Nonexistent and Fictional Entities

The Problem and Its Ontological Scope

People search for legendary cities, fear imaginary dangers, dispute the characteristics of fictional detectives, and correctly deny that a purported planet exists. Such practices make the ontology of nonexistent and fictional entities unavoidable. If the terms in these judgments denote objects, philosophers must explain what sort of objects they denote. If they fail to denote, philosophers must explain how the judgments remain intelligible, informative, and sometimes true. If literary characters are objects of a special cultural kind, philosophers must explain their origin, identity, and dependence on human activity. These demands overlap, but they are not identical.

Consider five judgments. “Vulcan does not exist” corrects an unsuccessful astronomical hypothesis. “The golden mountain is golden” can express the contents of a description without reporting a geological discovery. “Sherlock Holmes is a detective” is ordinarily accepted in a literary context. “Sherlock Holmes is better known than many actual detectives” belongs to critical discourse about a fictional character. “A round square cannot be drawn in Euclidean space” concerns an inconsistent specification. Any account that treats all five sentences as straightforward reports about the same kind of entity has failed to identify the different explanatory tasks.

The leading ontological question is therefore conditional: what entities, if any, must there be for these representational, semantic, and cultural practices to work as they do? A second question concerns semantic dependence: must a meaningful singular term have a referent, or can its contribution to thought and discourse be explained through sense, description, pretense, conceptual role, or an intentional act? A third question concerns philosophical method: should an ontology count every entity mentioned in competent discourse, or only entities required by its best literal theory?

The phrase “nonexistent object” is itself contentious because ordinary quantification appears to imply existence. Under standard classical first-order semantics, saying that something has a property ordinarily requires a value in the domain of quantification. Meinongian theories may distinguish an object's being from its existence, and some theories employ existence-neutral quantification. These are substantive choices about the interpretation of logic, not verbal licenses for adding unusual inhabitants to reality. The modern research landscape is mapped in Nonexistent Objects and Fictional Entities in the Stanford Encyclopedia of Philosophy.

This dossier treats the ontology of fictional entities as a focused case within the broader problem of nonexistent objects. It does not assume that all fictional characters fail to exist in every sense. An artifactualist can deny that a detective corresponding to Holmes walked the streets of London while affirming that a culturally constituted character named Holmes exists. An eliminativist can explain the same practice without accepting a character as an additional object. Their disagreement requires exact formulation at the level of reference, inference, and metaphysical cost.

Kinds of Apparent Nonexistence

Failed empirical identification

An unsuccessful posit is introduced in an attempt to describe an independently existing object. The nineteenth-century hypothetical intra-Mercurial planet Vulcan is a standard philosophical example. Its intended role was empirical: to identify a physical body. The failure of that hypothesis does not retrospectively create a literary character, and a linguistic history of the name does not establish the planet's existence. It establishes a history of investigation and failed identification.

This distinction matters because a fictional name usually has a different introduction. An author may introduce a detective in a narrative with no intention that readers search for a corresponding real detective. The same surface grammar can therefore conceal different conditions of success. To discover a planet is to find the object demanded by the hypothesis; to introduce a fictional character is to institute a representational practice whose success can occur without discovering a person of the described kind.

Incomplete and inconsistent descriptions

The expression “a golden mountain” is compatible with many possible completions. It supplies a pair of attributes while leaving height, location, origin, and countless other features unspecified. An expression of the form “round square,” understood as requiring contradictory geometric properties in the same respect, presents a different case. The first lacks an actual instance without its description being intrinsically inconsistent; the second may fail even to describe a possible instance under the relevant geometry. A theory that assigns objects to both must explain why incompleteness and impossibility do not collapse its logic.

One can discuss the mathematical inconsistency of a specification without assuming that an impossible object literally exemplifies its incompatible properties. The inference from meaningful specification to objecthood is exactly what a general theory of objects must defend. Russell's famous challenge to naïve object theories targets this unrestricted transition, and later neo-Meinongian approaches restrict properties or distinguish different ways of having them (Meinong, 1904; Nonexistent Objects, §4–5).

Past objects, future possibilities, and abstract items

A deceased person does not presently exist as a living organism, yet biographical statements may still refer to that historical individual. A possible future building may never be built; its design may nonetheless exist. A mathematical object, if accepted, is not a concrete object at all. These cases show why the term “nonexistent” should not be assigned indiscriminately to everything absent from the current field of perception. Temporal absence, modal nonactuality, abstractness, and fictional status make different metaphysical claims.

The research question is strongest when it identifies a specific failure of instantiation or reference. A sentence can fail to describe an actual concrete object while successfully referring to an abstract artifact, a text, a concept, or an event in intellectual history. Such alternatives require independent argument: describing a sentence's use does not automatically prove the existence of its alleged referent.

Ancient Foundations: Being, Difference, and Representation

The ancient philosophical background concerns the possibility of meaningful discourse about what is not. In Plato's Sophist, the Eleatic Stranger confronts a difficulty inherited from the stringent opposition between being and nonbeing. If the false statement asserts what is not, how can false assertion possess an object or articulate a determinate content? Plato's analysis of nonbeing as otherness supplies a crucial answer. At Sophist 257b–258d, negation is understood through difference: to say that something is not beautiful need not posit an independent entity called Nothing. The negative predicate identifies a relation of difference from what is beautiful. The argument forms part of Plato's wider account of the combination of kinds and the possibility of false speech (Plato, Sophist 257–258).

This was not a twentieth-century theory of literary characters. Its enduring contribution lies in an argumentative distinction. A negative judgment can possess intelligible content without introducing nonbeing as a mysterious object opposite to being. Contemporary deflationary accounts of empty reference likewise seek to explain negative discourse by analyzing propositional structure. The analogy is methodological, and the historical theories retain their separate ontological assumptions.

Aristotle's account of poetic representation adds another foundation. In Poetics 9, at 1451a36–1451b11, Aristotle distinguishes the historian's concern with what has happened from poetry's concern with what could happen according to probability or necessity. Poetic construction can present intelligible sequences of action independently of whether their represented agents existed historically. This passage concerns plot, universality, and mimēsis; it does not settle the ontology of characters. It nonetheless clarifies why a representation can be assessed for coherence, typicality, and explanatory power without being assessed as a factual chronicle of particular persons.

Together these ancient arguments identify two dimensions that modern debates often separate: the possibility of significant negative speech and the possibility of intelligible representation without actual exemplification. Neither requires a single universal ontology of all objects of thought. The historical lesson is that truth, representation, and reference must be analyzed together while retaining their distinct criteria.

Intentionality and the Directedness of Thought

Franz Brentano's Psychology from an Empirical Standpoint (1874, book II, chapter 1) makes intentionality central to the characterization of mental phenomena. A presentation is a presentation of something; fear is fear of something; judgment concerns something. The traditional expression “intentional inexistence” raises an immediate interpretive danger. It names a manner in which an object is implicated in an act; the phrase by itself does not prove that every act is directed toward a peculiar nonexistent thing. The historical text and subsequent interpretations are surveyed in Intentionality.

The issue arises sharply when a person fears a mythical creature. The fear is psychologically real and may have measurable consequences. Yet the feared animal need not exist. The simplest relational model, in which every genuine mental relation must contain an independently existing object as its second term, appears to fail. Three broad repairs become available. First, alter the metaphysics by recognizing intentional objects. Second, alter the semantics of intentional attribution by treating the verb phrase as a representational construction that need not denote a real relation. Third, identify the relata as existing representations, contents, or practices rather than the putative creature. Each repair preserves a different aspect of the original intuition.

A useful test concerns substitution. Someone may believe that the morning star is bright while failing to believe the same about the evening star, even though both terms designate Venus. The truth of an intentional attribution depends on how the target is presented, not merely on the identity of its referent. Empty-name contexts intensify this opacity because there may be no ordinary referent with which to begin. A credible analysis of intentionality must explain both directedness and differences in cognitive significance.

Tim Crane's The Objects of Thought (2013) revisits this broad problem, arguing that understanding the mind's representational power requires taking its capacity to represent what does not exist seriously (Oxford University Press). The book is relevant as a contemporary systematic intervention; the present dossier relies on its publisher's abstract and the broader accessible debate rather than claiming independent inspection of every chapter. The central methodological question remains whether the “object of thought” must be understood as an ontological item or as a way of specifying what the thought represents.

Frege: Sense Without Reference

Gottlob Frege's “Über Sinn und Bedeutung” (1892) transformed the problem by distinguishing Sinn, usually translated as “sense,” from Bedeutung, usually translated as “reference.” Two expressions can have the same referent while presenting it differently. Frege uses the morning star and evening star to show that an informative identity differs from a trivial repetition. The sense of an expression contributes a mode of presentation through which a referent is determined, if there is one (Frege, 1892, original publication record; English translation).

Frege explicitly considers the name of Odysseus. A reader can understand the thought expressed by a sentence about Odysseus while questioning whether the name has a referent. This establishes an important theoretical opening: an intelligible expression need not achieve successful ordinary reference. In a literary context the reader can follow relations, expectations, and narrative development without requiring an actual bearer corresponding to the name. Frege also connects the reference of a complete declarative sentence to its truth value in the mature semantic theory. Sentences containing empty names therefore pose special problems for the transition from meaningful thought to truth-evaluable reference.

Frege's distinction provides a principled explanation of meaningful empty terms, but it leaves questions. What makes the sense associated with a name sufficiently determinate for different readers to grasp it? How should a truth-value be assigned to a sentence containing a name without reference? Can literary interpretation be explained by a common thought alone, or does discourse about a character require an object of literary criticism? Answering these questions needs more than a repetition of the sense–reference distinction.

The principal achievement is the decomposition itself. The semantic accomplishment of understanding a name and the metaphysical success of finding its referent are separate matters. Later theories disagree about how far this separation can go. A direct-reference theorist may reject the idea that a proper name simply abbreviates a description, while still acknowledging that fictional names participate in organized discourse. Frege provides the starting problem and an influential solution, not the final consensus.

Meinong and the Independence of So-Being

Alexius Meinong's “Über Gegenstandstheorie” (1904) proposes a distinctive inquiry into the object considered as such. The original German text begins from the breadth of judgment and presentation: one judges something, even where the object of judgment does not exist. Meinong's Gegenstandstheorie concerns the range of objects that can be intended, described, or investigated without restricting inquiry in advance to actual existents (Meinong, 1904, German transcription; 1904 collection, catalog).

A guiding maxim is that an object's Sosein, its so-being or determinate character, is independent of its Sein, its being. The golden mountain can be specified as golden and mountainous regardless of whether it exists. This independence promises a unified treatment of intentional targets, mathematical or logical constructions, counterfactual descriptions, and fiction. It also threatens to generate objects corresponding to mutually incompatible property specifications. The round square is not an accidental embarrassment; it tests the limits of the very generality that gives object theory its explanatory appeal.

Meinong's position has a developmental history. His account cannot be fairly reduced to the naïve unrestricted rule that every collection of predicates is exemplified by a concrete object. Later refinements distinguish constitutive properties, considerations of existence and possibility, and forms of incomplete or impossible objecthood. Scholarship accordingly separates the original theory from revised Meinongian programs. A textually responsible reconstruction distinguishes the project of studying an intended object independently of its existence from any one formal axiom system attributed to that project (Nonexistent Objects, §2 and §5).

The most attractive Meinongian argument proceeds as follows. A meaningful intentional act must be directed toward something. Some intentional acts are directed toward what does not exist. Therefore, the range of what can be an object of an act exceeds the range of what exists. Its strength depends on the first premise. If directedness can be analyzed as a feature of content rather than a relation requiring an object, the inference to nonexistent objects is blocked. The second premise is widely accepted as a description of human practice; the metaphysical conclusion remains disputed.

Meinong's challenge survives even where his own answer is rejected. Any opposing theory must give a positive explanation of the predication, identification, and inferential behavior that makes a nonexistent target seem available to thought. Calling such discourse merely mistaken is inadequate when the discourse supports systematic judgments that distinguish one imagined target from another.

Russell and the Logic of Denoting

Bertrand Russell's “On Denoting” (1905) provides one of the decisive alternatives. A definite description need not function as a genuine name of an object. A sentence containing “the present King of France” can be analyzed through quantification, identity, and predication rather than through an alleged individual corresponding to the phrase. Russell's paper appeared in Mind 14(56), 479–493, with the original DOI 10.1093/mind/XIV.4.479.

Consider “The present King of France is bald.” In a familiar Russellian reconstruction, let Kx mean that x is presently King of France and Bx mean that x is bald. The assertion has the form:

There is an x such that Kx; every y that is K is identical to x; and Bx.

This decomposes the grammatical singular expression into an existence claim, uniqueness claim, and predication. Since the existence condition fails, the whole assertion is false under the ordinary classical interpretation. It does not require a mysterious king to possess the property of nonexistence. The negative sentence “The present King of France does not exist” can be analyzed as the denial that a unique present King of France exists.

Russell also exposes a scope distinction. “The present King of France is not bald” may be read with negation inside the description's asserted existence-and-uniqueness scope or outside the entire quantified assertion. One reading says that there exists a unique king who is not bald; the other denies that there exists a unique king who is bald. These differ precisely when the description has no satisfier. A semantic theory that misses the distinction can falsely diagnose a contradiction where the speaker merely denies the existence of a king.

The theory's interpretive reach has limits. Russell's analysis is explicitly about denoting phrases, while many literary names appear syntactically as ordinary proper names. To apply the theory to fictional names one must either treat such names as disguised descriptions or develop a further semantic paraphrase. That extension is controversial, especially after theories of direct reference and causal-historical naming. Russell's lasting contribution is to show that surface grammar is unreliable evidence of ontological commitment; the existence of a nominal phrase does not by itself license the existence of its seeming object.

A further dispute concerns Russell's criticism of Meinong. The charge that a golden mountain specified as existent must therefore exist targets an unrestricted bridge from included properties to actual exemplification. Revised object theories challenge that bridge through restrictions or multiple kinds of predication. Russell's objection remains serious because the stronger a theory's formation principle is, the more exact its treatment of existence and impossible properties must become.

Quine, Carnap, and Ontological Commitment

W. V. Quine's “On What There Is” (1948) reframes the problem through the commitments of a theory rather than the immediate appearance of a name. Quine's famous encounter with the purported entity Pegasus dramatizes the temptation to infer being from the possibility of denial. A person can truly deny the existence of Pegasus without first accepting Pegasus as an item over which an unrestricted quantifier ranges. Ontological commitment is to be determined from the values required by the bound variables of an adequately regimented theory (Quine, 1948, publication record; Quine bibliography).

This criterion is powerful as a discipline of clarity. It requires a theorist to display the logical form under which talk about characters, mathematical entities, or merely possible things becomes necessary. The criterion is not a complete metaphysical proof that no such objects exist. One must still decide what the best interpretation of the relevant discourse is and which paraphrases preserve its explanatory functions. An artifactualist may willingly quantify over fictional characters in a theory of literature; a nominalist may seek an equivalent theory without them. Quine supplies a criterion for locating the commitment, not an automatic adjudication between rival analyses.

Rudolf Carnap's “Empiricism, Semantics, and Ontology” (1950) offers a different diagnosis. Questions of existence may be asked internally to an accepted linguistic framework, or externally concerning the choice of the framework itself. Within a framework of numbers, certain numerical existence claims can be straightforwardly true. The pragmatic question whether to adopt such a framework should not be confused with an empirical investigation into a newly discovered population of numbers (Carnap, 1950, Revue Internationale de Philosophie 4(11), 20–40).

Applied to fictional discourse, this distinction suggests that the truth conditions of sentences under a literary or critical practice need not settle an unrestricted metaphysical question. A framework for discussing characters can be indispensable to criticism without its every quantified form being read as a primitive claim about mind-independent objects. The philosophical cost of this maneuver is explaining how frameworks relate to each other, why some ontological claims seem to transcend them, and which standards make a framework more than a convenient vocabulary.

The Quine–Carnap dispute reveals a structural tension. Regimentation makes commitments explicit; the choice of regimentation can itself reflect pragmatic, semantic, and explanatory decisions. The ontologist cannot simply point to ordinary language as proof, and the eliminativist cannot simply stipulate a paraphrase as success. Both must defend the adequacy of their theory of practice.

Empty Names, Negative Existentials, and Free Logic

An empty name is a singular expression without the ordinary referent demanded by a particular semantics. Empty names appear in unsuccessful scientific theories, fiction, mythology, and erroneous descriptions. The same name may operate differently across contexts, so “empty” requires specification relative to a semantic interpretation. A historically introduced fictional name may be empty as a name of a concrete person and nonempty under a theory in which it names an abstract literary artifact.

Classical first-order logic ordinarily treats every denoting constant as assigned an element of its domain. Consequently, interpreting a fictional name as a genuine constant can produce existential commitment as a side effect of the formalism. Free logics revise this background assumption. In a negative free logic, for example, atomic predications containing a non-denoting name may be false, while a specially defined existence predicate records successful denotation. Positive and neutral free logics adopt other truth-value treatments. The conceptual point is that ordinary inference rules permitting existential generalization from any atomic predication must be reconsidered when names can be empty.

Let E!(a) mean that the name a denotes an existing object under a chosen system. Then the inference from F(a) to the existential claim that some F exists requires assumptions about a's denotation and the semantic status of F(a). A logic that permits non-denoting singular terms must not silently preserve every classical rule unchanged. Such technical revisions are motivated by a general philosophical distinction: the grammatical availability of a singular term does not itself establish a value in the relevant existential domain.

Negative existentials form a useful stress test. If “Vulcan does not exist” is true, what supplies its truth conditions? A Russellian paraphrase denies the satisfaction of a description; a Fregean approach recognizes a sense without a referent; a free-logic approach permits a denotation-failure claim; an intentional-object theory interprets the name as targeting an entity that lacks existence. These are rival explanatory mechanisms, not alternative spellings of one logical result.

Further puzzles arise with fictional identity statements. “Holmes is Holmes” appears trivial, but “Holmes is the detective introduced in A Study in Scarlet” involves historical and literary identification. “Holmes is not Poirot” may be accepted by readers without invoking a physically instantiated detective. To represent the inference one must specify whether identity is applied within the story, between character objects, between linguistic roles, or between artifacts established in different narratives. Logical elegance is achieved only when the chosen interpretation makes these levels explicit.

Neo-Meinongian Objects and Dual Predication

Terence Parsons's Nonexistent Objects (1980) develops a modern theory with resources unavailable to a simple unrestricted property-bundle model. A central device distinguishes nuclear properties, which help determine the nature of an object, from extranuclear properties such as existence, impossibility, fictional status, and other predicates that do not simply contribute to an object's character in the same way. A golden mountain may be characterized by ordinary mountain-like properties without having the extranuclear property of existence. This allows the theory to block the inference from an object characterized as existent to its actually existing (Parsons, 1980, Yale University Press; Nonexistent Objects, §5).

The nuclear/extranuclear distinction has an explanatory advantage: it identifies the mistake in a paradox rather than merely declaring paradoxical objects beyond discussion. It has a corresponding burden. The theory must give a nonarbitrary criterion for the classification of properties, explain how ordinary linguistic predicates select one category, and justify its unusually broad object domain. If an entity is introduced simply because any specification can be made, the metaphysical basis for individuating similar specifications becomes problematic.

Edward N. Zalta's Abstract Objects: An Introduction to Axiomatic Metaphysics (1983) develops a different distinction, influenced by the Mallyan tradition. Abstract objects encode properties rather than ordinarily exemplifying every property they encode. A fictional character can encode being a detective while failing to exemplify that property in the way an actual detective does. This distinction can represent fictional descriptions without transferring them into the actual world (Zalta, 1983, publisher record; author's book outline).

The theory makes a precise philosophical proposal: one must distinguish what an object represents from what it instantiates. Yet the encoded/exemplified distinction cannot do all the explanatory work alone. The theorist must also explain which abstract objects are admitted, which properties they encode, and how fictional practices determine the relevant object. In particular, the mere existence of an abstract object corresponding to a description does not establish that a literary author created that very abstract object in the ordinary historical sense.

Graham Priest's Towards Non-Being (2005) develops an intentionally broad framework for intentionality and nonbeing, rejecting the assumption that quantification must automatically carry ordinary ontological commitment. The theory explores how thought can be directed toward items not found in the actual world, including impossible cases (Priest, 2005, Oxford University Press). Its formal ambition shows that mainstream classical interpretations do not exhaust the options, but it intensifies the question of how logical vocabulary is connected to metaphysical reality.

Across these approaches, the strongest common insight is that distinct forms of predication or quantification may be needed to capture the difference between a thing's being described and its being instantiated. Their major divergence concerns what is genuinely added to ontology. Parsons's nonexistent objects, Zalta's abstract objects, and Priest's intentionally neutral quantifiers are not interchangeable merely because all permit systematic discourse about fictional or impossible content.

Possibility, Impossibility, and Worlds of Representation

The possible-worlds approach appears to offer an easy solution: fictional stories describe worlds, and a character exists in the worlds that satisfy the story. This can be useful for formalizing what could have been the case. It nevertheless risks conflating a fictional character with a merely possible person. If a detective could actually have lived in London, it does not follow that some possible London resident is literally the fictional Holmes. A fictional character may remain the same across adaptations that alter nationality, age, setting, or profession; the conditions of character identity differ from those of ordinary transworld identity.

The distinction also bears on impossible fiction. Stories can contain inconsistent, logically impossible, or internally indeterminate events. An ordinary possible-world model may have no world at which every inconsistent detail of a story is true. A theory can respond by restricting the selected contents, employing impossible worlds, allowing contextual inconsistency, or treating fictionality in terms of prescribed imagination rather than exhaustive representation by possible worlds. Each repair should be evaluated for inferential discipline and expressive power.

A story is generally incomplete. It may settle the protagonist's profession without settling how many hairs were on the protagonist's head. Classical possible worlds are maximally determinate descriptions under many semantic frameworks; fictional stories rarely are. The move from narrative information to a complete world therefore adds a completion procedure. The procedure may be legitimate for selected semantic tasks, but its results should not be mistaken for facts literally supplied by the text.

David Lewis's “Truth in Fiction” (1978) develops influential possible-worlds analyses of the fiction operator. A statement true according to a fiction can be understood through suitable worlds compatible with what is explicitly narrated and with an appropriately selected background of facts or beliefs. Lewis introduces different analyses because the choice of background is itself philosophically consequential (Lewis, 1978, American Philosophical Quarterly 15(1), 37–46). The important lesson is not that fictional worlds are established as concrete universes, but that counterfactual and intensional semantics can model systematic narrative inference.

Modal fictionalism makes the relation between representation and ontology even more explicit. It treats discourse about possible worlds as having a useful fictional or representational role without accepting possible worlds as the objects the discourse appears to name. This is a theory about modal ontology, not straightforwardly a theory that literature and mathematics have the same semantics (Modal Fictionalism). Philosophers must separately specify which claims are accepted literally, which are paraphrased, and what grounds the success of the representation.

The Ontology of Fictional Characters

Eliminativism and paraphrase

An eliminativist about fictional characters denies that ordinary literary practice commits us to such objects. The theory may explain “Holmes is a detective” as shorthand for “According to the relevant stories, a detective with those attributes is represented.” Critical claims about a character's fame can be paraphrased as claims about texts, readers, references, adaptations, and their relations. The ontology includes existing people, works, inscriptions, and social practices rather than an additional fictional detective.

This program has substantial merit. It makes visible the real activities that sustain fiction and avoids the automatic multiplication of objects whenever a narrator invents a name. Its difficulty is systematic paraphrase. “Holmes has influenced later fictional detectives” and “Holmes is more famous than some historical detectives” combine narrative identity with historical cultural effects. An eliminativist must provide an interpretation that preserves the conditions under which such judgments are true, their inferential connections, and their role in literary explanation.

Realism about characters

A realist treats fictional characters as objects in some nonordinary category. Peter van Inwagen's “Creatures of Fiction” (1977) develops a particularly influential challenge to eliminating all reference to fictional characters from literary-critical discourse. The objects of literary discussion can be posited as theoretical entities while the depicted persons remain absent from the physical world (van Inwagen, 1977, American Philosophical Quarterly 14(4), 299–308). This argument should be understood as a proposal about the ontology required by interpretation and criticism, not as evidence that fictional detectives possess flesh-and-blood bodies.

Saul Kripke's posthumously published Reference and Existence: The John Locke Lectures (2013) similarly distinguishes fictional entities as objects of discourse from the concrete individuals represented within fiction. The book develops the significance of historical introduction and the distinction between a character's existence as a fictional character and its existence as a person (Kripke, 2013, Oxford University Press). This creates a compelling challenge to theories that infer total nonexistence simply because the person portrayed by the fiction never lived.

The realist's explanatory strength is direct treatment of comparative literary statements. If characters exist as abstract entities, a critic can quantify over them, compare them, and attribute historical influence to them without elaborate paraphrase. The metaphysical cost is an account of their category, creation, persistence, and identity. Treating the character as an abstract object cannot be merely a terminological victory. The theory must explain what makes one such object the character in one novel rather than a similar figure in another.

Abstract-artifact theories

Amie L. Thomasson's Fiction and Metaphysics (1999) argues that fictional characters can be understood as abstract cultural artifacts dependent on authorial activity and literary works. Unlike timeless abstracta, these characters have historical origins and depend on the practices that bring literary works into existence (Thomasson, 1999, publisher description; chapter on abstract artifacts).

This approach locates fictional entities among familiar cultural objects, such as musical works and institutions, which likewise may be created, transmitted, identified, and maintained through practices. It also explains why a particular work's origin and its author's action matter: the creation of the character is a historical event even when the represented adventures are not. A fictional character can exist as an artifact while not existing as the concrete person portrayed in the narrative.

The abstract-artifact view must specify its dependence relation. An author can imagine a person without establishing a publicly identifiable fictional character. A character can persist after the author's death and through translations, new editions, adaptations, and substantial textual changes. The theory must identify which relationships of origin, recognition, and transmission are necessary, which are contingent, and which permit survival. Its strength is that these questions are placed inside a general ontology of culture rather than treated as supernatural exceptions.

Phenomenological literary ontology

Roman Ingarden's The Literary Work of Art, originally published in German as Das literarische Kunstwerk (1931), is a significant antecedent to theories of dependent literary objects. Ingarden investigates the stratified structure of the literary work, the role of represented objectivities, and places of indeterminacy that readers may concretize. The 1973 English translation by George G. Grabowicz should be distinguished from the original publication (Ingarden, Stanford Encyclopedia of Philosophy).

Ingarden's approach emphasizes that the ontological status of what is represented cannot be extracted from isolated proper names. A character belongs to a work with linguistic, semantic, and aesthetic organization. The represented object has determinations and indeterminacies governed by the work's structure. This perspective anticipates contemporary attention to interpretive dependence without simply identifying the story's character with the author's psychological image or the reader's private experience.

The historical comparison is illuminating. The Meinongian asks what objects may be intended independently of their existence; the artifactualist asks what culturally dependent entities are created; Ingarden asks how a work's structure makes represented objectivities available to readers. These analyses can be jointly informative because they explain different relations among specification, cultural creation, and aesthetic apprehension.

Truth in Fiction and Narrative Inference

Explicit and implicit narrative content

A fictional assertion can be explicit in the text or inferred from what the text presupposes. If a story describes a detective working in a nineteenth-century city, readers ordinarily infer that the detective inhabits a world with streets, buildings, weather, and social institutions of the relevant kind. Yet inferential enrichment has limits. Textual silence does not entitle readers to assert every proposition true of some possible completion of the story.

The problem is best described as a problem of regulated inference. We need criteria determining when an unstated claim belongs to the fictional content. The criteria may appeal to ordinary background knowledge, authorial stipulation, genre conventions, the narrative perspective, and special rules introduced by the work itself. A historical novel and a fantasy narrative may license very different defaults. A plot twist can explicitly defeat an inference that initially seemed unavoidable.

Lewis's possible-worlds analyses reveal the importance of selecting a background while Walton's account treats fictional truth as what an authorized game of imagining prescribes (Lewis, 1978; Fiction, §2). Neither approach should be reduced to the simple claim that every sentence describing a story is true in another universe. Inference requires an account of representational authority.

Inside and outside a fiction

“Holmes is a detective” may express a proposition licensed inside Doyle's fiction. “Holmes was created by Conan Doyle” expresses a historically oriented claim outside that fiction. If an abstract-artifact realist accepts the latter literally, the same noun phrase refers to an abstract cultural object rather than a bodily detective. Treating these sentences as if they used identical predication conditions invites category mistakes.

This distinction also affects apparent paradoxes. “Holmes does not exist” can be true when it denies the existence of an actual historical detective corresponding to the stories. “Holmes is a fictional character” can be true in critical discourse. The two statements are consistent once their predicates and reference conditions are specified. The conceptual work occurs in explaining the distinction, not simply asserting that there are two Holmeses.

Contradictory, unreliable, and indeterminate fictions

Narrators may lie, misremember, or misunderstand events. A story can therefore contain assertions that are narrated without being true according to the work. Character speech likewise does not automatically establish the actual fictional state of affairs. A competent interpretation differentiates the narrator's report, the character's belief, and the work's authorized content. In especially radical fiction, the work may deliberately suspend a stable resolution.

The logic of inconsistent fiction is also distinct from the logic of false scientific hypotheses. A novel can stage apparently contradictory events for aesthetic purposes; a scientific theory containing contradictions normally faces a different norm of repair. This does not make logic irrelevant to literary understanding. It requires a description of which contradictions the work licenses, which inferences readers should draw, and why uncontrolled explosion from one fictional contradiction would destroy rather than improve interpretation.

Fictional Discourse, Pretense, and Make-Believe

John Searle's “The Logical Status of Fictional Discourse” (1975) examines how an author can use ordinary meaningful sentences within fiction without making the ordinary factual assertions associated with them. The distinction turns on the speech acts performed in telling a story. Fictional utterances can take the form of assertions while operating under conventions that suspend their ordinary assertoric commitment (Searle, 1975, New Literary History 6(2), 319–332; 1979 collected chapter).

This approach explains why an author ordinarily does not deceive readers by describing an invented detective. The words preserve their linguistic meanings, while the communicative act belongs to storytelling. Its difficulty lies in cases that move beyond the author's original act: readers compare characters, critics debate them, and cultural institutions refer to them as objects of study. A theory of fictive utterance must either analyze this later discourse separately or show how the original convention extends into it.

Kendall Walton's Mimesis as Make-Believe (1990) makes prescribed imagination the basis of a broad theory of representation. A work can function as a prop in a game of make-believe. Its elements establish what participants are to imagine and what counts as fictional truth within the relevant practice (Walton, 1990, Harvard University Press). The account accommodates literary and pictorial representation and places aesthetic engagement within organized social activities.

The make-believe model explains a great deal without positing a concrete bearer for every character. It also sharpens the question of normativity. If a work prescribes some imaginings, who determines the rules and their extensions? The author, the text, a culturally established practice, or an interpretive community may each contribute. When readers disagree, the theory requires criteria for adjudicating what the authorized game actually prescribes.

Pretense theories and artifactualist realism need not be exclusive in every respect. The fictional practice may constitute an abstract cultural character while also providing the rules for imagining that character as an actual person within the story. A complete theory can accept both claims only if it differentiates the act of imaginative participation from the ontological status of the artifact it sustains. Without that distinction, the phrase “the character exists in the fiction” becomes a source of avoidable ambiguity.

Identity, Authorship, and Historical Dependence

When are two characters the same?

Identity is a demanding test for every realist theory. Two stories can describe individuals with identical properties while introducing different characters. Conversely, a single established character can survive revisions that change many described properties. A purely qualitative criterion therefore looks inadequate. If character identity depends entirely on the bundle of predicates attributed in a story, revisions would too easily create new individuals, while independent authors might accidentally create a numerically identical character.

A historically oriented solution connects identity to practices of introduction and continuation. A sequel, adaptation, translation, or authorized reinterpretation may preserve continuity with the first character even while changing the representation. The historical chain is important, but not sufficient by itself: mere reuse of a famous name might create parody, an allusion, or an unrelated character. An adequate condition combines provenance, representational intention, conventions of continuity, and public uptake.

This resembles problems about the identity of musical works, legal institutions, and other cultural artifacts. Their historical persistence can be compatible with changes in embodiment, performance, and interpretation. An abstract-artifact theory gains plausibility from this analogy, but the analogy transfers the task of specifying identity conditions rather than removing it. Thomasson's treatment of dependence is especially relevant here (Fiction and Metaphysics, chapter contents).

The difference between fictional and historical figures

A historical novel may include actual persons alongside invented characters. An author can represent Napoleon differently from the historical Napoleon without thereby creating an independent Napoleon who supersedes the historical individual. Literary criticism must distinguish the historical referent, the fictional portrayal, and any autonomous character introduced under a related name. The distinction may be straightforward in one novel and deliberately unstable in another.

A historical individual may also become the object of myths, inaccurate biographical traditions, and imaginative reconstructions. These representational developments do not erase the individual's historical existence. The correct ontology should respect the evidence governing historical claims while recognizing a distinct object of literary analysis where a culturally established fictional portrayal is being discussed.

Authorship and collective continuation

Some fictional characters have a single identifiable creator; others emerge through collective storytelling, long oral traditions, collaboration, improvisation, or accumulated adaptations. A theory that treats all characters as instantaneous products of one individual act is too narrow. The constituting activity may be distributed, and its boundaries may be established retrospectively through circulation and recognition.

This does not imply that every fleeting imagined description is already a publicly individuated character. Public reidentification demands enough continuity for speakers to converge on which character they mean. The standards vary across literary genres and cultural practices, but the underlying ontological question is stable: what explains the continued reference to one cultural identity through changing representations?

Fiction, Emotion, and the Human Sciences

Readers respond emotionally to fiction while usually knowing that its characters and events are invented. Fear, compassion, admiration, and grief can arise in response to representations whose depicted objects have no concrete actual counterpart. This phenomenon is sometimes called the paradox of fiction. A strong account must explain the intentional content of the emotion and its relation to belief, imagination, and the conventions of aesthetic engagement.

One response revises an alleged premise that emotion always requires belief in the represented object's actual existence. Another emphasizes imaginative participation and the evaluative structure of emotion. The disagreement is not settled by observing that fictional feelings are intense. Intensity establishes a feature of the human response; it does not by itself determine whether the fictional target is an entity, an intentional content, or a structure of make-believe.

Psychology and cognitive science can investigate how narratives shape attention, simulation, memory, and emotional appraisal. These inquiries may supply causal explanations of engagement. The philosophical question concerns what such engagements are about and what, if anything, their successful explanation implies ontologically. A neuroscientific correlate of imagining a character would establish a property of the reader's activity, not the existence of a flesh-and-blood person described by the narrative.

Aesthetic interpretation introduces further conditions. Literary characters are often deliberately underdetermined, culturally reinterpreted, and assessed through social context. No experiment measuring a reader's emotional response by itself determines the work's authorized fictional content. Empirical research can illuminate mechanisms of interpretation while textual and normative analysis specifies what interpretations are warranted. The two forms of inquiry cooperate when their explananda remain distinct.

Digital Narratives and Artificially Generated Characters

Contemporary digital environments make fictional ontology visible in new forms. A character may persist across interactive games, episodic media, collaborative online narratives, generative systems, and changing databases. What appears to the user as a continuous character may be implemented by many separate computational processes, model states, prompts, files, and interactions. The metaphysical question is whether continuity belongs to an underlying digital object, a representational practice, a culturally recognized identity, or some combination of them.

An interactive character may behave differently for different participants. This does not automatically establish multiple characters. Stage actors have long represented one character in multiple performances; similarly, a digital narrative may preserve one culturally instituted character through varying enactments. Yet computational variation can become constitutive when there is no canonical script or stable framework determining which representations belong to a single character. Identity then depends on the rules of the work and its practices of continuation.

A particularly important distinction separates an artificial system that produces a character from the fictional character represented by its outputs. An implemented generative system is a real technological arrangement with causal history and material or computational infrastructure. The represented figure is identified under the fiction's narrative and cultural conventions. The two may be causally connected without being numerically identical. Whether artificial systems can themselves bear other forms of identity or agency is a separate philosophical investigation.

Theories of abstract artifacts may extend to characters created in distributed or artificial production processes, provided the relevant conditions of creation, attribution, and persistence are specified. The issue does not require assuming in advance that every generated sentence institutes a new cultural object. Creative selection, continuity, dissemination, and recognition remain candidate conditions. The proliferation of automatically produced narratives makes these conditions more urgent and exposes the insufficiency of simply counting every named representation as a distinct entity.

The broader philosophical horizon of Artificial Era raises questions about nonbiological authorship and the organization of public cultural knowledge. In this dossier, the relevant conceptual point is the separation of representational content, producer, artifact, and historical provenance. Those distinctions support investigation across both human and artificial production without replacing the ontology of fiction with claims about machine consciousness or introducing new canonical categories.

Critical Comparison of the Leading Positions

The major theories should be compared according to specific explanatory demands rather than sorted into winners and losers by a single ontological slogan. At least six tests are necessary: intelligibility of empty names; truth of negative existentials; truth in fiction; critical discourse about characters; identity through revision and adaptation; and treatment of impossible or inconsistent content.

Russellian description theory is particularly strong where an apparent name or description can be paraphrased into satisfiability conditions. It exposes illicit existential inferences and negation-scope ambiguities. Its burden increases with ordinary fictional proper names, direct-reference phenomena, and discourse about a character's cultural influence. A successfully eliminated reference must preserve more than the truth value of a single isolated sentence.

Fregean sense theory explains how meaning can remain available without ordinary reference and why ways of presenting an object matter to thought. Its difficulties concern the determination of sense, the truth value of sentences with empty names, and the ontological treatment of literary-critical statements. A purely semantic account of understanding need not provide a complete metaphysics of cultural artifacts.

Meinongian and neo-Meinongian systems place intentional and nonexistent objects in the metaphysical theory itself. They offer a direct semantics for objects of thought and can represent fictional properties systematically. The strongest objection is not that their theories are automatically formally inconsistent. Sophisticated contemporary systems can avoid the obvious paradoxes. The objection is that their broadened domains and specialized relations of predication require independent interpretation and metaphysical justification. Their formal possibility leaves open their explanatory necessity.

Possible-world semantics is highly effective for counterfactual interpretation and the regulated evaluation of narrative statements. It faces the incompleteness of stories, impossible fiction, and the difference between a character and a merely possible person. The introduction of impossible worlds or selection functions may improve the model, but it should be accompanied by a philosophical account of what the formal structures represent.

Pretense and make-believe approaches identify the normative activity through which representations are understood. They explain why fiction can be emotionally and intellectually consequential without ordinary assertoric commitment. Their challenge lies in systematic statements outside the game of imagining: critical comparisons, historical influence, and public reidentification. Such practices may be modeled by paraphrase or by an additional cultural ontology.

Abstract-artifact realism provides an economical literal treatment of many external claims about fictional characters. Its central cost is metaphysical specification. It must explain how an abstract entity begins to exist, depends historically on acts and works, survives changes, and is distinguished from another independently introduced character with a similar profile. The existence of other cultural artifacts makes the proposal plausible but does not entail its truth.

These comparisons reveal that the apparent conflict between semantics and ontology is partly an allocation of explanatory responsibilities. A theory that explains a fictional sentence's interpretation has not necessarily shown whether there is a fictional entity. A theory that posits a fictional entity has not necessarily explained how a reader determines what the story authorizes. The most defensible research program evaluates both directions.

Outstanding Philosophical Questions

Criteria of character constitution

What turns a private imagining into a shareable fictional character? Textual inscription, authorial intention, public recognition, and interpretive continuity are plausible factors. None is universally necessary across oral, collaborative, and digital forms in exactly the same way. A general criterion must distinguish momentary conceptual specification from a historically identifiable cultural creation without excluding nonliterary traditions.

Individuation and cross-work identity

When does a sequel preserve a character, and when does a similar representation establish a new one? Legal ownership, naming, canonical status, continuity of description, and explicit acknowledgement can come apart. A theory of identity must be sensitive to these variations while retaining principled grounds for numerical sameness.

The ontology of impossible content

The conceptual utility of impossible descriptions is clear; their metaphysical import is disputed. A formal system may allow precise reasoning about inconsistent objects while treating those objects as representational constructions. The outstanding question is which additional explanatory achievements would justify treating impossible objects as ontological inhabitants rather than mathematical or linguistic devices.

The semantics of external fictional assertions

Statements such as “Hamlet is a fictional character” or “Holmes has been adapted many times” appear straightforwardly true. It remains contested whether they require fictional entities as referents or can be paraphrased through works and practices. Competing theories should be tested against full networks of actual literary-critical inferences, not selected easy examples.

Dependence, persistence, and public recognition

If characters are abstract artifacts, what happens when the last copy of a text disappears, its creator is forgotten, or no community continues its narrative? Historical creation may be sufficient for persistence on one account, while continued cultural maintenance may matter on another. The choice has implications for the metaphysics of all dependent cultural entities.

The place of fictionalism outside literature

Mathematical, scientific, and modal fictionalist approaches use the vocabulary of fiction for several different explanatory purposes. They cannot be assumed to share a single claim about whether mathematical objects, idealizations, or possible worlds exist. Each fictionalism requires specification of the propositions accepted literally, the representational devices employed, and the practices in which those devices earn their role (Fictionalism in the Philosophy of Mathematics; Modal Fictionalism).

Philosophical Synthesis

The completed analysis supports a layered account of how nonexistent and fictional discourse functions. These layers are analytical distinctions that reconstruct the findings of the research; they are not proposed as a new canonical ontology or a comprehensive rival theory.

First, representational determinacy permits speakers to think and reason about descriptions even when they lack actual satisfiers. The content of a description may be stable enough for assessment without identifying an object that exemplifies it. Frege's distinction between sense and reference, Russell's contextual analysis of descriptions, and the resources of free logic each explain different aspects of this achievement.

Second, the logical form of a sentence determines whether its surface naming expressions commit its theory to existential quantification. Quine's methodological insight survives in the requirement to state an exact regimented theory. Yet an adequate metaphysical verdict also depends on the adequacy of that regimentation. The choice between quantifying over characters and paraphrasing such quantification cannot be made independently of the explanatory practices at issue.

Third, narrative content is evaluated relative to a work or representational practice. The predicates of the story do not transfer without qualification to the actual world. A fictional detective is a detective in the story; that does not make an abstract cultural character an actual detective. Possible-world semantics and make-believe explain why this qualification is principled and why fictional inference is nevertheless answerable to rules.

Fourth, external discourse about characters is historically situated. Authors introduce, audiences recognize, publishers disseminate, and interpreters continue narrative identities. These are actual events and practices. An abstract-artifact realist interprets a character as a culturally constituted entity dependent on that history. An eliminativist interprets the same history through existing works and agents. The strongest version of the debate concerns which ontology provides the better explanation of reference, persistence, and systematic critical judgments.

Fifth, the type of representational failure matters. An undiscovered planet, a fictional detective, an impossible geometric figure, and a merely possible person occupy distinct positions in inquiry. Their differences concern the purpose of introduction, conditions of satisfaction, treatment of inconsistent predicates, and relation to historical cultural practices. A universal category called “the nonexistent” would obscure these distinctions unless furnished with an equally precise internal structure.

The synthesis yields an operative method for evaluating claims about a purported entity. Identify the relevant discourse and its purpose; establish whether the claim is internal to a representation or external criticism; specify whether its names denote, describe, or function within pretense; reconstruct the logical commitment of the best interpretation; and test whether the resulting ontology explains identity and dependence across contexts. At every point, truth conditions and metaphysical commitment are related through argument rather than inferred from grammar alone.

This method also explains why ontology and philosophy of language need each other. The ontology cannot decide what must exist without interpreting its sentences. The semantics cannot completely explain socially persistent fictional identities without considering the structures and histories on which their identification depends. Aesthetics contributes the norms of representation and interpretation; logic provides formal discipline; ontology asks which commitments the integrated account ultimately warrants.

Conclusion

Nonexistent and fictional entities are a fundamental philosophical subject because meaningful thought and organized discourse extend beyond objects presently instantiated in the actual world. The extension is undeniable as a feature of representation. Its metaphysical explanation is contested. The historical trajectory from Plato to contemporary abstract-artifact realism reveals increasingly exact ways of separating negation, sense, reference, description, quantification, intentionality, fictionality, and cultural identity.

Meinong makes the breadth of intentional objects the central metaphysical problem. Russell demonstrates how apparent singular reference can be analyzed away. Frege explains meaningful presentation without successful reference. Quine demands explicit accounts of quantificational commitment. Parsons, Zalta, and Priest show that sophisticated theories of objects and predication can formalize alternatives to classical assumptions. Lewis and Walton analyze the inferential and imaginative practices through which fictional propositions become assessable. Ingarden, van Inwagen, Kripke, and Thomasson illuminate the dependence, existence, and identity of literary characters as objects of cultural discourse.

The philosophical result is a disciplined separation of representation from instantiation and of a represented person from the historical reality of a work. From that basis, realism and eliminativism about fictional characters become substantive rival interpretations, each accountable to the same semantic, logical, historical, and aesthetic evidence. A completed theory must explain what a term contributes to meaning, why some sentences containing it are true, how its referent or apparent referent is individuated, and what grounds the persistence of discourse about it.

This research establishes the conceptual architecture within which these questions can be answered. It leaves open the ultimate choice between competing ontologies where the evidence of successful interpretation permits more than one metaphysical regimentation. That openness is itself a determinate result: the existence of fictional characters is neither established by the intelligibility of their names nor refuted by the nonexistence of the people those names portray. It is decided, if at all, by the comparative explanatory success of fully articulated philosophical theories.