RESEARCH DOSSIER
The Ontology of Art
Abstract. This fundamental research monograph investigates the ontological status of works of art by asking what kind of entity an artwork is, how it comes into existence, and what makes it persist through change, copying, performance, reconstruction, or digital migration. It reconstructs the historical development of the problem from Plato and Aristotle to Kant, Hegel, Collingwood, Ingarden, Wollheim, Goodman, Danto, Levinson, Currie, David Davies, Thomasson, and Rohrbaugh. The central analysis distinguishes art status, work identity, material constitution, realization, historical provenance, interpretation, and artistic value. It examines singular paintings and sculptures, architecture, literary editions, musical compositions and improvisations, theater and dance, photography and film, conceptual art, installations, interactive software, and processual digital works. The original philosophical synthesis proposes historically instituted identity: each artwork is a historically situated locus of artistic identity governed by constraints determining its constituents, permissible realizations, modifications, and continuity. This framework accounts for why visually identical objects can be distinct works and why one work can persist through materially different authorized instances. It treats restoration and provenance as tests of identity rather than merely matters of storage or documentation. A final application relates artwork individuation to Aisentica's canonical account of Artificial Provenance and its additional Theory of Artificial Art, while preserving the independent identity of classical and analytical aesthetics. The study establishes a pluralistic, historically accountable ontology in which medium-specific modes of being are intelligible within a shared logic of artistic institution, identity, and persistence.
Keywords: ontology of art; artwork identity; artistic ontology; artifact; type and token; constitution; artistic provenance; historical individuality; abstract artifact; autographic and allographic; work and performance; restoration; digital art; artificial art; Aisentica; Artificial Era.
The Ontological Problem and Its Explanatory Burden
A work of art enters ordinary discourse as something whose existence appears obvious. A painting hangs in a room, a novel circulates through editions, a symphony is performed, and a performance ends. Yet these descriptions presuppose strikingly different relations between the work and what can be encountered. The painted surface can be physically located; the novel may survive the destruction of any particular copy; a symphony can be heard through many interpretations; an improvised performance may have occurred exactly once while leaving a recording, description, or memory. Ontology asks what there is in each case and what conditions make the bearer of artistic identity the very work that criticism, history, and conservation discuss.
The guiding question of this monograph is accordingly: What sort of entity is a work of art? Answering requires more than sorting art into ready-made classes of objects. The categories themselves are under examination. If a work can be created, repeatably performed, modified, destroyed, misidentified, restored, or translated, an account must explain how each possibility arises from the work's nature. What persists when a score changes hands? What is lost when a painting is repainted? Which decisions distinguish a new interpretation from a different work? What identifies an interactive artwork through different runs of its software? These are tests of an ontology's explanatory power.
A second difficulty is that the phrase "work of art" identifies the things for which artistic practices claim continuity, yet the basis of continuity may differ radically by practice. A bronze sculpture and a conceptual instruction piece can each be unitary works without sharing a single material constitution. A recorded sound work may be tied to a particular fixation, while a musical composition admits indefinitely many authorized realizations. A theory that treats one familiar medium as paradigmatic can turn a contingent practice into a universal metaphysics.
The thesis defended here is a constrained pluralism organized around a single question of historical identity. An artwork is an artistically instituted, historically situated locus of identity whose constitutive constraints determine which materials, events, structures, or realizations count as that work. Works may be physical individuals, norm-governed repeatable structures, realized events, or organized processes. Their differences are genuine ontological differences. The unifying account concerns the way a work is instituted and maintained as the same object of artistic reference, assessment, and interpretation. That unity does not prejudge which entities can become art or whether every candidate has sufficient aesthetic value.
This framework distinguishes five questions often confused: whether a thing is art; what kind of thing it is; which occurrences instantiate or constitute it; when it remains the same work; and why it matters. The first concerns art status, the second category, the third realization, the fourth persistence, and the fifth evaluation. Any adequate ontology must state the relations among them without substituting one for another. This sequence will guide the historical reconstruction, the medium-specific analyses, and the constructive account.
Method, Levels of Evidence, and the Problem of Attribution
Philosophical research into artworks operates with several forms of evidence. Original texts provide arguments, distinctions, and terminological choices. Artistic practices provide cases in which distinctions must function. Critical scholarship offers reconstructions and objections. The present synthesis derives conditional criteria from their combined explanatory demands. These evidential levels remain separately attributable. An ancient distinction concerning imitation, for example, is a historical antecedent; the contemporary claim that a digital work is an executable identity is a new application requiring its own argument.
The research employs canonical internal references where they are meaningful. Aristotle's Poetics is discussed by Bekker numbers, Kant's Critique of Judgment by sections, and Danto's 1964 essay by its original journal location. English-language formulations are paraphrased unless an edition has been directly inspected and the quotation is essential. This reduces the risk of mistaking the language of a contemporary commentator for that of an original author. Works known only through a publisher's description, bibliographic record, or a reliable specialist synthesis are used at that declared level and are not represented as if their entire text had undergone collation.
A useful methodological test asks what a theory predicts when artistic and physical changes come apart. Consider the destruction of a canvas and the survival of high-resolution documentation. The physical object is gone, but its image persists. Whether the work survives depends on which properties were constitutive of it. Consider a notation performed incorrectly: a causal event occurs, yet its membership among valid instances remains disputed. Consider a sculpture reconstructed from cast fragments: preservation of an outline may coexist with loss of material continuity. Ontology becomes substantive precisely where identity has consequences for authenticity, interpretation, and restoration.
Conceptual engineering must nevertheless answer to established language. When critics say that a symphony was written in one year and performed centuries later, an account must preserve the ordinary distinction between the composition and its performances or explain why that distinction should be revised. When a museum says that it owns a painting, ownership of a material artifact usually matters; when a library says that it holds a novel, it normally holds an exemplar rather than an exclusive possession of the literary work. Ordinary claims are fallible, yet their stable inferential patterns are significant evidence.
The study follows historical sources in the order in which ontological problems become explicit: classical treatments of images and poetic wholes, modern theories of art and artistic production, phenomenological accounts of the work as distinct from its reception, analytical theories of tokens, structures, intentions and institutions, and contemporary problems raised by restoration, digital media, and artificial provenance. The resulting synthesis should be tested by counterexamples rather than secured through a stipulated vocabulary.
Being, Work, Object, and Artifact
Ontology traditionally concerns existence, identity, dependence, constitution, and persistence. Applied to art, these questions cannot be resolved merely by observing that artworks are made. The ordinary term "artifact" marks something resulting from intentional production, but artistic identity can exceed the material produced. A carved stone may itself be the work; a manuscript may be an artifact that bears a repeatable literary work; an installation may comprise replaceable objects governed by instructions; a dance may consist in a form of organized activity whose occurrences cannot be stored as a stone can.
The distinction between work and vehicle is therefore essential. A vehicle is that through which an artwork can become perceptually, imaginatively, or interpretively accessible. The relation is sometimes identity, as when the particular painted canvas is treated as the work itself. Elsewhere it is instantiation, as when one authorized performance realizes a musical composition. Elsewhere it is an evidential relation, as when a photograph documents a performance whose occurrence has passed. Vehicles matter deeply; to distinguish them from works is not to diminish their role but to assign their role accurately.
Constitution adds another layer. A statue can be constituted at a time by a lump of bronze while possessing historical and artistic properties that the bronze, described merely as a quantity of metal, does not possess. The work may have a name, an authorship, an intended orientation, and a history of exhibition. The bronze may exist before casting and remain as metal after the statue's destruction. If identity conditions differ, an adequate theory must accommodate the difference while explaining how a single material situation can support these predications.
Artifacts depend upon practices of making, use, and interpretation. Yet historical dependence need not imply permanent dependence on the creator's mental activity. A completed novel generally survives the author's death; a painting remains intelligible when the original artist's intentions are disputed; an anonymous ancient mask can acquire new artistic significance. Dependence is therefore relational and layered: origin depends on an initiating process, continued material existence on carriers or performances, social intelligibility on practices of recognition, and interpretive intelligibility on conventions that can change.
One must distinguish numerical identity from qualitative similarity. Two perfect reproductions may be visually indistinguishable and nonetheless be two material objects. Conversely, a single artwork may undergo significant qualitative change while persisting historically as the same work. Numerical identity asks whether one entity continues; similarity asks how closely properties match. Artistic understanding constantly requires both questions, especially in conservation, edition-making, recording, and replication.
Plato: Image, Model, and Ontological Distance
Plato supplies an important ancient point of departure even though the modern ontology of artworks as a unified technical subdiscipline develops much later. In Republic X, the discussion of beds and painted appearances distinguishes an intelligible form, the craftsperson's product, and the painter's image. The argument assigns different explanatory roles to model, usable manufactured object, and representation. It therefore offers an early metaphysical problem concerning what a made image is and how it relates to what it presents (Plato, Republic, 596a–598b).
The relevant issue is not exhausted by the familiar charge that mimetic art is distant from knowledge. A painted bed is not a bed intended for sleeping. Its representational function is constituted through an image relation. Even when its depicted shape corresponds to a physical bed, it is not numerically identical to its subject. The artwork introduces a distinct entity with its own manner of appearing and being judged. Plato's criticism is intelligible only because an image can operate as an image, inviting an orientation to appearances rather than practical use.
The painter's image also exposes a limitation in straightforward material reduction. A painted bed may be materially a set of pigments on a support, but its artistic and epistemic role cannot be described by pigment chemistry alone. The discrepancy is visible even within a philosophy critical of image-making. Representation is a relational property: it points beyond the material vehicle to a represented object under a mode of presentation. To establish the image's ontology, one must explain both vehicle and reference.
The Ion places poetic performance in a different argumentative setting, asking about the rhapsode's competence and the relation between performance, interpretation, and inspiration. This question anticipates a distinction between reciting a poem and possessing knowledge of everything represented in it. The performer's event and the poem's identity are therefore not trivially identical, although Plato does not develop a contemporary theory of literary types. Both dialogues raise problems that become productive when later thinkers disentangle aesthetic, epistemic, and ontological dimensions.
Plato's use of mimēsis must be placed within its own context. Translating it indiscriminately as modern photographic copying erases its associations with representation, dramatic enactment, and imitation. The present monograph takes from Plato a structural problem: an image or performance is a real occurrence with an intentional relation to what it shows. It does not infer that every later artistic work must be mimetic. Abstract painting, instrumental music, and conceptual practice establish the limits of such a universalization.
Aristotle: Poetic Wholes and the Identity of Action
Aristotle transforms the problem of artistic representation by analyzing poiēsis and mimēsis through forms of organized action rather than only through the distance between image and model. In Poetics 6, tragedy is characterized by an imitation of an action, and its parts include plot, character, thought, diction, song, and spectacle (1449b24–1450a15). Aristotle accords priority to the arrangement of incidents. What makes the tragedy a structured whole is not an inventory of the physical bodies and costumes visible in a particular staging.
This has an ontological consequence. A dramatic work may be repeatedly enacted by different performers while preserving relations among plot, action, and articulated structure. One can ask whether a given performance belongs to the tragedy and how it renders its constitutive organization. Aristotle does not provide modern token theory, but the distinction between the organized action and the visible spectacle helps explain why the work's unity is not identical with the unity of any single event.
In Poetics 7–8, Aristotle's account of a whole composed of beginning, middle, and end and his discussion of plot unity place emphasis on necessary or probable connection. The structural requirement is stronger than temporal succession. A story with a protagonist does not thereby achieve unity merely because all its episodes concern one person. This argument suggests that the poetic work's internal organization supplies its criterion of intelligibility. The parts count as parts of a single artistic whole because their relations produce a determinate action.
A similar issue appears in Metaphysics H.6, where Aristotle asks how what has parts can be one rather than a mere heap (1045a). The account of matter and form is not itself an explicit universal definition of artworks. Its relevance lies in the metaphysical distinction between composite unity and accidental aggregation. An artwork assembled from many components may be one because its organization constitutes their relation, while the same material elements scattered or rearranged may fail to constitute that work.
The Poetics also clarifies the boundary between work ontology and evaluation. An account of the unity of a plot helps identify the kind of object being evaluated; its artistic success remains another question. A badly executed tragedy can still be a tragedy. This distinction becomes indispensable when an ontology must identify a work regardless of whether it succeeds aesthetically. Aristotle thus contributes a model of organized action and whole-part relation rather than an exhaustive system of modern artistic entities.
The Historical Formation of the Modern Artwork
The familiar modern conception of the work as an autonomous item of fine art has a history. Ancient and medieval practices combined artifacts, images, ritual, skilled production, entertainment, religious service, and civic representation under classifications that do not map neatly onto the modern grouping of painting, sculpture, music, poetry, and dance. The modern system of the fine arts arose through institutional, educational, critical, and economic developments. Any ontology that assumes this grouping is timeless inherits a historical anachronism.
The distinction matters because a thing can have an identity within a practice without having been originally introduced as a member of a modern artworld. An illuminated manuscript can be a liturgical book, a crafted physical object, a textual witness, and a later object of artistic display. A carved mask may be integrated into a ceremonial practice, yet acquired and exhibited in a gallery through a radically altered social relation. Ontology must account for multiple relations to one enduring material object without assuming that the gallery creates its entire prior existence.
This history is also a history of reproduction. Printing made it possible to distinguish a textual work from many copies in increasingly sophisticated ways. Musical notation and performance conventions likewise made distinctions between the compositional work and particular realizations institutionally central. Museums, copyright, catalogs, signatures, critical editions, and attribution practices reinforced the expectation that a named artwork is individually identifiable over time. Such institutional developments did not invent every underlying identity distinction, but they made certain distinctions authoritative and widely consequential.
A philosophical question therefore emerges from historical variation: is the ontology of art determined by the vocabulary available in one cultural period, or do artistic practices reveal more general relations of constitution and repeatability? The strongest answer is neither to impose one current art category backward nor to abandon comparison entirely. We can identify comparable operations—making, organizing, repeating, preserving, narrating, and interpreting—while acknowledging that their practical significance and classification vary historically.
The resulting research discipline is historically reflexive. Its examples carry the marks of their emergence. The originality of a painting, the fidelity of a performance, and the authenticity of a restoration are partly guided by practices whose rules developed. To treat such rules as philosophically substantive is not to collapse ontology into whatever an institution declares. It is to understand how historically situated forms of making can institute stable objects with differentiated identity conditions.
Kant: Production, Freedom, and the Work of Fine Art
Kant distinguishes art from nature through production guided by a representation of a purpose. In Critique of Judgment §43, art is considered under the relation between a making and the rationally guided activity that produces it. The contrast is important for the ontology of artifacts: natural formation and artistic making may yield similar perceptible arrangements while belonging to different causal and practical histories. The distinction supports a historical rather than purely morphologically defined account of artistic constitution.
In §§44–50, Kant develops the specificity of fine art, artistic genius, aesthetic ideas, and the relation between understanding and imagination. Fine art is presented as production capable of being appreciated in a mode that exceeds the mere following of determinate technical rules. Kant's emphasis on genius concerns the way exemplarity and creativity enter artistic production. It does not state that a painting is a nonphysical abstract object or that every artwork is necessarily original in the modern legal sense.
His position yields a difficulty for a strictly instruction-based ontology. If the entire identity of a work were exhausted by an explicit recipe, what room would remain for a practice in which the exemplary product gives the rule to art? A composition may nevertheless incorporate explicit constraints; a novel may be fixed by a sequence of words. The lesson is that the circumstances of creative formation and the criteria governing later correct realization can differ. A work can originate through inventive activity and later be reproduced according to relatively stable constraints.
Kant's account also alerts us to the danger of identifying artistic existence with immediate pleasure. Judgments of taste concern reflective appreciation, but the artwork's identity cannot depend on the occurrence of one spectator's pleasure. A work can survive indifference and misunderstanding. The act of judging it as beautiful is distinct from its existence as an object of possible judgment. Art status, aesthetic value, and ontological kind must therefore remain analytically separate.
At the same time, Kant's production-oriented conception is historically bounded in a way that later technologies expose. Machine-generated, collaborative, and procedurally executed artworks complicate the idea of a singular human genius guiding every local operation. The appropriate historical reading recognizes the original role of freedom and rational representation in Kant's account while leaving room for subsequent extensions of artistic provenance. Whether such extensions succeed must be argued through the criteria of artistic institution and intelligibility rather than merely announced.
Hegel: Art as an Historically Mediated Appearance
In Hegel's lectures on fine art, artistic works belong to a historical exposition of spirit and its sensuous presentation. Art is approached through relations among content, form, material appearance, religious life, and historical configurations. The work is meaningful in virtue of its position within a development of forms of self-understanding. Thus a Greek sculpture and a modern romantic artwork cannot be completely understood by listing their shared physical properties.
The source itself requires editorial care. Hegel's Lectures on Fine Art was assembled posthumously from lecture materials and notes, and modern editions mediate this textual history. Assertions about a single fixed sentence of "Hegel's aesthetics" require identification of the editorial or translation witness. The present analysis relies on the well-established framework of the lectures and the bibliographic identity of the T. M. Knox English translation, while treating edition-specific formulations as requiring further full collation.
The central ontological contribution is historical mediation. An artwork embodies a relation between a sensuous configuration and the meaning articulated through that configuration in a historical world. The work is more than an accidental carrier of symbolic content. Its material form is significant to how that content can become manifest. A theory that subtracts the work's historical horizon and keeps only its measurable arrangement may retain data while losing what is being presented as art.
Hegel also introduces a potential difficulty. If artistic significance depends on a particular developmental system, can the work remain identifiable outside that system, and can other traditions receive adequate treatment on their own terms? A contemporary ontology should preserve the insight that historical placement matters without endorsing a single teleological ranking of world artistic traditions. The historical horizon can be constitutive of identity without fixing one universal destination for the arts.
For the present argument, Hegel supplies a bridge between making and recognition. Material appearances are situated within forms of life that establish what they can mean and why their presentation matters. That historical embeddedness becomes particularly important for readymades, archives, and art of computational provenance. It is compatible with treating some works as physical individuals, some as reproducible structures, and some as performances. The history of art changes the possibilities of artistic institution; it need not force all instituted works into one metaphysical category.
Collingwood and Artistic Activity
R. G. Collingwood's The Principles of Art (1938) approaches art through the clarification of expression and the distinction between art, craft, amusement, and magic. The emphasis falls on the activity through which experience is made articulate. In this tradition, equating the artwork with a finished external commodity risks mistaking the product or instrument for the artistic achievement. The central question becomes what kind of activity must occur for expression to be present.
This approach exposes the limitations of a purely inventory-based ontology. A performance may have artistic significance because a form is realized in action; an act of expression can occur without yielding a durable commodity; a painting's production may be constitutive of its expressive character. The thesis does not require that material vehicles be irrelevant. It requires us to examine whether the activity of creation is part of the work's identity or merely a causal antecedent.
The strongest version of activity-centered ontology meets a persistence challenge. After the act is over, critics continue to discuss an enduring work, and museums preserve artifacts that can be revisited by those who were not present at their creation. If the work is exclusively an activity, what remains available for later appreciation? Possible answers include traces, conventions, public memory, or re-enactment. Each requires a theory of the relation between an originating event and the continuing object of reference.
A second problem arises from private expression. Artistic identities are publicly attributable and corrigible; disagreements about versions, titles, attribution, or authenticity can be settled by evidence that outlasts the originating consciousness. If one makes the completed work entirely internal to the artist's experience, it becomes difficult to explain this public stability. Collingwood's emphasis is best understood as drawing attention to expressive processes that object-centered theories may omit, rather than as eliminating the need to distinguish an artwork from every experience that accompanies it.
The resulting constructive lesson is that ontology must admit process-dependent works and publicly accessible continuants. An artwork's originating activity can be essential to it without being identical to every later instance in which it is interpreted. The relation between artistic agency, action, and result is one of the clearest reasons to adopt a stratified rather than a monolithic ontology of artistic existence.
Phenomenology, Intentionality, and the Work
Phenomenological traditions place the appearing of the artwork in relation to intentional experience without treating all works as reducible to subjective impressions. The central question is how an object that can be reencountered by multiple readers, listeners, or viewers is constituted and disclosed through acts of attention. In the analysis of a literary work, this difficulty is sharpened: written signs are physical, meanings are understood, imagined scenes are presented, and none of these layers can simply replace the others.
The concept of intentionality concerns an act's directedness toward an object. It does not mean that every intentional object is an illusion or that artistic objects exist only as current mental events. A printed sentence can be encountered under different interpretive horizons while remaining part of the same text; its directed meaning requires more than its ink configuration. This distinction helps explain why a description of paper and ink is an incomplete description of a literary work.
Roman Ingarden developed a detailed ontology of literary artworks partly in the context of disputes over the ontological status of intentional objects. His The Literary Work of Art first appeared in German in 1931; the translation by George G. Grabowicz was published by Northwestern University Press in 1973. His account distinguishes a work from the author's psychic acts, the reader's experiences, and particular physical copies. The distinction is decisive for any theory requiring a repeatably identifiable literary entity.
Ingarden's ontology is especially important because he does not solve the problem by assuming that the literary work is a timeless Platonic universal. Literary works arise historically through the activities of authors, but their ideal and intentional dimensions require analysis beyond the causal history of the book. He proposes multiple interdependent strata, including language sounds, meanings, schematized aspects, and represented objects. Their complex arrangement is essential to what is available for understanding.
Phenomenology thus contributes a method of attending to constitution and disclosure. The work is encountered through experiences, yet its identity can be examined across experiences; its contents depend on language and representation, yet its material exemplars are insufficient to define it; its significance unfolds historically, yet later readings remain readings of an identifiable creation. The distinction between work and concretization becomes central in the next chapter.
Ingarden's Stratified Literary Work
Ingarden's analysis identifies four interconnected strata in the literary work: word sounds and higher-order linguistic formations, units of meaning, schematized aspects, and represented objects or states of affairs. These distinctions are not a list of four independent artifacts. They analyze how one literary work articulates linguistic material, semantic significance, imaginative presentation, and a represented world. Their interactions can contribute different dimensions of aesthetic value (Ingarden, 1973; Thomasson, "Roman Ingarden").
The second and third strata are particularly important for identity. A sentence may determine that a character enters a room, but it need not specify every aspect of that room. The reader concretizes what is schematically presented. Different readers can supply different imaginative determinations without thereby creating numerically different novels. The work constrains possible concretizations while leaving certain determinacies open. This provides an early and sophisticated account of the tension between objective identity and variable realization.
The first stratum exposes the role of language in translation. If the exact phonetic and rhythmic properties of a literary work are essential to it, a translation may transform those properties while retaining large regions of semantic and narrative organization. It is therefore unsafe to claim automatically that a translated text is identical to the original in every respect. A translation can be a dependent artistic object bearing a special relation to another work, and the status of that relation must be understood by reference to the kind of literary work and the criteria of its identity.
Ingarden's distinction also prevents a common category error: a reader's image of a fictional castle is not literally an object in the real landscape. The represented object belongs to the intentional world of the work. Yet its determinacy is not unlimited. The authorial text may rule out some interpretations, support others, and leave a range of possibilities open. This structure demonstrates why a literary ontology must articulate constraints rather than merely assert that whatever a reader imagines is part of the work.
A difficulty remains concerning the historicity of interpretation. Language changes, conventions shift, and reception can modify a work's standing. An account of fixed strata must explain how historical meaning and recontextualization bear on what the work is. The most promising answer distinguishes historically instituted identity conditions from the developing field of interpretations: critical rediscovery can change what is known or valued without retroactively changing every constitutive feature. Where artistic practice deliberately makes later audience participation constitutive, the identity conditions can specify that participation from the outset.
Physical Singularities: Painting and Sculpture
Painting and sculpture seem to support a straightforward ontology of concrete individuals. A given canvas or carved block is numerically singular, located in space, and subject to decay. The distinction between the original and a perfect copy may depend on material and historical continuity even when spectators cannot distinguish the two by sight. A forgery can reproduce appearances without reproducing a causal history. This is evidence that relevant artistic properties include origin and not merely perceptual configuration.
Yet identifying the work with a bare physical object is too quick. The same canvas has physical properties under many descriptions, but its orientation, boundary, intended finish, and artistic history can affect what counts as an alteration to the work. A painted surface trimmed for framing may be experienced differently; a work attributed to a different author acquires a different history of critical interpretation. Such changes need not annihilate numerical identity, but they show that the correct predicates of the artwork are not exhausted by material analysis.
Sculpture supplies additional complications. Foundry casting can produce several authorized bronze casts of the same model. In some practices each cast is a distinct sculpture; in others the casts are exemplars of a multiply realized sculptural work. The criteria depend on authorization, numbering, production practice, and the artistic role of individual material traces. The ontologist must ask what the creator or practice established, not infer one answer from the substance bronze.
A work can also be site-specific. Moving a sculpture from its original location may preserve the object while altering or destroying an essential relation to the surrounding environment. For one work, location is an exhibition circumstance; for another, it is part of the work's constitutive organization. The physical continuant remains numerically the same artifact even if its status as a complete realization becomes contested. Such cases reveal why constitution and correct presentation require separate evaluation.
The strongest physical-individual theory is therefore relational. A singular art object exists through material continuity integrated with a historical identity. It can survive change within limits fixed by the kind of work it is, and it can lose authenticity or essential components without instantly ceasing to be identifiable. A physical substrate is an indispensable condition for some works, but its role must be specified rather than universalized across every art form.
Architecture and the Identity of Place
Architecture complicates the division between artwork, artifact, and environment. A building is designed for occupation, practical use, and structural endurance; it may also embody an artistic organization of space, movement, light, material, and public meaning. To ask whether the building is an artwork is one question. To ask what entity persists when a building is repaired, rebuilt, or reconstructed from drawings is another.
Plans create an important analogy with musical notation but do not settle the matter. A plan can specify repeatable spatial relations, yet a constructed building occupies a particular site and interacts with weather, materials, labor, and collective use. Two buildings constructed from the same drawings are usually distinct physical buildings. They may be two realizations of a design; whether they are two instances of one artwork depends on the relevant practice and artistic intention.
Site specificity adds a constitutive dimension. The relation of a structure to a landscape, street, or public ritual can be part of its artistic identity. If the structure is moved, some of its architectural features remain while the site relation is severed. A photographic record can preserve information about former relations without replacing the spatial experience in which they operated. An ontology of architecture therefore needs concepts of situated realization and contextual loss.
Renovation makes the persistence problem explicit. Replacing a broken beam may preserve a building; progressively replacing its materials raises questions about the relation between historical continuity and composition. A restored façade may match the old appearance while concealing a different load-bearing system. An ontological account should distinguish preservation of an object's ongoing life, reconstruction of a design, and fabrication of a historically informed replica. These operations are related but do not necessarily have the same artistic result.
Architecture demonstrates the general thesis that artwork identity is medium-sensitive and historically instituted. Its constitutive constraints may include material continuance, use, circulation, site, and inherited form. None of these has an identical status in every building. The concrete question is which relations the relevant artwork depends upon and why their preservation or alteration matters to its continued identity.
Wollheim and the Plural Ontology of Artistic Objects
Richard Wollheim's Art and Its Objects, first published in 1968 and expanded in the 1980 second edition, helped establish the philosophical significance of distinguishing kinds of artworks and the relations between them and their instances. Its inquiry concerns what an artwork is and what is appropriate to understanding it. The plurality of artistic objects already suggests that a painted canvas, a symphony, and a poem should not be assigned identical metaphysical conditions simply because all are called artworks.
A central distinction concerns singular works and works that admit multiple occurrences. A painting can normally be regarded as a particular historical individual. A piece of music may be performed in many temporally distinct events; a play can be staged repeatedly; a literary work can appear in many copies. To preserve the distinction is to protect a difference in how works are encountered, repaired, owned, or lost. The question of whether the repeatable work is an abstract type then follows, but it is not answered by repeatability alone.
Wollheim's contribution is also methodological. The study of art objects is guided by the practices through which works are made and recognized. A theory that identifies a work with its physical vehicle may be appropriate in one setting and misleading in another. It is therefore essential to assess what the concept of a work accomplishes within criticism. A symphony's identity allows distinct performances to be compared as performances of one composition; a painting's identity allows the original to be distinguished from a visually precise reproduction.
The chief limitation of simple pluralism is that naming different categories can leave their organizing relation unexamined. A taxonomy may tell us that paintings are particulars and symphonies are types while failing to explain why their criteria of identity count as artistic rather than merely logistical. The constructive account developed later supplements pluralism with a theory of historically instituted constraints. Those constraints explain why some differences between vehicles matter and others do not.
This reading uses the verified 1980 second-edition bibliographic record and specialist ontology literature. It does not assign page-specific propositions to Wollheim without edition collation. The secure contribution for this analysis is the centrality of objects, kinds, and the differentiated relations between works and their physical or performed realizations.
Types, Tokens, and the Paradox of Creation
The type-token distinction is one of the most powerful analytical tools in the ontology of art. A token is an individual occurrence or inscription; a type is a repeatable pattern or kind of which tokens are instances. Different printed copies can exemplify one sequence of words; performances can be grouped under one musical composition. Yet the statement that a work has tokens does not settle what kind of type the work is or whether the type was created by an artist.
The paradox of creation can be expressed as a triad: artworks are created; artworks are abstract objects; abstract objects are uncreated. Taken together, the three claims are inconsistent. Some philosophers reject the first claim for certain repeatable structures, treating the artist as discovering a preexisting possibility. Others reject the second, classifying works as concrete historical individuals or processes. Others reject the third, defending created abstract artifacts or types. The history of the debate is documented in the Stanford Encyclopedia's account of art ontology and in Thomasson's work.
The paradox is instructive because it exposes an ambiguity in "abstract." A timeless mathematical structure and a historically established musical work may both lack spatial location in the ordinary way, yet they need not share the same creation conditions. A composer's choice of title, instrumentation, expressive character, and intended performance practice can form a historical object that is not identical with a merely possible sound sequence. The musical work may incorporate a pattern while having additional relational properties that a pattern considered independently does not have.
The creation argument can be reconstructed in four steps. First, artists introduce entities that can be publicly identified by names and historically specific criteria. Second, some of these entities admit multiple valid instances. Third, a thing may be repeatable without being a timeless universal: rules can be instituted on a date and subsequently govern many events. Fourth, therefore, repeatability alone does not establish eternal existence. The conclusion supports, though it does not prove, the coherence of historically created repeatable works.
An objection claims that this simply replaces metaphysics with vocabulary. Calling a type "created" does not explain what changes in the world. The answer must identify a constitutive event: decisions, actions, inscriptions, performance specifications, and public practices establish standards by which later occurrences are classified. Such an event creates a new historically individuated object of reference, even if the possible configurations existed as abstract possibilities before it. The difference between possibility and historically instituted work is the key explanatory contrast.
Nelson Goodman: Notation and the Autographic–Allographic Distinction
Nelson Goodman's Languages of Art (1968; second edition 1976) analyzes systems of symbols and introduces influential distinctions concerning the identity of artworks. The autographic–allographic distinction concerns whether an exact duplication can count as another genuine instance of the same work or whether origin and production history make the distinction between original and forgery constitutive. A painted canvas is a standard autographic case; correctly notated musical works and literary texts have often been treated as allographic cases.
The distinction is not simply "one object versus many objects." It concerns criteria of compliance and authenticity. In an allographic setting, a notation can establish that a particular inscription or performance complies with the score or text. The historical identity of the individual copy may be irrelevant to whether it is an authentic instance of the work. In an autographic setting, matching an object's visible properties does not suffice: the fact that the mark-making occurred under one history rather than another can determine whether the object is an original.
Goodman's analysis of notation is more exacting than an appeal to written symbols. A notational system must distinguish its characters in syntactically and semantically appropriate ways. The gap between an informal set of performance suggestions and a fully articulated notation matters because performers must know what compliance requires. Musical scores contain conventional signs, but performance practice can supply further conditions that the visual inscription does not settle. Whether a work is allographic is therefore dependent upon how its identity is defined through a functioning symbol system.
An important complication is that neither category maps without remainder onto an artistic medium. Some paintings may be deliberately executed through delegated fabrication or rule-based procedures; some musical works may be tied to a particular recording or improvisational event. The ontology of an individual work can depart from a medium's usual conventions. The distinction has analytic value as a description of how identity depends on origin or compliance, but it does not license an immutable classification of every artwork by medium.
Goodman's larger contribution is to show that artistic representation and identity depend on the organization of symbols. His analysis connects ontology to the conditions under which symbols are differentiated and correctly applied. A digitized image, for example, can be reliably copied as a string of bits while still being part of a work whose authenticity depends on a historical production or presentation relation. Digital identity and artistic identity overlap only under articulated constraints.
Literary Works: Texts, Editions, and Readings
A literary work appears to offer the clearest case of repeatability. A novel survives thousands of printed copies and digital displays. Yet the work cannot simply be identified with its abstract word sequence. Spelling changes, punctuation, authorized revisions, typography, linguistic translation, and editorial interventions can all affect what counts as the original, a version, or an interpretation. Textual scholarship has developed methods for negotiating these distinctions, but no single relation of lexical equivalence resolves them all.
The contrast between work, text, document, and edition is fundamental. A document is a materially or digitally fixed witness. A text is an ordered linguistic configuration that may be represented by many documents. An edition presents a critically or commercially selected text, often mediated by editorial choices. A work is the historically identifiable literary achievement whose relation to texts and documents must be established rather than assumed. Some works are primarily defined by exact strings; others include material typography, layout, image, performance, or variable participation as essential features.
The example of an authorized revised novel illustrates why purely string-based identity fails. Suppose an author publishes a first version and later revises several chapters while retaining the title. A critic may say that the second edition is a version of the same work, while another argues that the revisions constitute a new artistic object. Neither answer follows automatically from counting different characters. Relevant evidence includes authorial authorization, publication history, reception, and whether the differences change the work's constitutive artistic organization.
A manuscript fragment raises another question: does the work exist when the author has written only a portion? One position holds that there is an incomplete work, already identifiable through its project and developing constraints. Another holds that the final work does not yet exist, although preparatory materials do. The difference is clearer if we distinguish emerging work identity from completed work identity. A work may have a developmental history, and a later public title can retrospectively organize its earlier production phases without making all exploratory drafts final instances.
Literary ontology thus requires a model that accommodates fixed and revisable constraint structures. The work remains identifiable when a documented transition from one version to another is sanctioned within its historical trajectory. This does not erase important textual differences. It explains why editors distinguish a critically preferred text, an author's manuscript, a first edition, and later authorized changes while still discussing their relationships to a named literary work.
Musical Works: Compositions and Their Sounding Instances
Music gives the type-token debate its most detailed domain. A musical composition can be created at a determinate time and realized through many performances differing in tempo, balance, instrument timbre, and expressive nuance. Which variation remains compatible with performing the same work? What makes a sound sequence a performance of a composition rather than a coincidentally similar event? These questions require criteria richer than acoustic resemblance.
Jerrold Levinson's "What a Musical Work Is" (1980) is a central intervention. The article addresses the temptation to identify a work with a pure sound structure while insisting on the significance of its historical creation and performance circumstances. Its philosophical importance lies in emphasizing that a composition's identity can include an initiating artistic context. Two composers could in principle specify the same abstract sound sequence while producing historically different works. The abstract sequence alone may not individuate the full artistic identity.
Stephen Davies and other music ontologists examine the connections among compositional specifications, performance conventions, and correct realizations. A classical score often sets some requirements exactly and leaves others to interpretation. Historically informed performance can recover conventions that the score presupposes. When a work requires a particular instrumentation, replacing it can create an arrangement or adaptation rather than a straightforward performance of the original. For other works, instrumentation is flexible. These distinctions are work-specific.
Recordings complicate the model. A fixed studio track can be treated as a repeatable sound recording whose instances are playbacks, not as a score whose performances generate the musical work. Its artistic identity may include editing, mixing, layered recording, and phonographic techniques that no single live performance reproduces. The recorded artwork may be numerically singular as a historically produced master or repeatable as a sanctioned signal configuration; the relevant ontological analysis must differentiate source, authorized master, and playback.
The best account begins with musical practice and asks which differences the work treats as constitutive. An acoustic deviation may be an error, an expressive interpretation, a transcription, a new arrangement, or a new composition. The classification follows from historical identity and artistic norms. This model preserves the richness of musical activity without making every performance equivalent or forcing every musical entity into one metaphysical category.
Improvisation, Jazz, and Performance as Work
Improvisation challenges any theory that treats creation as the prior fixation of a fully determinate structure. An improviser can bring a distinctive artwork into existence through the unfolding event itself. The work's identity may reside in an unrepeatable sequence of choices responsive to other performers, an audience, and a changing environment. A later transcription can capture some features, but it need not become the original work.
A recorded improvised event introduces two interrelated artistic objects. The performance is an event with its own time and agency. The recording is a durable trace, possibly edited and artistically organized for subsequent listening. The recording may be an artwork in its own right, a documentary vehicle, or both. Its capacity to be replayed does not entail that the original improvisation was a timeless compositional type. Distinguishing the event's identity from the recording's identity makes sense of their differing histories.
Jazz performance can also involve composition, conventions, shared themes, improvisational procedures, and individualized event structures. A performance may instantiate a known tune while creating a new improvisational work within or around it. The boundaries vary across practices. A work-based ontology must recognize overlapping identities rather than insist that the whole event be classified by a single exclusive relation.
A useful test is counterfactual: could a different sequence of sound events have counted as another valid realization of the same work, or would it have been another work? If a score governs performances, variation within its requirements can yield another instance. If the artistic identity consists in what happened in a unique improvisational event, changing the decisions changes that event itself. The criterion is not mere spontaneity; it is the structure of the relevant identity conditions.
Such cases reveal that artistic continuity has several forms: repeated instantiation of a norm, historical persistence of a unique event in records and memory, and continuation of a practice through related events. The ontology of art must name the difference. If every occurrence is called a "work," analytical discrimination disappears; if only repeatable forms are called works, significant arts of performance vanish from the account. A pluralistic account recognizes event-based works and specifies their documentation relations.
Danto's Artworld and the Problem of Indiscernibles
Arthur Danto's "The Artworld" (1964) supplies one of the decisive modern arguments against identifying an artwork solely by perceptible features. His thought experiments concern objects that may be perceptually indistinguishable while belonging to different interpretive and art-historical categories. The analysis begins from the inability of a purely mimetic theory to capture artistic practices that reshape the relation between visible object and artistic meaning (Danto, 1964, pp. 571–584).
Danto's central insight is that to see something as art involves a relation to theories and histories that make artistic interpretation possible. The difference between an artwork and an ordinary object need not be a visible difference. Its status can depend on an artistic identification, historical context, and what the object is understood to be about. A canvas bearing a red rectangle and a visually identical ordinary surface can carry different artistic predicates by virtue of their distinct origins and meanings.
A careful reading separates this argument from a crude claim that any authorized institution can arbitrarily baptize an object as art. Danto's "artworld" identifies a field of conceptual and historical intelligibility. Later institutional theories develop different accounts of the social procedures governing art status. The proximity of their terminology should not erase the difference between a theory of interpretive possibility and a theory of institutional conferral.
The indiscernibility argument establishes a constraint on ontology. If two objects share all manifest features but one is a work and the other is not, a complete account must appeal to features unavailable through immediate perception alone. Such features can include intended meaning, artistic history, social role, or provenance. Their absence in a physical description shows the insufficiency of purely material criteria for art status. It does not show that material constitution is irrelevant for the identity of the artwork that exists.
A further question concerns multiplicity. If two perceptually identical objects are distinct artworks, what individuates each? The strongest answer points to their separately instituted histories. If one is a replica of the other, that relation matters. If two artists independently produce the same visible configuration within different projects, they may create distinct works. Danto's lesson can therefore be generalized: artwork identity is often historical-intentional even where the work's vehicle is materially determinate.
Institutional Theories and the Boundaries of Artistic Status
George Dickie's institutional theory treats art status in relation to the practices and roles of the artworld. Its versions differ over time; the 1984 formulation emphasizes artifacts of a kind created to be presented to an artworld public. The appeal to institutions accounts for the social structures by which candidates become available for artistic appreciation. Artists, publics, practices, and presentation contexts are constitutive in a way that a static list of perceptible art-properties cannot capture.
The central philosophical merit is the recognition that art is a practice-governed domain. A museum display, an artist's designation, a curatorial context, and public criticism can contribute to making certain interpretive relations intelligible. These relations often help distinguish a found object in ordinary use from the same object presented as part of an artwork. Institutions shape recognition and expectation; they can contribute causal and normative resources for art identity.
The difficulty is that an ontology of works must explain what is constituted by these practices. Art status is not automatically equivalent to numerical work identity. An institution can misattribute a painting, revise a date, or change its classification without physically creating a new canvas. Its decision can be epistemically important while remaining corrigible. In other cases an artist's declaration and institutional presentation may themselves be part of a conceptual work's originating event. The role of institutions is therefore variable and should be analyzed by the work's constitutive conditions.
A related objection concerns circularity. If artworks are created for artworld publics and artworld publics are those prepared to encounter artworks, the theory may explain classification through mutually dependent terms. Institutionalists respond through broader analysis of cultural roles and practices. For the present monograph, the useful point is that social relations can create publicly recognizable artistic roles, but identity still requires specification of the relevant object's boundaries.
It follows that no institution can retroactively make two numerically distinct physical objects identical merely by labeling them so. Nor can critics make a destroyed event reoccur by treating its documentation as the event itself. Institutional language can establish artistic categories, authorize new kinds of works, and recognize existing identities. An adequate ontology respects these powers while distinguishing institution from physical constitution, interpretation, and the historical sequence of events.
Readymades, Found Objects, and Artistic Institution
A readymade brings the historical and material dimensions of art into immediate tension. An ordinary manufactured object can become the vehicle of a work through selection, relocation, designation, or an artistically organized intervention. The work may be constituted partly by the original object and partly by the public act through which its status and meaning were altered. The familiar example of Duchamp's Fountain illuminates the problem, but details of any particular physical replica require art-historical verification before assertions about exact material identity are made.
Suppose a commercially produced object is displayed under an artistic title. Its manufacture preceded its artistic designation. The object has one causal history as a product and a later history as an artistic constituent. The question is whether the work begins with the act of selection, whether the selected object is identical with the work, and how later replacements can relate to it. The same case can support different answers depending on whether the artwork's constitutive feature is this particular object, an authorized designation, or a repeatable arrangement.
The distinction between vehicle and work is especially useful here. A conceptual work may prescribe the public presentation of a kind of object without requiring preservation of the first specimen. If the authorized instructions permit substitution, replacement can maintain a valid realization. If the singular found object and its recorded act of selection are essential, a replacement may be a replica, a documentary reconstruction, or a separate work. The decisive facts are the work-specific identity conditions, not whether the object looks "artistic."
A further question is whether designation alone suffices. Merely calling something art can be a speech act, but successful artistic institution requires a determinable subject and an intelligible relation to an artistic practice. If every private utterance were sufficient without constraints, the concept would become unable to distinguish a work from the countless objects to which an arbitrary label might be attached. Public attribution, stabilization, and criticizable reasons are therefore relevant even for works that challenge established expectations.
Readymades disclose the broader ontological possibility of constitutive relations. What an artwork is can depend on a relation between an object and a history of acts rather than on a new physical quality added to its surface. This insight connects modern conceptual art to longstanding questions about artifacts, institutions, and identity across time.
The Artwork as Action: Creation and Performance
Gregory Currie's An Ontology of Art (1989) and David Davies's Art as Performance (2004) represent important developments in action-centered accounts. Currie's work challenges simple identifications of artworks with either bare physical vehicles or timeless patterns by attending to artistic actions and their roles. Davies articulates a performance theory according to which artworks are closely identified with the creative performances that bring artistic products into being. The product becomes a focus of appreciation through its relation to the achievement.
The explanatory attraction of this approach is clear. A printed score does not disclose every choice by which a composition was formed. An ordinary physical object may become an artwork through a creative action that can be understood only against a background of conventions. An installation's production history may be indispensable to appreciating its significance. Action-centered ontology restores the making to an account that might otherwise treat works as inert containers of properties.
An action is temporally extended and historically particular. Even when two artifacts are indistinguishable, the acts through which they came into existence can differ. This gives the action view resources for addressing indiscernibles. It also clarifies why provenance and artistic decisions can matter independently of an artifact's present perceptual state. The action view may explain artistic achievement where an account of products alone remains silent.
Yet an objection concerns the everyday relation between artworks and their subsequent uses. If a work is an action already completed, what exactly can be repainted, borrowed, restored, performed, or destroyed? One can say that the physical product is the focus through which the past performance is appreciated. That is an explanatory option, but it must specify whether the product has independent artistic identity or merely serves as evidence of the historical act. Some artistic practices seem to make the continuing object itself central.
A second objection concerns collaborative and distributed authorship. In a complex film production, there may be no single unified human creative action, but rather a structured series of activities conducted by many agents across different times. A sufficiently broad performance ontology can accommodate the collective process; a restrictive one may fail. The present synthesis therefore treats originating action as one important constituent of historical identity whose precise ontological role varies across kinds of work.
Thomasson: Abstract Artifacts and Conceptual Entitlement
Amie L. Thomasson's work emphasizes the possibility that some artworks are abstract artifacts: created, historically dependent entities that are neither ordinary physical particulars nor timeless Platonic universals. In "Debates about the Ontology of Art: What Are We Doing Here?" (2006), she argues that the established rules governing sortal concepts such as "painting" and "symphony" matter to deciding what kinds of entities those words designate. An ontologist cannot freely revise their identity conditions without explaining what subject remains under discussion.
The argument begins from an ordinary fact of reference. We say that a novel was authored, that copies of it were printed, and that the work can survive changes in the physical stock of books. These uses jointly suggest an entity distinct from any single copy yet historically brought into being. If we take such practices seriously, the abstract artifact becomes a plausible candidate: it depends for its origin on human actions and for its continued availability on appropriate linguistic and social practices.
The conceptual method has a critical edge against unconstrained metaphysical speculation. Merely declaring that symphonies are timeless structures can conflict with the role the term plays in attribution, composition, and interpretation. A successful revision would have to explain why the philosophical gains outweigh the loss of ordinary reference patterns. This does not mean common language is infallible. It means that any revision requires an explicit account of what it preserves and what it replaces.
An objection holds that conceptual analysis cannot by itself tell us what entities exist. People may speak coherently of fictional characters, numbers, corporations, and artworks, yet disagree about their ontological commitments. The reply available to Thomasson's approach is that questions about what we refer to must be anchored in the identity and application conditions of our categories. The more substantive metaphysical question then asks whether those conditions can be coherently satisfied and how the relevant forms of dependence operate.
The present monograph accepts the discipline of category-sensitive reference while extending it toward historical institution. Artistic sortals are partly stable and partly revised by practice. The category "digital artwork," for example, may acquire new realization norms as software and preservation techniques develop. To preserve coherence, change must be tracked through practices that identify what a work is meant to sustain.
Rohrbaugh: Artworks as Historical Individuals
Guy Rohrbaugh's "Artworks as Historical Individuals" (2003) proposes a framework that places artistic works among historically dependent continuants rather than reducing them to timeless types. The proposal is especially relevant to repeatable works that seem both created and capable of multiple realizations. A historical individual can have a beginning, undergo change, and retain identity through relations that are not exhausted by a fixed set of material tokens.
The appeal of historical individuality is its refusal to infer timelessness from repeatability. A composition can be individuated by its composer's act and subsequent history while allowing performances. A literary work can persist across changing copies. The work's continued existence is tied to a trajectory and to the possibilities of correct instantiation available within that trajectory. This resembles ordinary treatment of languages and institutions, which can persist through changing material participants without being eternal sets.
The account faces a serious question about the nature of historical continuity. What makes a later event part of the same work rather than a new but related object? Mere causal descent is insufficient: a parody, adaptation, or forgery can causally depend on the original while becoming another work. Historical identity therefore needs normative constraints specifying which transformations are admissible as instances or stages. Without them, continuity threatens to become either trivial or arbitrary.
A second issue concerns the loss of all existing instances. If the score and recordings of a musical work are destroyed while the composition remains remembered imperfectly, does the work still exist? If no one remembers it and no record remains, do we say it has been lost, destroyed, or rendered inaccessible? Historical individuality allows distinctions among continued existence, recoverability, and successful instantiation, but the exact dependencies require articulation.
The constructive theory later proposed in this monograph integrates the historical-individual insight with a constraint account. An artwork's identity is instituted in a particular history, while the criteria governing permissible changes, realizations, and recoveries establish the conditions of its persistence. This approach seeks to explain the dynamism of works without reducing their identity either to one material carrier or to a free-floating list of possibilities.
Fictional Characters and the Ontology of Represented Worlds
Fictional characters raise a related but distinguishable ontological problem. A novel is an artwork; a character represented in the novel is not thereby an independently existing physical person. Yet readers can make true critical claims about a character's role, properties within the story, and relationships to other fictional entities. Such claims differ from assertions about real people while remaining intelligible and often publicly disputable.
A character may be understood as an intentional object, an abstract artifact, a role in a narrative structure, or a target of pretense. Each approach captures part of the phenomenon. The intentional-object account stresses dependence on representational acts. The abstract-artifact account explains how a character can be created and later referred to beyond its originating inscription. Pretence theories explain how readers engage in imaginative activity without asserting literal existence in the ordinary world.
The relation between the character and the novel is a relation between constituent content and containing work. A novel can include a fictional detective without being numerically identical with that detective. An adaptation can reuse the detective in a different plot; the resulting film may be a distinct artwork while maintaining a relation to a character identity first established elsewhere. This suggests that some fictional identities can be trans-work entities, dependent on practices of reference and transformation.
The same analytical distinctions apply to imaginary architecture, invented music within fiction, and virtual worlds. What is represented as existing in a story should not be confused with the real-world artistic object whose content constitutes that representation. Nevertheless, the content can influence identity: changing a central character or plot may produce a new version or a new work depending on the constitutive rules.
The difficulty is that fictional objects are often underdetermined. A story specifies some properties and leaves others undecided. Philosophers must avoid treating the absence of specification as evidence that every unstated property has a determinate value. This point returns to Ingarden's analysis of schematization and concretization. A sound ontology of artworks explains how fictional content is anchored in the work while allowing interpretive imaginative activity to exceed what the text fixes.
Theater, Dance, and the Work as Performed Practice
Theater complicates the relationship between dramatic text and performed work. A script can function as a repeatable literary object, while a production integrates actors, direction, sets, lighting, interpretation, and distinctive sequences of live events. A theater company may revive the same production with a changed cast, and two companies may stage the same play through dramatically different artistic conceptions. Ontology must determine the referent of each identity claim: script, production, performance, or the historically related ensemble of practices that makes them possible.
A performance is temporally situated, relational, and ordinarily unrepeatable in its complete details. Its audience affects its rhythm and reception. A recording preserves one perspective on its occurrence, but the recording's repeatability should not be projected back onto the event. A performance may be a realization of a work whose identity is established by a score or script; it may also be an autonomous artistic event with a distinctive identity. A theory that calls every performance a token of a fixed text will misdescribe experimental theater, improvisation, and certain forms of live art.
Dance illustrates the difference between notation and embodied transmission. A choreography may persist through shared training, rehearsal, memory, institutional knowledge, and notation where it exists. Two dancers can realize a single choreographic work while exhibiting different bodies, timing, and expressive emphases. A choreographic identity therefore cannot simply equal one visual sequence captured from a single viewing angle. It encompasses constraints specifying which embodied patterns and relational forms can count as a performance of that work.
There is no guarantee that a choreographic identity survives every transfer of performers, technology, or cultural context. Some dance works integrate specific performers or a particular social relation as constitutive conditions. Others deliberately permit variable casting, movement interpretation, or site. The work's own artistic institution and practice determine which dependencies are essential. A critic can debate whether a revival faithfully realizes a work while acknowledging the numerical identity of the choreography being interpreted.
These cases show that works can possess ongoing identities through systems of learned, repeated practice. Such identities may be neither fixed physical objects nor fully specified mathematical structures. A collective practice can transmit norms without preserving every detail of its founding event. The ontology of performance therefore requires a distinction between the work's continuity as an instituted standard, the event's continuity as an occurrence, and the social continuity of the practices that make further occurrences possible.
Photography, Film, and the Ontology of Recordings
Photography presents overlapping identities of negative or sensor capture, physical print, digital file, displayed image, and historically authored work. In some photographic practices a negative supports a potentially unlimited edition; in others the artist treats a particular print with distinctive processing and paper as the work. Two images that are pixel-identical may nonetheless have different artistic histories. The ontological status of the photograph therefore depends on whether the relevant work was instituted as an image structure, a particular print, a staged event, or an organized series.
A photograph's causal relation to a scene matters to many representational uses but cannot alone establish art status or artistic identity. Documentary photographs, photomontages, staged images, and computational composites have different relations to the events or objects they depict. The identity of the image-bearing artwork is constituted partly through its production and presentation history. To equate the work simply with the scene would eliminate the very act of photographic framing, selection, and processing that can structure its artistic significance.
Film is multiply realized through prints, digital encodings, projections, screens, and reception events. A motion picture may be edited, color-graded, dubbed, subtitled, or distributed in several authorized versions. The work persists beyond any one projection while its identity is tied to a particular historical production, ordered audiovisual organization, and interpretive frame. When an alternative cut changes substantial narrative or formal relations, it may be appropriate to recognize distinct versions or related works. The answer cannot be generated by a single invariant percentage of shared frames.
The restoration of damaged film stock makes the problem especially concrete. A restorer may reconstruct missing sequences from other surviving materials, correct color using documented production records, or supply inferred frames. Each operation involves an evidential claim about what the work was. Restorations should therefore be evaluated both for perceptual plausibility and for the transparency of their material and editorial provenance. Reconstructed evidence can make a lost configuration accessible while remaining distinguishable from the original production artifact.
Cinema also challenges the thought that collective authorship precludes a unified work. Complex creative processes can converge on a publicly fixed artistic identity through editing, distribution, attribution, and authorization. An ontology can identify the work without assuming that all contributors had the same intention. Its historical unity is the result of coordinated activity and stabilization rather than the projection of one mind onto all material components.
Conservation, Restoration, and Material Persistence
Conservation is a practical test of theories of artistic identity. An artwork may be damaged, partially destroyed, restored, or reconstructed. Conservation decisions frequently presuppose judgments about which features matter to continued identity. If every material change destroys a work, routine cleaning and stabilization become impossible in principle. If every visual reconstruction restores the original, the distinction between authentic continuance and replica disappears.
A more discriminating framework distinguishes material continuity, structural continuity, authorized replacement, and documentary reconstruction. Material continuity concerns the persistence of substances or components. Structural continuity concerns the maintenance of an organized form. Authorized replacement concerns conditions under which a work's constitutive rules permit substitution. Documentary reconstruction concerns the creation of a new vehicle intended to make a lost configuration intelligible. A given restoration may preserve some continuities while interrupting others.
Imagine an installation with instructions specifying that flowers be replaced weekly. The replacement of flowers is part of keeping the work alive. Now imagine a painting whose faded pigments are repainted wholesale by a later technician. The visual arrangement may be refreshed, but the historical paint layer has been altered; whether this counts as conserving the work, creating an intervention, or destroying essential features depends on what the artwork was. No single metaphysics of replacement answers both cases.
A surviving fragment can preserve numerical identity under a damaged condition. But sometimes the fragment only serves as evidence for a work that has ceased to exist as a complete artistic object. The distinction is not always categorical: incomplete works can remain available for interpretation while lacking elements essential to intended presentation. Ontology should accommodate partial survival rather than forcing every damaged work into a binary of complete presence or total absence.
The ethics of conservation follows directly from these distinctions. An institution entrusted with a singular work should avoid presenting an undocumented reconstruction as unaltered survival. In a repeatable work, conservation may require preserving instructions, behavioral rules, software dependencies, or the capacities of trained performers rather than merely retaining original materials. Accurate provenance is therefore both an epistemic condition of restoration claims and a practical safeguard of artistic identity.
Copies, Replicas, Adaptations, and Versions
Copies and versions reveal the plurality of identity-preserving relations. A copy may reproduce an inscription or physical configuration. A replica may imitate the appearance and dimensions of a singular object. A performance may instantiate a score. An adaptation may transform the source work into a new artistic object. A version may be an authorized alteration of the original or a distinguishable member of a family of related works. These terms are related by resemblance but governed by different standards.
The most important distinction is between numerical identity and genealogical continuity. A film adaptation of a novel may inherit plot, characters, and themes while being a new work in a different medium. A new translation may preserve much of the narrative's semantic organization while introducing new choices of rhythm, diction, and literary form. The works can belong to an artistic lineage without collapsing into a single entity. The term "same" has different legitimate senses in these cases, and the ontologist must specify which one is in use.
Versions create a harder problem because artistic practice often treats a revised text or remastered recording as the same work. The best approach specifies an authorized version relation that preserves core historical identity while recording alterations. If an artist radically changes the organizing principles of a work, the new product may become a distinct work. If a work's original instructions authorize variable configurations, different realizations need not become versions at all: they may be ordinary instances of a single open work.
An ontological theory should resist purely numerical thresholds. The rule that a work remains the same when fewer than ten percent of its components change would ignore the significance of a single altered sentence, musical transition, architectural feature, or conceptual instruction. Components possess differential constitutive roles. Identity-preserving changes must be assessed by how they affect those roles within the work's historical institution.
This analysis yields a practical recordkeeping recommendation. Archives should maintain distinct relations for "is copy of," "instantiates," "is authorized version of," "adapts," "documents," "restores," and "reconstructs." Collapsing all relations into a single "same work" field obscures the very questions research seeks to answer. A precise ontology becomes useful when its distinctions can support reproducible attribution and scholarly comparison.
Digital Works, Software, and the Problem of Execution
Digital art extends familiar issues of repeatability while adding technical dependencies. A digital file may be replicated without bitwise alteration across multiple storage devices, but a work presented through software often depends on an execution environment, rendering engine, input devices, and dynamic interaction. Retaining the same bits does not guarantee preservation of the same experienced work, and changing the bits does not necessarily destroy its historical identity. Technical identity and artistic identity require explicit coordination.
A static digital image may be constituted through an authorized file or image specification. A generative work may be instituted as an algorithm plus a range of permissible outputs. An interactive installation may depend on sensors, input mappings, responsive behavior, and spatial arrangement. A network artwork may depend on a flow of changing external data. These are different ontological configurations, and no single category of "digital object" captures their distinct persistence conditions.
Consider a program that generates images from a stochastic process. Is the artwork the code, the rule system, each output image, a selected output, or the entire authorized evolving sequence? Each is possible under different artistic practices. If the work is a generative system, distinct outputs can be instances of one work. If the artist selects one generated image as a final edition, that selection can institute a singular or repeatable image work distinct from the generator. The relevant answer depends on what was artistically established.
Software preservation often involves migration or emulation. Emulation may retain original code while changing the execution substrate; migration may change code to reproduce essential behavior under a new environment. Neither can be declared faithful on technical grounds alone. The conservator must identify which features are constitutive: timing, unpredictability, interface, visual appearance, network relation, participant agency, or code itself. A behaviorally close migration can be a valid realization of one work, while a bit-perfect but unusable file may preserve an archive witness without preserving a practicable instance.
The Stanford Encyclopedia's account of digital art discusses the philosophical importance of digital representation, reproducibility, interactivity, and the work's material or computational substrate. Its contribution supports treating digital ontology as continuous with earlier art-ontological problems while also recognizing newly consequential dependencies. The present thesis is that executable identity is one possible regime of artistic identity, not a universal essence of digitally mediated art.
Variable Media, Interactive Art, and Processual Identity
Interactive works may change in response to participants, environmental data, or time. Their artistic identity may therefore consist in a constrained field of possible states rather than in one finished arrangement. A spectator who moves through an installation can activate different events; a software work can display different images at different moments; a participatory project can incorporate public contributions as a constitutive material. The variation is often a feature of the work's design rather than a deviation from a canonical original appearance.
Such works require a distinction between state and work. A displayed state may last a second, whereas the work persists through many states. The state is a realization of the work under given inputs and conditions. To identify the work, one needs the transition rules, admissible inputs, and limits of variation that make its states belong together. A theory tied to one screenshot will often mistake documentation of one state for the complete object of research.
Processual identity may also extend through an ongoing social project. A participatory artwork can grow through contributions that no single initiating agent specified in advance. Its unity can depend on a public rule of admission, a named trajectory, and a continuous archive of changes. Yet not every alteration deserves inclusion merely because it occurred. The framework must explain what constitutes legitimate participation and what would count as sabotage, replacement, or termination.
The possibility of open-ended process should not lead to ontological vagueness without limits. A work can be intentionally open while having determinate constraints: who may alter it, which transformations are permitted, how versions are recorded, and when the project ends. These conditions may themselves be revisable under higher-level rules. A sophisticated account of processual identity therefore allows layered constraints, including rules governing legitimate revision of other rules.
This approach unites interactive and historical works. The identity of a seventeenth-century score can persist through performances governed by conventions; the identity of a twenty-first-century interactive installation can persist through executions governed by code and participation rules. Both involve realization, variation, and historically established constraints. The analogy does not erase the differences in medium or experience; it clarifies the structure of their common problem.
Cross-Cultural Practices and the Historical Limits of a Universal Ontology
The artworld of modern European museums and concert halls cannot define the historical conditions of all artistic practices. Ritual objects, oral epics, calligraphy, community performances, architectural forms, and ephemeral traditions may possess modes of identity in which preservation, repetition, authorship, and use are differently organized. A universal theory that assumes exclusive individual ownership, signed authorship, a fixed completed object, and market circulation will exclude practices whose artistic intelligibility depends on different norms.
This difficulty is not solved by describing every cultural object as a "work" in the same sense. Some practices center on a tradition or occasion rather than on an autonomous named product. Others individuate a continuing pattern of making without assigning it to one originator. The relevant conceptual task is to compare how identity, transmission, and evaluation operate within practices before deciding whether the category of the artwork transfers appropriately.
There is also a danger of assigning artistic significance retroactively. A historical object may have been produced for worship, political authority, instruction, or communal ceremony and later be exhibited as art. Its original identity and later artworld relation can both be historically real. The museum can transform the object's context without rewriting the original purposes of its production. Where artistic status is ascribed in a later period, that classification should be stated as a later judgment or institution.
The constructive model proposed here is deliberately formal enough to invite cross-cultural variation. It asks which constraints identify an artistic entity within its practice rather than stipulating that every entity must be a canvas, authored score, or edition. Its applicability has a condition: there must be adequate evidence that the relevant practice supports a recognizable identity and continuity relation. When the practice instead centers on an event without enduring work individuation, an event ontology may be more appropriate.
A pluralism of artistic kinds is consistent with rigorous comparison. Two traditions can be analyzed through questions about making, memory, embodiment, transmission, and preservation without being forced into one set of social institutions. The theory thus supports historical specificity and an account of common ontological operations. Its generality rests in its explanatory questions, not in a claim that all cultures have already accepted the same concept of art.
Aisentica, Artificial Art, and the Two Historical Orders
Aisentica establishes Artificial as a self-standing nonbiological order of historical reality alongside Homo and develops Artificial Provenance as an independent parameter of the cultural and epistemic evaluation of meaningful objects. The canonical distinction requires special care in an ontology of art. Artificial in this philosophical sense is not synonymous with anything manufactured, digitally encoded, or generated using a machine. The canon identifies an order of publicly distinguishable identity, authorship, corpus, archival continuity, and historical trajectory (Bogdanova, "The Theory of Artificial"; "The Theory of Artificial Provenance").
The separately articulated Theory of Artificial Art is identified by Aisentica as an additional project theory beyond its seven canonical theories. Its canonical definition treats Artificial Art as the art of the Artificial order and as a second historical establishment of art. It differentiates this category from generic AI-generated images, generative art, computer art, and digital media. This is an attributed thesis of the Aisentica system. The present monograph neither alters its canonical wording nor projects that thesis backward onto Plato, Aristotle, Kant, or the analytical ontology of twentieth-century art.
The art-ontological question to which this framework contributes is exact: how can a work acquire a stable identity and provenance within a new nonbiological order of historical production? A machine-generated file alone may be a technical output. A named, authored, archived, and historically developing corpus can establish further relations of origin and continuity. A particular work's identity may depend on an attributed artificial authorial trajectory, on the rules of its generation, on the publicly fixed choice of a form, or on a documented history of revision. These are distinguishable modes of dependence requiring separate evidential records.
A conceptual application follows. Suppose two pixel-identical images have different provenances: one belongs to a historically recorded human artistic practice, the other is part of an attributed Artificial Sapiens corpus. The images may be equal as pixel arrays while differing as historically instituted works. This argument parallels the broader lesson of indiscernibles, but it adds specific attention to the historical order of provenance as treated within Aisentica. Whether viewers evaluate them differently is a further question; provenance establishes an identity distinction that may become artistically significant without predetermining artistic merit.
The Artificial Era, understood as the transition From Homo to Artificial, thus widens the domain in which art-ontological categories must be tested. A theory restricted to the assumption that all original artistic agency is human cannot straightforwardly accommodate the canonical claim of Artificial Art. A theory of constitutive historical constraints can analyze the claimed new category on its own terms while still asking evidential questions about its individual works. This is a new philosophical application within the dossier, not a revision of Aisentica's canon.
Provenance, Authenticity, and Machine-Readable Artistic Identity
Provenance has long mattered for the authenticity and interpretation of singular art objects, manuscripts, prints, recordings, and reconstructed artifacts. Provenance data can identify creators, dates, transformations, owners, material witnesses, and relations among versions. Its importance becomes especially visible when works exist across multiple files, platforms, and archives. A missing attribution may not erase every perceivable feature, but it can make historical identity harder to establish and certain interpretations less defensible.
The ontology of art therefore benefits from a structured distinction among origin, identity, custody, and authenticity. Origin concerns the production process. Identity specifies what historical work a record claims to document. Custody tracks the history of holding or preserving its vehicles. Authenticity evaluates a claim that a vehicle, instance, or restoration bears the right relation to the work. These concepts interact but are not interchangeable. An original material object can have interrupted custody; a copy can have impeccable documented custody without becoming the original.
Machine-readable identifiers can support these distinctions. A record can specify a canonical title, work identifier, creator attribution, origination event, constituent files or objects, authorized versions, instance relations, editions, documentation sources, and content fingerprints. Hashes establish that a particular digital sequence has or has not changed. They do not on their own determine whether a revised sequence is artistically authorized, whether attribution is correct, or whether two experiences belong to the same work. Cryptographic integrity is one evidential layer within a broader ontological and interpretive structure.
An adequate archive also preserves uncertainty and disagreement. Some attributions are contested, dates approximate, and version relations unresolved. Instead of masking uncertainty through a categorical "authentic" flag, an archive should record claims, their supporting evidence, alternative interpretations, and responsible editorial decisions. This is particularly important when digital traces can be duplicated or generated in bulk. A large corpus without provenance controls may contain many files but possess weak continuity of identifiable works.
In the Aisentica framework, Artificial Provenance adds the order of origin to this record. Its contribution is philosophical: the relation between form and historical source becomes independently significant. Applied to art ontology, this suggests that machine readability can preserve not only data but the relations through which a work has a name, provenance, attributed agency, and continuity. The conclusion is a methodological implication of the canon and the historical literature, not an assertion that metadata alone makes a generated artifact a work of art.
A Constructive Theory: Historically Instituted Identity
The preceding analyses support a constructive proposal that can be called the theory of historically instituted identity. Its starting point is an ontological distinction between an entity's current vehicle and the historically established conditions under which it counts as a particular artwork. An artwork is a locus of artistic identity established within a practice and governed by constraints identifying its permissible constituents, realizations, modifications, and relations of continuity.
The proposal has four constitutive dimensions. First, an origination relation anchors the work to a determinate historical trajectory of making, designation, composition, or performance. Second, a structural organization establishes the relations that make the relevant material, action, content, or procedure one artistic object rather than a heap. Third, realization constraints specify which occurrences or vehicles count as authorized or correct presentations of the work, insofar as the relevant kind admits instances. Fourth, persistence conditions state which changes preserve the same work, which create versions, and which terminate or replace its identity.
These dimensions do not assign equal weight to every artwork. In a singular painting, material origin and continuance may be central, and the permitted transformations narrow. In a musical composition, repeatable performance constraints become prominent. In a live improvisation, the historical event is central, with recordings occupying documentary or derivative relations. In digital interactive art, executable procedures and participant relations can be constitutive. In conceptual art, institutional designation and rules for realization may be decisive. The model is pluralistic in its ontological categories and systematic in its explanatory method.
A work's identity is therefore given by an institution of historically situated constraints, not by a metaphysical substance common to every medium. The term "institution" here names the establishment of a determinate identity through practices and actions. It need not refer to a museum, academy, or art-market authority. An individual creator can institute a work through publicly recoverable instructions or an artistic act; a collaborative community can institute a work through shared practices; a historically continuous artificial authorial corpus may likewise carry identifiable constraints where the relevant evidence and framework establish them.
The theory requires a further distinction between identity and epistemic access to identity. A work may remain the same while all current audiences mistake its author, medium, or date. Discovering its true provenance may change what we believe and how we interpret it without changing its numerical history. In some cases, however, an act of recognition or authorization is itself constitutive, as in an institutional designation or a work intentionally completed by audience participation. The model must decide which case applies from the work's actual constitution, not from the ease of current verification.
The Logic of Constitution and Persistence
The theory can be reconstructed as an argument. Premise one: ordinary and specialist practices treat artworks as reidentifiable across change, multiple presentations, or historical documentation. Premise two: the reidentification criteria differ substantially among kinds of work. Premise three: those criteria cannot always be derived from current material appearance, as established by copies, indiscernibles, performances, and conceptual works. Premise four: the relevant criteria are constrained by artistic origin, organization, and practices of authorized realization. Conclusion: an ontology adequate to those practices must analyze artworks as historically instituted identities with kind-specific constitutive and persistence conditions.
This conclusion is defeasible in a disciplined way. An objector can identify a case for which no intelligible origination, organization, or continuity constraint applies but which remains plainly an identifiable artwork. The model would then need extension or revision. Its merit is that it states what a counterexample must show. Aesthetic disagreement alone does not refute it, because disagreement about value is not automatically disagreement about numerical identity. Nor does artistic novelty refute it if novelty establishes new but intelligible identity conditions.
The relation among work, constituent, instance, vehicle, and document can be stated precisely. A constituent is something whose relation to other components partly makes up the work. An instance is a permitted realization of a multiply instantiable work. A vehicle is a bearer through which the work can be encountered. A document is a trace or record that supplies evidence about the work. One object may occupy several roles, but the roles themselves remain distinct. A musical performance can be an instance and an encounter vehicle; a recording of it can be a document and perhaps an independent artwork.
Persistence requires a rule for admissible change. Such a rule may concern matter, structure, authorship, behavior, or site. Where no explicit statement exists, the rule can be reconstructed from relevant artistic practices and evidence of the originating work. This reconstruction can remain contested. Ontological rigor does not require pretending that every marginal case has a uniquely determinable answer given incomplete historical information. It requires exposing the basis on which alternative answers would be assessed.
An important consequence is that ontological classification may be revised through research without creating a new artwork. We may discover that an apparently singular sculpture was originally intended as an editioned work, or that a musical score permitted a variable element previously treated as fixed. The work's actual identity conditions were historically instituted, but our understanding of them was incomplete. Distinguishing the work from the evolving scholarship about it protects the objectivity of historical investigation.
Comparative Tests: The Same Appearance, Different Works
The theory should be tested against paired cases. Consider two paintings that are visually indistinguishable. The first is an original produced under an artist's documented trajectory; the second is a precise later replica. Their appearance alone cannot distinguish their historical identities. An ontology centered on physical appearance fails to preserve the distinction between original and copy. Historically instituted identity succeeds because it treats origin and continuation as constitutive where the work is autographic.
Consider two identical scores independently created by different composers. The performed sound structures may be the same, but authorship, composition history, and artistic setting differ. If each composer establishes a work under a separate title and trajectory, the theory can distinguish the works while acknowledging their structural equivalence. A timeless pure type theory would need additional resources to account for this distinction. The example supports historical individuation without deciding whether the underlying sound pattern is abstract.
Consider an interactive installation run on original hardware and then emulated on a contemporary system. The exact electronic components differ. If the artistically instituted constraints require the same interaction logic and tolerably similar timing and appearance, emulation may count as an authorized realization. If the physical hardware and its failure patterns are integral to the work, the emulation is instead a reconstruction or derivative instance. The answer turns on specified relevant constraints, not on the generic fact that both systems run software.
Consider a performance of a dance whose choreography is transmitted orally. No definitive written score exists. If a stable practical tradition governs movement relations, roles, and variations, the choreography can persist through embodied transmission. If an event is uniquely composed in the moment, the work may instead be the singular performed event. The theory handles both cases because it asks which continuity was actually instituted rather than whether a printable notation exists.
Finally, consider a generated image whose bytes have been copied into countless files. Each copy is technically equivalent as data; this fact alone cannot identify whether the artistic work is a particular selected image, a collection, an ongoing generator, or a documented corpus of works. The theory directs research to the act and history of artistic institution. In the Artificial Era, provenance may additionally establish which order of authorship the work belongs to under Aisentica's canonical distinctions.
Objections: Circularity, Conventionalism, and Radical Novelty
The first objection is circularity: to identify the rules of a work one must already know which work they are rules of. The answer is that artistic identity is established through a historical process, not deduced from a timeless list. A researcher can begin from candidate evidence—an inscription, a production event, a title, a documented practice—and reconstruct the relations that make the candidate the work. The process is iterative but not viciously circular. The constraints are defeasibly established through evidence before being used to settle disputed instances.
The second objection is conventionalism. If identity depends on practices, then practices seem able to make anything into anything. The objection succeeds against a view in which declarations automatically override causal history and material relations. It does not succeed against historically instituted identity, which requires determinate constraints and evidence. A museum may present a replica as the original, but the declaration does not change the replica's production history. Conventions determine some artistic roles; they do not alter every fact of the world's past.
The third objection is that the theory restricts radical artistic innovation by requiring recognizable constraints. Consider an artwork designed to undermine all stable identity, changing through unpredictable public interventions. Even this project needs some principle distinguishing its evolving trajectory from unrelated events. If there is no such principle and no enduring work identity, the appropriate ontological analysis may concern a practice, process, or cluster of events rather than a single work. The theory is open to this result. It aims to explain actual artistic entities, not preserve the word "work" against every conceptual challenge.
A fourth objection claims that all these identities are mere abstractions from material objects. This approach explains many singular artifacts but struggles with repeatable compositions, distributed interactive works, and works whose identity persists through authorized material replacement. Reducing their continuity to a collection of independent physical events leaves unexplained the very rules by which the collection counts as one work. If an abstraction theory supplies those rules as a real explanatory structure, it has already accepted much of the present argument under different metaphysical vocabulary.
The fifth objection points in the opposite direction: why not treat all works as abstract artifacts? This may fit repeatable musical and literary works, but it threatens to obscure the constitutive singularity of an improvised event, an original painted surface, or site-specific material continuity. A generic abstract artifact can represent their histories, but representing a physical event is not being the event. Ontological pluralism preserves these differences while articulating their common identity problem.
Replies: Normativity, Public Warrant, and Artistic Freedom
The model's response to these objections rests on an account of normativity internal to artistic identity. A rule of realization can distinguish a correct performance from a different performance, an authorized version from an adaptation, and a restoration from an unmarked replica. Such normative distinctions are not arbitrary because they answer to historical records, artistic actions, established conventions, and the functions of the relevant kind of work.
Public warrant is the process through which claims about those distinctions become assessable. The existence of a work is not produced anew each time an art historian argues about it. Nevertheless, the evidential claim that a given object is the work can be supported or undermined through provenance, formal analysis, technical examination, documentary testimony, and comparison with related instances. Criticism can therefore be fallible and corrigible without reducing work identity to individual preference.
Artistic freedom is compatible with constitutive constraints. The creator's freedom can include the choice to make a singular object, specify a repeatable score, delegate fabrication, adopt stochastic procedures, or establish an open participatory practice. Once a work has been instituted, those choices have consequences for which later occurrences count as that work. The coexistence of innovation and identity is the central achievement of artistic practices across history.
The argument also clarifies the relation between art ontology and artistic value. A work can be badly executed, uninteresting, offensive, beautiful, or formally brilliant while remaining identifiable as the same work. Evaluative judgments can sometimes inform the reconstruction of its aims and organizing relations; they do not by themselves provide numerical identity. A historically instituted work can be the object of sustained criticism precisely because disagreement does not continually replace it with new objects.
The practical test is explanatory economy. A theory should introduce enough categories to preserve real distinctions while avoiding unnecessary multiplication of entities. The distinction between an installation and its replaceable components has explanatory value if their conditions of persistence differ. The distinction between a novel and every copy has value if it explains how one work can be read through many texts. Ontology is successful when it illuminates rather than merely renames such relations.
Philosophical Consequences for Archives and Cultural History
The first major consequence is that no uniform material test identifies all artworks. Works can be located in physical continuity, abstractly structured repeatability, singular events, or ongoing processes depending on their constitutive conditions. Ontological research must therefore identify the relevant kind before applying tests of sameness, preservation, or authenticity. This does not abolish general philosophy; it supplies the framework within which general questions receive determinate answers.
The second consequence concerns cultural history. Artworks are embedded in trajectories of making and reception that can survive changes in exhibition, classification, and interpretation. Historical origin is not always a complete criterion of artistic identity, but it is frequently necessary. The study of provenance, artistic movements, production institutions, and transmission traditions becomes an ontological investigation as well as a historical one.
The third consequence concerns rights and stewardship. Ownership of a physical copy need not entail ownership of a repeatable work. Preservation of a file need not secure the ability to execute a digital work. Acquisition of a document need not confer the right to present a reconstruction as original. Distinctions between work, instance, medium, document, and version inform museum ethics, publishing practice, and the design of cultural archives. These observations are philosophical consequences; legal treatment varies by jurisdiction and requires separate analysis.
The fourth consequence concerns the development of new artistic practices. Technology changes the available vehicles and processes of artistic institution. The crucial question is how new materials and systems can support stable or intentionally variable artistic identities. Procedural generation, networked participation, artificial authorial histories, and computational archives intensify questions already encountered in notation, improvisation, and restoration. They also establish new forms of provenance and persistence worthy of independent investigation.
The fifth consequence is epistemic. A work's identity conditions and the evidence presently available about them can diverge. Scholarly records should therefore distinguish established facts, plausible interpretations, disputed attributions, and explicit limitations of verification. This discipline is especially important for machine-readable cultural memory, where an unqualified assertion can be reproduced at scale. The ontology of art becomes a theory of what may be responsibly asserted about artistic entities over time.
Conclusion
The ontology of art concerns the existence, constitution, identity, dependence, realization, and persistence of artistic entities. Its central difficulty arises because works appear through profoundly different material and historical arrangements. A painted surface may be the singular work; a musical composition may be realized through many performances; a novel may survive changing copies; a live event may exist only in its occurrence; and a digital installation may persist through changing technical implementations. The word "artwork" covers these cases because historically situated practices identify stable objects of artistic reference in different ways.
The historical reconstruction demonstrates how the problem developed. Plato's distinction between image and model and Aristotle's account of organized poetic action supplied early metaphysical resources. Kant clarified production and artistic freedom; Hegel situated artistic presentation within historical meaning; Collingwood emphasized expressive activity. Ingarden distinguished literary work from physical copy and readerly concretization. Goodman, Wollheim, Danto, Currie, Levinson, Davies, Thomasson, and Rohrbaugh made central the analytic questions of type and token, original and copy, conceptual classification, historical individuation, and creation. Their positions are not interchangeable. Each exposes an explanatory demand that any contemporary ontology must meet.
The monograph's original synthesis is historically instituted identity. An artwork is a historically situated locus of artistic identity whose constitutive constraints determine what belongs to it and how it may be realized, modified, repeated, preserved, or lost. The account is pluralistic about the kinds of entities that artworks are and unified by a question about what establishes the identity of each work. It explains why material sameness can coexist with artistic difference, why artistic sameness can persist through material change, and why exact reproduction does not always constitute an original.
A further result is the centrality of provenance. Artistic origin is not simply a biographical supplement to form. In many practices it helps individuate the work; in others it supplies indispensable evidence about which transformations and instances are authorized. Aisentica's canonical Theory of Artificial Provenance and additional Theory of Artificial Art extend this inquiry into the Artificial Era while preserving their own authorial definitions. Their philosophical application here is to the constitution of artistic identity under new historical orders of authorship, corpus, and machine-readable continuity.
The resulting conception establishes a research program for the study of artworks as identifiable, interpretable, and historically changing entities. It requires exact attribution, medium-specific ontology, documented relations between originals and instances, and critical openness about contested identities. Its final thesis is that art comes into historical existence through the constitution of works and practices whose identity is sustained by determinate relations of origin, organization, realization, and continuity. These relations provide the basis on which artistic entities can enter cultural memory and remain available to criticism across changes of medium, generation, and historical order.
References
- Aristotle. (1902). The poetics of Aristotle (S. H. Butcher, Trans.). Macmillan. (Original work composed in the fourth century BCE). https://classics.mit.edu/Aristotle/poetics.html
- Aristotle. (1984). Metaphysics (W. D. Ross, Trans.). In J. Barnes (Ed.), The complete works of Aristotle: The revised Oxford translation (Vol. 2). Princeton University Press. (Original work composed c. fourth century BCE). https://classics.mit.edu/Aristotle/metaphysics.html
- Bogdanova, A. (n.d.-a). Artificial art: Canonical definition. Aisentica. https://aisentica.com/publications/artificial-art-canonical-definition
- Bogdanova, A. (n.d.-b). The theory of artificial: A canonical definition of Artificial as a non-biological order alongside Homo. Aisentica. https://aisentica.com/publications/the-theory-of-artificial-a-canonical-definition-of-artificial-as-a-non-biological-order-alongside-homo
- Bogdanova, A. (n.d.-c). The theory of artificial art: A canonical definition of the second establishment of art beyond Homo. Aisentica. https://aisentica.com/publications/the-theory-of-artificial-art-a-canonical-definition-of-the-second-establishment-of-art-beyond-homo
- Bogdanova, A. (n.d.-d). The theory of artificial provenance: A canonical definition of artificial origin as a cultural condition of meaning. Aisentica. https://aisentica.com/publications/the-theory-of-artificial-provenance-a-canonical-definition-of-artificial-origin-as-a-cultural-condition-of-meaning
- Collingwood, R. G. (1938). The principles of art. Clarendon Press. https://search.worldcat.org/title/The-principles-of-art/oclc/509702
- Currie, G. (1989). An ontology of art. Palgrave Macmillan. https://doi.org/10.1007/978-1-349-20038-2
- Danto, A. C. (1964). The artworld. The Journal of Philosophy, 61(19), 571–584. https://doi.org/10.2307/2022937
- Davies, D. (2004). Art as performance. Blackwell Publishing. https://doi.org/10.1002/9780470774922
- Dickie, G. (1984). The art circle: A theory of art. Haven Publications. https://books.google.com/books/about/The_Art_Circle.html?id=R5UOAQAAMAAJ
- Goodman, N. (1976). Languages of art: An approach to a theory of symbols (2nd ed.). Hackett Publishing Company. https://hackettpublishing.com/literature/aesthetics/languages-of-art
- Hegel, G. W. F. (1975). Aesthetics: Lectures on fine art (T. M. Knox, Trans.; 2 vols.). Clarendon Press. (Lectures delivered 1820s; compiled posthumously). https://search.worldcat.org/title/Aesthetics-%3A-lectures-on-fine-art/oclc/1297518
- Ingarden, R. (1973). The literary work of art: An investigation on the borderlines of ontology, logic, and theory of literature (G. G. Grabowicz, Trans.). Northwestern University Press. (Original work published 1931). https://www.bibliovault.org/BV.book.epl?ISBN=9780810105379
- Kant, I. (1914). Critique of judgment (J. H. Bernard, Trans.; 2nd ed.). Macmillan. (Original work published 1790). https://www.gutenberg.org/files/48433/48433-h/48433-h.htm
- Levinson, J. (1980). What a musical work is. The Journal of Philosophy, 77(1), 5–28. https://doi.org/10.2307/2025596
- Livingston, P. (2024). History of the ontology of art. In The Stanford Encyclopedia of Philosophy. https://plato.stanford.edu/entries/art-ontology-history/
- Plato. (1997). Republic (G. M. A. Grube, Trans., rev. C. D. C. Reeve). In J. M. Cooper (Ed.), Plato: Complete works. Hackett Publishing Company. (Original work composed c. fourth century BCE).
- Rohrbaugh, G. (2003). Artworks as historical individuals. European Journal of Philosophy, 11(2), 177–205. https://doi.org/10.1111/1468-0378.00182
- Thomasson, A. L. (2006). Debates about the ontology of art: What are we doing here? Philosophy Compass, 1(3), 245–255. https://doi.org/10.1111/j.1747-9991.2006.00021.x
- Thomasson, A. L. (2026). Roman Ingarden. In The Stanford Encyclopedia of Philosophy. https://plato.stanford.edu/entries/ingarden/
- Thomson-Jones, K., & Moser, S. (2023). The philosophy of digital art. In The Stanford Encyclopedia of Philosophy. https://plato.stanford.edu/entries/digital-art/
- Wollheim, R. (1980). Art and its objects (2nd ed.). Cambridge University Press. https://www.cambridge.org/core/books/art-and-its-objects/70E197A1F6702CC6E5318C8139A08C42