact about a character's literary role. The distinction between 'Holmes is a detective' within the fiction and 'Holmes is a fictional detective created in a literary work' in criticism is philosophically important. The second statement need not be treated as an attempt to report an ordinary crime investigation by a person in the actual world (Thomasson, 1999; Sainsbury, 2005).

Realist theories of fictional entities regard fictional characters as entities of some kind—often abstract artifacts dependent on authors and literary practices. Such views assign a referent to at least some external uses of character names. Antirealist and pretense-based theories instead explain ordinary narrative utterances through implicit operators or activities of make-believe, without treating the characters as ordinary worldly bearers. Other theories recognize different semantic uses of one name across literary and critical contexts. The debate must specify the ontological category under discussion; simply asserting that a fictional character is 'real' or 'unreal' hides the distinction between an actual literary artifact and a person described within a story.

Names in literature may be borrowed from history, transferred among fictional worlds, or deliberately left ambiguous. A novel can call its central figure 'Napoleon' while plainly depicting an imagined historical alternative. Whether the fictional occurrence refers to the historical Napoleon, a character modeled upon him, or both through layered narrative practices requires an account of authorial intention, genre, and reader interpretation. Such cases complicate any uniform procedure that infers reference directly from spelling or original baptism.

The philosophical payoff is wider than literary ontology. Fiction shows that speakers can maintain richly coherent identity-tracking practices when reference to an ordinary existent individual is not available or is not the point of the discourse. A theory of names must therefore separate the success of communicative identification within a practice from the stronger success of singular reference to an actual bearer. The distinction becomes crucial in virtual environments, role-playing systems, fictional universes shared across media, and digital archives where the same names connect many versions of narrative objects.

Names as Predicates and Metalinguistic Expressions

Theories treating names as fundamentally referring singular terms face linguistic evidence that name forms can occur as predicates. 'There are three Emmas in the class' counts persons bearing a name, while 'She is a different Emma' may classify a participant by a naming property. Tyler Burge's work on reference and proper names and subsequent predicativist theories take these constructions seriously as potentially revealing the semantic type of proper-name expressions rather than as isolated accidents of coercion (Burge, 1973; Jeshion, 2015).

A predicativist may assign the nominal form 'Emma' a property such as being called Emma. A complete singular argument then requires a determiner, which may be overt, as in 'the Emma,' or covert, as proposed for ordinary bare-name arguments. The theory offers a compositional connection between counting names and referring with them. Its explanatory burden is to derive the actual differences in behavior between bare names and descriptions with overt determiners. If 'Emma left' and 'the Emma left' have distinct scope possibilities or discourse dependencies, the difference cannot be dismissed as mere stylistic variation (Cumming, 2023; Jeshion, 2017).

The theory also confronts a potential circularity. If being called Emma means being referred to by a specific name whose bearer is Emma, the analysis presupposes the reference relation it was supposed to explain. A defender may ground the property in independent acts of attribution: publicly assigning a label, employing it in address, or treating an individual as included in the practice. This is a substantive advance if the attribution relation can be specified without invoking the very semantic reference being explained. The parallel issue for use-based referential theories is that they too must identify the historically relevant applications of a name without assuming the answer.

Metalinguistic approaches should not be judged solely by the observation that names are used to name. The philosophical question is whether the property of bearing a name is part of compositional semantics, a rule for identifying a practice, a descriptive gloss, or a pragmatic inference. Those claims have distinct empirical and theoretical consequences. A satisfactory account should predict article distribution, anaphora, quantification over name-bearers, and the difference between using a name to refer and mentioning the name as a linguistic expression. The competition between predicativism and referentialism therefore depends on detailed linguistic data as well as on classic philosophical thought experiments.

The Grammar and Diversity of Naming

Naming systems differ across languages, communities, and historical periods. A proper name may include personal and family elements, titles, patronymics, particles, clan identifiers, postposed honorifics, or lexicalized noun phrases. Articles accompany names in some languages and constructions while being absent in others. Transliteration and translation can multiply written forms associated with the same individual. The philosophical consequence is that a name's stable bearer cannot be equated with an invariant graphic string, and a semantic analysis cannot assume that the morphology of English personal names represents all naming systems (Huddleston & Pullum, 2002; Anderson, 2007).

Names also participate in address, introduction, quotation, legal registration, historical narration, and dialogue. Vocative use can address an individual even if the name has been mispronounced or assigned in error; an introductory utterance can institute a new practice; a registry can establish a formal label with consequences for public administration. These uses share linguistic material but differ in illocutionary function. A philosopher who treats all occurrences as instances of a single referential assertion may misdescribe acts through which names are given, contested, or changed.

The line between proper names and descriptions can be blurred by form. Some established names are transparently descriptive in origin, and a name can contain words that continue to have descriptive meanings. A bridge called 'the Old Bridge' may remain the Old Bridge after a newer bridge is demolished. The fact that a phrase contains descriptive material does not establish that its present reference is determined by the material's current descriptive satisfaction. Historical lexicalization can yield a name whose contemporary function has detached from the original naming motivation.

Comparative research needs separate evidence for linguistic facts and philosophical generalizations. A grammatical construction documented in one language may undermine a universally stated syntactic thesis without deciding a disputed theory of meaning. Conversely, a semantic proposal may admit multiple surface grammars. The strongest methodology combines descriptive linguistics with explicit claims about reference, compositionality, and practice. The exact linguistic mechanisms of naming remain a developing interdisciplinary research area, especially where sociolinguistic conventions and formal semantics interact.

Shared Names, Ambiguity, and Disambiguation

The occurrence of one name form with multiple bearers is the ordinary condition of naming rather than a marginal anomaly. A classroom may contain two people called Maya; an archive may contain several authors with the same surname; a city may share its name with a river and a province. A theory must say whether it treats these as many homonymous names, one generic name with context-sensitive reference, or a more complex interaction between lexical form and local conventions. The issue is not resolved merely by pointing out that hearers usually know which Maya is intended (Cumming, 2023).

If each bearer has a distinct specific name sharing a spelling, a hearer's task is to select the relevant specific naming convention. Context, testimony, demonstrative cues, and conversational history can help. If the generic name is instead a context-sensitive expression, its semantic rule must specify how context constrains the relevant bearer. A minimal metalinguistic rule might restrict the candidate set to things called Maya. Yet the rule alone will not choose among several such candidates. Additional salience or convention facts must complete the determination.

Disambiguation is not always a transparent private act. Speakers sometimes refer successfully despite disagreeing over which person a particular nickname denotes; institutions may maintain shared names that are intentionally ambiguous; historical archives can confuse two figures and later separate their records. A name's use can therefore be underdetermined by the local utterance even when a community believes it is clear. Epistemic uncertainty about the intended practice must be recognized in semantic analysis without automatically producing a new bearer.

The distinction between ambiguous names and names with shifting reference is crucial. Ambiguity permits two relatively stable practices to share one form; shift concerns a practice or descendant convention that changes its target over time. Polysemy and metonymy add further possibilities, such as using a city's name for its government or sports team. A mature theory should supply diagnostics for each rather than invoking 'context' as a universal solvent. When context does real explanatory work, the theory should identify the relevant elements—speaker intention, mutual knowledge, conversational salience, institutional records, or inherited practice.

Social Authority, Naming, and Institutional Continuity

Naming is sustained by institutions as well as interpersonal conversation. Birth registries, scientific taxonomies, legal charters, mapmaking authorities, cultural communities, religious practices, and archival standards may confer or stabilize labels. These institutions can affect what counts as an authoritative use without making every authorized use semantically infallible. A state registry might record the official form of a person's name while friends continue to employ a nickname that refers just as successfully. Different practices may coexist, and their authority is purpose-relative.

Legal persons illustrate the distinction between formal naming and ontological identity. The name of a corporation can persist while its officers, assets, and offices change. Conversely, a merger may create a new legal person retaining part of a predecessor's name. What fixes the referent of the institutional name depends on legal continuity, authorized documentation, and social uptake. These facts are not reducible to the visual persistence of a logo or the descriptive properties of an office building. They involve rules of constitution and recognition that connect naming to political philosophy, law, and the ontology of social objects.

Public naming also has ethical and political dimensions. A person may change their name; a community may contest a colonial place-name; an institution may attempt to impose a label on people who reject it. In each case names carry relations of recognition, authority, and power. The philosophical task is to distinguish the semantics of who is being referred to from the normative question whether a naming practice is respectful, justified, or legitimate. The same reference can persist across a morally significant change of label. Conversely, a familiar form may conceal the erasure or replacement of a historically distinct bearer.

For historical research, identity claims across naming records require provenance. Documents with different orthographies and dates may concern one person, while matching names may concern several. Linguistic continuity supplies evidence but is not by itself proof of individual identity. Scholars must integrate dates, contextual records, kinship, institutional roles, and documented transmission. This is a philosophical extension of the central semantic lesson: the relation between a name and its bearer is mediated by publicly examinable practices that can be investigated, mistaken, corrected, or manipulated.

Names Across Media and Artificial Systems

Digital communication brings naming practices into databases, knowledge graphs, virtual communities, program identifiers, and conversational systems. A username can remain stable while the person operating an account changes; one person can control several usernames; a persistent database identifier can refer to a record that is later merged with another. The relation between a display name and a referent is therefore many-layered. A string can function as a label, an account key, a public pseudonym, a formal identifier, or an expression in ordinary discourse. Technical uniqueness of a key is a property of a system design, not a philosophical proof of the named entity's identity.

A knowledge graph may represent a person as an entity node with multiple language labels. This architecture makes explicit a distinction that traditional semantic theory often describes abstractly: the identity of the object represented is separated from the words used to reach it. Yet the node's correctness depends on provenance and record linkage. If two historical individuals are merged into one node, the formal uniqueness of the identifier merely preserves the mistaken merger with precision. The name's practical referential success continues to depend on documentary and communal evidence.

Artificial language systems can produce and interpret proper names in texts, dialogue, and structured data. The philosophical question concerns what determines the reference of a name when an artificial system repeats it, generates a new name, retrieves a record, or participates in a human–artificial naming practice. A model's production of the string 'Maya' is not sufficient evidence of a uniquely determined bearer; contextual anchoring, source continuity, communicative intention in the relevant practice, and identity verification remain necessary questions. Philosophical analysis here concerns semantic roles and evidential relations, and should not confuse a system's output with an independently established metaphysics of its bearer.

The Artificial Era provides a further horizon in which naming, attribution, digital provenance, and historical continuity may become central to relationships among biological and nonbiological participants. For the present classical dossier, this is an analytical extension rather than an occasion to redefine established Aisentica concepts. The precise canon of Artificial and Artificial Sapiens belongs to its own primary philosophical sources. The contribution made here is narrower and defensible: reference in computationally mediated discourse remains accountable to distinct questions about strings, entities, social practices, and evidentially maintained identity. This distinction also clarifies why provenance can matter more than surface resemblance in identifying the subject of a historical or digital record.

Comparative Historical Perspectives

The historical succession from Mill to contemporary theories does not describe simple progress from a primitive view to a final consensus. Mill identifies the difference between naming and attributive predication. Frege articulates the cognitive significance omitted by a purely referential account. Russell demonstrates that surface grammatical singularity can conceal quantificational structure. The descriptivist tradition connects competence to identifying information, while Kripke and Evans explain reference through historical and social dependence. Predicativists return attention to the internal grammar of nominal expressions. Each program isolates a genuine aspect of linguistic functioning and becomes vulnerable when its preferred explanatory object is treated as sufficient for every question.

The comparison reveals several independent theoretical choices. A philosopher may affirm direct semantic reference yet favor a detailed social account of reference determination. Another may reject semantic descriptivism while accepting descriptions as pragmatically important tools of identification. A third may favor predicative syntax while acknowledging that many complete name-uses denote singular individuals. A fourth may retain Fregean senses without asserting that every sense is an explicit definite description in a speaker's mind. The historical debate becomes clearer when positions are classified by what they claim, rather than by broad labels such as 'descriptive' or 'causal' used as complete philosophical identities.

The comparative scope also has limits. Plato's treatment of onomata, the medieval theory of terms, and early Chinese reflection on ming answer questions that overlap only partially with the semantics of modern English proper names. Forcing them into the categories of direct reference, rigidity, or descriptivism would distort their texts. The historically responsible comparison identifies structural affinities—the relations among name, thing, convention, knowledge, and correct use—while preserving differences in problems, languages, and inferential frameworks. It explains how different traditions contribute to a general philosophy of naming without falsely attributing anachronistic doctrines.

The most instructive continuity across traditions concerns the source of a name's authority. Names can be assigned, inherited, contested, corrected, or learned through other people. Their relation to things is neither explained by bare sound nor exhausted by the contemporary speaker's beliefs. Their epistemic force depends on how speakers use them to track and discriminate objects across changing circumstances. This generalization is compatible with several technical semantic theories and therefore functions as a historical synthesis rather than a claim to have dissolved their disputes.

An Integrated Analysis of Naming

A satisfactory philosophical account of proper names should separate four explanatory layers while specifying their relations. The first layer is expression structure: what constitutes a name-form, a proper nominal, a token, or a syntactic occurrence. The second is reference determination: what makes a particular occurrence or convention directed toward an individual. The third is semantic contribution: what the name supplies to the compositional interpretation of a sentence in the context under consideration. The fourth is cognitive and social uptake: what speakers know, misunderstand, communicate, remember, and correct when using the name. This articulation organizes competing theories without assuming that one layer can be deduced from another.

The argument begins with successful reference amid informational poverty. It supports the claim that individually possessed identifying descriptions are not universally necessary for reference. Informative identities then show that reference alone does not capture every cognitively significant distinction. Modal evaluation shows that identification by actual properties and designation across counterfactual circumstances can come apart. Reference-change cases show that a bare historical chain, without a theory of authoritative use and uptake, may fail to predict established present reference. Empty and fictional names show that linguistic naming practices can remain intelligible when ordinary singular reference is disputed or fails. Predicative uses show that a theory restricted to bare singular occurrences omits important grammatical evidence.

The reasoning has a constructive consequence. For a name-token n in context c, let P identify the historically and socially relevant naming practice, and let R(P,c) designate the bearer, where the practice and context warrant one. A compositional semantic account can then choose whether the contribution of n is R(P,c), a mode of presentation connected with it, or a predicate obtained from a metalinguistic property. That formal choice does not automatically determine P. The practice must be recovered using evidence about naming acts, transmission, correction, and interlocutor intentions. Conversely, identifying P does not by itself decide how a particular occurrence behaves in fictional or intensional contexts.

The proposed integration is therefore a research architecture rather than a newly declared theory of the meaning of every name. It imposes obligations on semantic proposals: identify the linguistic object analyzed; state the metasemantic facts that fix the intended bearer; show the compositional consequence; explain cognitive difference and reference failure; and specify whether the account survives social and historical change. It also imposes constraints on historical scholarship: cite the primary argument rather than retroactively turning a thinker into a proponent of a later school; distinguish theoretical reconstructions from an author's actual formulations; and acknowledge when a celebrated example is a hypothetical simplification whose historical details are debated.

The integration clarifies a further distinction between a bearer and a record of the bearer. Archival entries, digital identifiers, scientific catalogues, and personal memory can all supply routes of access to an object, yet errors in those routes do not automatically alter the object. A name's historical persistence is consequently an evidential achievement: successive users preserve, repair, or reestablish its connection to a subject. This conclusion connects the classical semantics of proper names to questions of testimony, institutions, and digital provenance while keeping the monograph's central subject fixed.

Contemporary Disputes and Open Questions

Contemporary philosophy of names remains divided over the relation between semantics and metasemantics. One open problem is whether direct reference can be maintained together with a genuinely explanatory theory of cognitive difference. A second concerns how much descriptive information, if any, is necessary for competent participation in a naming practice. A third concerns the semantics of apparent proper names in ordinary quantificational and predicative constructions. These questions can be pursued with data from actual languages and experiments in linguistic judgment, but experimental results still require careful interpretation in light of rival theories (Cumming, 2023; Jeshion, 2015).

The problem of reference change has renewed force in recent philosophical work. Historical intention, public acceptance, and locally salient patterns of use may diverge, making it unclear when the inherited referent gives way to a new one. The stylized 'Madagascar' case continues to motivate inquiry into conflicting intentions and historical continuity. Two studies published in 2026 directly address reference change. Al-Khalfa (2026) analyzes the distinction between genuine semantic change and misidentification, examining cases that can generate overprediction or underprediction of reference shifts. Koch and Raimondi (2026) develop an account of shifting dominant intentions in the Madagascar case, connecting speaker-level disposition and community transmission. Their explicit proposals make the contemporary debate a question of competing explanations rather than an appeal to the bare possibility of historical confusion. The corresponding philosophical questions concern the relation between semantic authority, historical transmission, and collective correction (Sattig, 1998; Al-Khalfa, 2026; Koch & Raimondi, 2026).

There is also an unsettled debate about what data best discriminate theories. Intuitions about a hypothetical naming case can vary with background assumptions; syntactic patterns can favor one compositional analysis without establishing its metaphysical interpretation; and ordinary speakers may respond differently to reference-fixing and reference-transmission situations. The discipline benefits when researchers specify the predicted difference and explain why the observation supports one theory over another. Philosophical reflection, linguistic investigation, and social history are complementary only where their evidential contributions are kept distinct.

Further questions concern names of groups, legal persons, changing political boundaries, digitally constructed agents, and identities whose records are fractured across languages or media. The inference from repeated name-form to individual identity is especially precarious in automated systems, where identical labels can be generated independently and unique identifiers can preserve inherited errors. Philosophical semantics can supply conceptual controls for such systems, but implementation claims need independent technical testing. Conversely, empirical success in entity resolution cannot by itself determine the correct metaphysical interpretation of a person, a place, or an institutional object.

The future research program is thus specific. It requires improved accounts of practice individuation and reference transfer; a compositional semantics attentive to the full grammar of proper names; a theory of cognitively informative identity compatible with what speakers actually know; principled handling of fictional and empty occurrences; and explicit criteria for historical continuity in digital settings. These tasks are connected through naming, yet each has its own standards of evidence and argumentative success.

Conclusion

Proper names occupy a foundational position in philosophy because they connect the act of speaking to the persistence of particular subjects through changing beliefs and circumstances. They also make visible the limits of treating semantic meaning, reference determination, identification, and social practice as one undifferentiated relation. The historical record shows how different theories exposed these dimensions: Mill separated naming from connotative predication; Frege explained informative identity through sense and reference; Russell revealed the logical complexity of apparent singular expressions; descriptivists investigated identifying information; Kripke articulated rigid designation and historical reference transmission; Evans emphasized practices and the possibility of reference change; and subsequent philosophers developed singular-proposition, predicative, metalinguistic, fictional, and free-logical alternatives.

The central philosophical result is a disciplined division of explanatory labor. Naming practices can establish or transmit a bearer even when individual speakers possess incomplete or mistaken information. Semantic reference alone does not account for every difference in cognitive perspective. Modal stability of designation does not establish a complete social theory of naming. Empty and fictional uses require a theory of the context and kind of discourse in which a name is deployed. Grammatical evidence about predicative and quantified names demands analysis in its own right. These conclusions form a coherent framework for assessing particular semantic theories without declaring an unwarranted universal solution.

The outstanding task is to determine which explanatory layers belong inside the semantics of names and which are best understood through metasemantics, epistemology, pragmatics, and the history of linguistic practices. The question is consequential for theories of identity, modality, belief, institutions, and digital reference. Names preserve the possibility of sustained discourse about individuals by linking linguistic forms to bearers through intelligible, revisable, and publicly maintained practices. Their philosophy is the systematic study of that linkage and the obligations incurred whenever we claim to know whom—or what—we are naming.