IN SEARCH OF THE ONTOLOGICAL COMMON CORE OF ARTWORKS: RADICAL EMBODIMENT AND NON-UNIVERSALIZATION

I propose that artworks represent a specific and homogeneous ontological kind, grounded in a common ontological core. I call this common core ‘non-universalizable embodied meaning’, and I argue that (a) this common core explains how artworks unfold their ontological identity at the physical, intentional, and social levels on the basis of an original and irreducible mode of material embodiment and cultural emergence; (b) this common core functions as the constitutive rule of art (conceived as a social kind) and institutes an axiological normativity, that is, normativity based on aesthetic value; (c) it has a central role both in the theoretical definition of art and in the identification, recognition, and fruition of the individual artworks; (d) it enables artistic creativity, historical transformations of art, and the current, multifaceted state of art.

are ontologically heterogeneous. 2 The ontology of art outlined here suggests a radically different point of view. Although artworks undoubtedly present heterogeneous properties, especially in their physical implementation, they are ontologically specific and homogeneous because they share the same ontological common core.

I. ARTWORK (NOT) AS A PHYSICAL OBJECT
In order to understand the ontological common core of artworks it is useful first to define some general ontological distinctions. Here, 'physical object' means an entity that exists in space and time independently of the mind. 'Imaginary object' means an entity that exists as a product of mere imagination, without perceptible, external, public, enduring correlates. 'Ideal object' means an entity that exists outside space and time, though it may be subject to psychological acts. 'Social object' means an entity that exists in space and time by virtue of collective intentionality, although it is not a fictitious and unreal entity.
The starting point of my proposal is the level of physical description. A picture at this level is a canvas on which there are coloured pigments. Simply put, without physical realization, artworks would not have an actual existence. Indeed, by virtue of their physical implementation, artworks are realized in space and time, and are perceptible, thus they differ both from imaginary entities, which are devoid of external, public, enduring expressions, and from ideal entities, which are devoid of space and time location.
But, even if physically realized, the artwork is not reducible to a physical object because it cannot be identified with the constituting material. The artwork and the constituting material, in fact, have different conditions of identification and persistence. 3 A statue of clay survives the replacement of a part of it with another suitably shaped piece of clay, whereas the clay piece does not survive. In contrast, the clay piece survives the transformation of the statue into another object, whereas the statue does not survive.
Moreover, artworks cannot be reduced to physical objects because they are characterized by very different modes of realization. Artworks can be multirealizable: if a print, photograph, or literary work (such as a poem or novel), the artwork is implemented in a physical substrate, a copy that is replicable; thus the artwork has a status similar to a universal: it supports multiple manifestations but none of them can count as the artwork itself. Artworks can be multi-executable: if it is a theatre performance or work of music, the artwork is implemented in a physical substrate that consists of a performance, that is, an event occurring in the space-time dimension, thus it has a status similar to a set of instructions.
Artworks can be mono-realizable: if a painting or sculpture (at least a sculpture that is not based on a mould), the artwork is implemented in a physical substrate that is not reproducible and it has a status similar to a concrete and irreplaceable individual.
But it is easy to show that the modes of physical realization are more complex than this. Firstly, the different instances of a single category of art do not always share the same mode of implementation. Musical works realized as total improvisations are not multi-executable. They could, however, be filmed or recorded, and therefore emulated. Yet could these emulations be considered instances of multi-executable works? Similarly, sculptures produced by fusion, as well as etchings and lithographs, are not mono-realizable. Furthermore, the text of a play, originally conceived as a set of instructions for a proper performance, may be treated also as a literary work and may have multiple copies. Conversely, literary works such as poems may be recited and may have multiple executions.
Secondly, the different modes of realization vary in different dimensions. For instance, within the category of multi-realizable art, literary works may be subject to translation, which is never a simple reproduction. Within the category of multiexecutable art, performances may involve changes and adaptations; they may also be improperly realized executions that involve a relevant number of errors.
In Search of the Ontological Common Core of Artworks Works of mono-realizable art may be restored, and then undergo changes to their material constitution. But at what point do these changes cease to alter the nature of the artwork?
Lastly, technical innovations can modify the ontology of different art subclasses.
From this perspective, does digital photography share the same ontological nature as analogical photography? And if, for instance, future technical inventions allow us to perfectly reproduce mono-realizable art, will paintings and sculptures still be considered unique and concrete individuals? 4 Considering the complexity of the various modes of realization, since their early formulations, the analytical ontologies have rejected the so-called 'physical hypothesis' , the idea that artworks coincide with the physical object in which they are embodied. 5 But, in a large part of the analytical tradition, the discussion has been focused on the 'type/tokens' paradigm, and has produced conflicting solutions. 6 As Thomasson has argued, 7 the basic difficulty with this paradigm is that artworks do not match the traditional bifurcations between mind-dependent entities and mind-independent ones (mono-realizable art falls between these two categories) and between concrete individuals and abstract universals (multirealizable and multi-executable art falls between these two categories). I propose that it is possible to overcome this impasse concerning the nature of artworks only if we refer to the intentional level of description.

II. ARTWORK AS AN INTENTIONAL DEVICE
Generally speaking, artworks constitute representational devices that convey meanings. More precisely, they are physically embodied and culturally emergent devices. Now the crucial point here is to identify the specific and original mode in which artworks realize the embodiment/emergence relationship. From this point of view, pace Margolis, 8 this relationship is in principle very different from that one between people. True, people and artworks are materially embodied: both people and artworks are not reducible to, and therefore not identical with, their physical embodiments because they have intentional properties that cannot be explained exclusively in physical terms. But there is an essential difference between people and artworks. The former have what Searle calls 'original intentionality' , 9 that is, they are an autonomous source of intentionality, whereas the latter have only a kind of 'derived intentionality': artworks are culturally emergent only by virtue of the intentionality of authors and interpreters. From this point of view, artworks are much more similar to artefacts than to people. 10 Yet they are very special artefacts. Whereas the meanings and functions of tools are immediately manifest because they coincide with the evident practical usefulness of tools, conveyed meaning cannot be entirely extracted from artworks even by means of a full explanation. The meanings of artworks cannot be constrained once and for all in an overall generalization; they cannot be confined in a complete and definitive formulation. Consequently, my main point is that since artworks share a specific and original mode of embodiment/emergence, the derived intentionality is grounded in a underlying structure; this structure affords a multiplicity of potential cues, associations, and meanings; linked to the underlying structure, interpretation can never be completely and definitively solved. That is to say, artworks exist in the manner of non-universalizable embodied meaning.
This thesis echoes a widespread conception proposed in various forms in modern and contemporary aesthetics, from the beginning (for instance, the 'aesthetic ideas' of Kant) all the way to the latest developments (for instance, the 'embodied meaning' of Danto). 11 Moreover, many recent experimental data confirm that artworks, especially high-quality ones, are multi-interpretable. 12 They recursively stimulate multiple cycles of perception, conceptualization, and emotional reactions. They suggest a variety of potential meanings, and are open to multiple possibilities of understanding. They represent continuous source of insights without a determinate and univocal interpretation. They are unsolvable. My proposal especially aims at making explicit and analysing In Search of the Ontological Common Core of Artworks 9 John Searle, 'Intentionality and its Place in Nature' , Synthese 61 (1984): 3-16. See also his 'Minds, Brains, and Programs' , Behavioral and Brain Sciences 3 (1980): 417-57. 10 the non-universalizable nature of artworks, arguing that it represents the condition that makes the embodiment of meaning in artworks radical and irreducible to any other form of embodiment.
First and foremost, this condition involves the fact that artworks, conceived as means of representation, tend to draw attention to the medium. In ordinary, scientific, and philosophical communication the medium that conveys information allows direct access to the object that it denotes; attention is transitively directed to the object to which the representation refers and not to the representation that conveys the reference. But interpreters cannot simply look through artworks' underlying structure to access the object. They must pay close attention to the medium as such, because the underlying structure is not an interchangeable vehicle of reference, a mere substitute for the object. Rather, it is meaningful in itself: moving towards the intentional properties of the objects, attention is also intransitively focused on the means of representation as such.
In artworks the medium of representation cannot be replaced, reduced, or eliminated by a general formulation, because it is essential to enable an open process of multi-interpretability. We can easily understand this point if we consider the radical difference between the embodiment of newspaper articles and that of poetry. In the former, the reader typically tends to extract only factual information from the means of representation. In the latter, the interpreter is completely absorbed into a network of deep relations between meanings and means of representation.
From the phenomenal point of view, this perspective entails that artworks are objects of attention as such, in their underlying structure and physical realizations. Subjective interaction with an artwork constitutes a meaningful experience in which the interpreter is completely absorbed in and identified with the object in its singularity. Whether an artefact or a performance, the artwork guides and prescribes the work of imagination, continually renewing attention and counteracting habituation; it is attractive and produces pleasure in itself, and it is valued and appreciated in itself. In virtue of this particular condition, the interpretation can never be realized as an automatic procedure, based on generalized knowledge and enacted by rule-governed processes. On the contrary, anchored to the network of associations and meanings embodied in the artwork, the work of interpretation typically fails to culminate in a definitive solution. Each conceptualization of the embodied meanings is not fully adequate, but stimulates the imagination to further processing. The generalization hypotheses are constantly renewable and a completely satisfactory interpretation is not available.
That is, interpreters cannot definitively conclude their process of understanding.
This process remains constitutively open. Thus, the circle of interpretation can be indefinitely repeated in relation to the same artefact. It is potentially endless: each determined conceptualization is never ultimate and definitive, but always transitory in such a way that the interpretation can be closed only on the basis of pragmatic, and easily revocable, agreement. The circle of interpretation is laced with conflict: the embodied meanings support multiple interpretations, even very different ones, in continuous dialogue with each other.
Thus, artworks (that is, all artworks, whatever their value) differ from other artefacts having the nature of embodied meaning because the means of representation in which the meanings are embodied cannot be replaced, reduced, or eliminated, but they are objects of attention as such, in their physical and concrete realizations. According to my proposal, non-universalizable embodied meaning enables deep interdependence between physical properties and intentional properties. On the one hand, intentional properties are internal to physical ones. Intentional contents, in fact, are embodied in a physical underlying structure that does not represent an interchangeable means. Rather, it is an irreplaceable support relevant to the elaboration and interpretation of the appropriate meanings. On the other hand, physical properties are internal to intentional ones because intentional properties, making the physical ones meaningful, determine which level is relevant for the interpretation.
From this point of view, there is no fixed role for physical properties and the levels at which they are situated: it is only the intentional attribution that establishes, from time to time, their meaningfulness and salience. For instance, in the paintings of Rothko, meaning is situated at the level of colour; in those of Mondrian, at the level of the lines; and in those of Burri, at the level of the constituent material. The same consideration applies to Cage's musical compositions, in which meaning is situated at the level of noise. From this perspective, although it remains undoubtedly true that in general a painting is not just a canvas with colour pigments and a musical composition is not just a sequence of sound waves, the intentional properties of the artworks can also make significant these basic physical and perceptual levels. So, what constitutes for many modern artworks an irrelevant physical and perceptual element, in many contemporary artworks can represent an essential support for interpretation.
Moreover, the meaningfulness of the physical underlying structure is not necessarily tied to the presence of the traditional aesthetic properties, such as beauty, grace, and proportion. For instance, in Duchamp's readymades, the attribution of meaning is precisely tied to the perceptual banality and ordinariness of the object. Only because the perceptual configuration is devoid of traditional aesthetic qualities does it become relevant, triggering the work of In Search of the Ontological Common Core of Artworks reinterpretation that involves the concept of the object and the elaboration of its meaning.
Based on the deep interdependence between intentional properties and physical properties, my proposal clearly echoes the common claim in the literature of analytic aesthetics based on 'not rule-governed supervenience ' . 13 According to this common point of view, aesthetic properties are not rulegoverned because their relation to the non-aesthetic physical substrate on which they depend is not rule-governed either. But intentional properties are different from aesthetic ones. 14 In my proposal, intentional properties concern meaning attribution and involve cognitive operations such as perceptual analysis, implicit memory integration, explicit categorization, and complex mental processes (conceptual blending, analogy, and so on). 15 Aesthetic properties concern evaluation and involve different kinds of subjective and personal reactions, from immediate affective appraisal to complex and sophisticated emotions. Obviously, intentional and aesthetic properties deeply interact in interpretation and aesthetic judgement. 16 Nonetheless, they are different in principle.
Moreover, the precise nature of supervenience is highly controversial, disputed not only in analytic aesthetics, but in all the domains in which the notion is applied -consider, for instance, the ontological conceptions of supervenience in the philosophy of mind. 17 For these reasons, I do not directly apply this notion here. In order to understand the common core of artworks, I consider it more useful to move from the intentional description of non-universalizable embodied meaning to the social one. Before I treat the social level of description, however, it is important to discuss a possible objection to my proposal. In criticizing my proposed conceptual framework, one might argue that not all artworks reveal this condition of radical embodiment, or, at least, not all artworks realize it in the way I have described.
From the point of view of the norms of belief, assertion, and evaluation, our perspective clearly involves the so-called 'acquaintance principle': if subjects desire a genuine aesthetic experience, the artwork can never be completely aesthetic and artistic value has questioned the status of the acquaintance principle. A number of compelling counterexamples have shown that a strict adherence to the literal, strong version of the principle is untenable. 18 Moreover, some artworks, in particular some works of conceptual art, do not seem to require a first-hand perceptual experience. 19 At any rate, the opponents of the principle (the advocates of optimism about the legitimacy of forming aesthetic judgements purely on the basis of testimony) acknowledge that the acquaintance principle is widely accepted in aesthetics: the idea that proper aesthetic evaluations are governed by a principle in the spirit of acquaintance has something like the status of orthodoxy and it is usually conceived to be something akin to a truism. 20 Moreover, opponents generally do not deny that there is a crucial difference between the cognitive states of the person who directly perceives the artwork and the person who derives his judgement from a secondhand description. 21 On the other hand, the recent version of pessimism rejects the 'unavailability option' (testimony does not make knowledge available at all) and embraces the 'usability pessimism': testimony alone cannot be the sole basis of the aesthetic judgement, but it may have an important role as signalling information in arriving at our aesthetic judgements. 22 Above all, if we consider the normative force of the acquaintance principle to be the basic rule concerning the social domain of the aesthetic evaluation, it is important to stress that, in line with Thomasson,23 risks of indeterminacies and vagueness arise whenever the criteria for applying a rule is determined by collective human intentionality. The acquaintance principle is therefore inevitably subject to a number of irresolvable counterexamples within the current, unspecified practices and concepts.
In Search of the Ontological Common Core of Artworks From this perspective, if we do not stipulate new collective conventions from those currently available, it is impossible to derive a consensual and satisfactory formulation of the acquaintance principle. That is, it is impossible to specify -with universal agreement, and once and for all, for any circumstance of interpretation -what can be seen as a valid surrogate for an artwork, capable of replacing the first-person experience of the artwork. It is precisely in these gaps that many contemporary artists have focused their research (for instance, Rauschenberg in Erased de Kooning Drawing or De Maria in Vertical Earth Kilometer), explicitly aiming at avoiding and thus falsifying the acquaintance principle. Although we should ask (and I will later) to what extent many of these products are only parasitic and derived manifestations based on the interpreters' previous familiarity with the considerably larger set of artworks for which the application of the principle is so essential as to be an implicit and tacit assumption taken for granted.
On this basis, it is now time to consider the social level of description in order to clarify the relation between non-universalizable embodied meaning and collective intentionality.

III. ARTWORK AS A SOCIAL ENTITY
Art and its many sub-genres are products of shared agreements; the work of imagination required to interpret an artwork is not a solipsistic suggestion; rather, it is grounded in an external occurrence that contains collectively accessible prescriptions; the assignment of the artistic status to a candidate artefact always requires intersubjective recognition. In particular, the level of social description must explain how non-universalizable embodied meaning, intended as the ontological core of art, represents the basis of forming beliefs and practices in art-related collective intentionality.
From this point of view, I propose that non-universalizable embodied meaning functions precisely as a'constitutive rule' in Searle's sense, 24 that is, as rule that not only regulates a social activity/entity, but also produces it. According to Searle, unlike the natural laws concerning physical objects, constitutive rules institute the form of the activities that they rule by virtue of a collectively accepted assignment of mind-dependent functions, as in chess, baseball, money, marriage, and so on. Their classical form is 'x counts as y in the context c' . In Searle's naturalism, x is an object or a physical event, whilst y is any social entity. Obviously, the formula can be recursively repeated, so that the term x at a certain level can be the term y at a previous level. For instance, the president of a state is also a citizen of that state. If the social entity established by the constitutive rule is still constrained to a physical structure, it is based on the simple 'assignment of function' . If the resulting entity is freed from any physical structure, it constitutes an 'institutional' fact based on the 'assignment of status function' . An example of the first case can be a social boundary formed by a physical barrier; an example of the second case can be a boundary constituted by an imaginary line.
The concept of constitutive rules has been subjected to various objections, both general, related to the indiscriminate application of the concept to all the social entities, and particular, related to its application to artworks. Although the analysis of these objections confirms the need to integrate the constitutive rules with other principles of construction concerning the social and institutional reality, 25 it shows the full applicability of this notion to art, in particular its relevance in explaining the social nature of non-universalizable embodied meaning.
The most common objection concerns Searle's general thesis about the 'logical priority of brute facts' , according to which collective intentionality adds new minddependent functions, whether social or institutional, to pre-existing physical objects. From this perspective, the attribution of function can be reiterated, but, at some early point, the process begins from physical, mind-independent objects.
The main limitation of this thesis is that it is appropriate only for a limited subset of social entities; in particular it does not explain those entities in which the level y cannot be reduced to any physical object or event, as in the sociopolitical dimension. For entities such as universities, political parties, corporations, classes, laws, governments, nations, and economies there is no element x that satisfies the formula. Nor are there physical elements that can correspond to negative entities, such as debts. It follows that, although the constitutive rule may be appropriate for the social entities that are 'invisible' but in some way 'tangible' , the rule fails to explain the 'evanescent' social entities that cannot be reduced to specific physical substrates. 26 This general objection, however, does not concern the definition of non-universalizable embodied meaning as the constitutive rule of art. It is true that artworks are as invisible as other social entities because, as I have argued with regard to the level of physical description, they do not coincide with a material object and cannot be reduced to sensory information. But, unlike the 'y-independent' social entities, artworks are not at all evanescent. On Thus, the assumption that in general represents a limit to the concept of the constitutive rule provides a useful explanatory tool in art because it sheds light on the essential role played by the condition of radical embodiment of meaning in the physical structure.
The most serious objection concerning the relationship between the concept of the constitutive rule and art claims that such a rule is inevitably exposed to a conventionalist drift. On the one hand, it is generally acknowledged that the constitutive rule effectively explains some artworks, such as Duchamp's readymades, a real example of the attribution of artistic status function, in which a mere thing is then 'transfigured' into an artwork and normatively judged according to the principles of artistic evaluation. On the other hand, it is often argued that by virtue of the assignment of status function, any x can be transformed into an artwork. 27 Thus, in line with Dickie's (implausible) institutional theory, 28 in art 'anything goes' , that is, anything can become art by virtue of the assignment of status from an artworld.
Nor, however, does this particular objection prevent the functioning of nonuniversalizable embodied meaning as the constitutive rule of art. Theoretically, the acknowledgement of the relevance of social practices and institutions does not involve conventionalism. Indeed, even if collective stipulations take the form of a real transfiguration, they are not, in principle, arbitrary decisions that can transform anything into art at their whim. Whatever its mode of presentation, whether proposed by official institutions of an artworld or simply outsider art, an entity that is a candidate for artistic status is (or ought to be) judged as art by the community only if it realizes, at least to a minimal degree, the normative criterion of non-universalizable embodied meaning. In this sense, it is precisely nonuniversalizable embodied meaning conceived as a constitutive rule that constrains the collective intentionality, providing rational warrants for the assignment of artistic status such that it prevents the omnipotence of interpretation and stipulations. Obviously, communities may in fact err -or, better said, communities might be directed primarily by market forces and conformism. Similarly, many experts argue that this art is both parasitic (that is, it can be understood as art only in relation to the other art, that is, not institutional art) and self-referential (that is, it has completely lost its transitivity, its openness to the world). 31 Non-universalizable embodied meaning allows us to identify the difficulties related to this art. the 'test of time' and become classic artworks in the full sense of the term, that is, artworks always available to endless interpretative processes? Or, on the contrary, is their persistence limited to the mere testimony of contingent agreements, without the opportunity to feed the debate and support a rich story of effects?
Whatever the correct answers to these questions, at this point it is relevant to underscore that my ontological proposal identifies a precise function of art.
Artworks are designed to prompt, to prescribe, and to guide open imaginative games of make-believe, repeated cycles of perception and conceptualization, interpretative processes that are terminable only in virtue of pragmatic stipulations that do not deplete the available affordances for further elaboration, and corresponding experiences of endogenous pleasure, autotelic appreciation, and self-contained satisfaction -that is, self-reinforced and self-rewarding experiences that do not go beyond themselves towards an external and practical goal, but are supported by the processing dynamics determined by the nonuniversalizable embodied meaning. 32 As I have argued, whatever the procedures of presenting artworks (or whatever the material changes undergone in the process of restoration, or the technical innovation concerning art), non-universalizable embodied meaning is a requirement that candidate entities must satisfy, at least minimally, in order to be assigned the status of art. Thus, my ontological perspective completely reverses the assumption followed by many contemporary ontologies of art, and not only the analytical ones.
According to that assumption, the ontology of art has nothing to do with aesthetic experience in general or with aesthetic value in particular. 35 On the contrary, from my point of view, the aesthetic value is a key component of the ontology of art. Obviously, this absolutely does not mean that the ontology makes value judgements on single artworks. On the contrary, it means that the ontology of art acknowledges that non-universalizable embodied meaning, as the constitutive rule of art, is a value-dependent and value-sensitive principle.
On the basis of its intimate relationship with aesthetic value, the condition of non-universalizable embodied meaning allows us to consistently explain theoretically the distinction between high-quality and low-quality art. It is equally true of low-quality and high-quality artworks that intentional and aesthetic properties are grounded in physical ones. The difference between bad-making or good-making intentional and aesthetic features depends on the degree of realization of the crucial principle of non-universalization. Similarly, it allows us to consistently explain theoretically the formation of artistic canons. If highquality artworks, as more complete realizations of non-universalizable embodied meaning, are multi-interpretable, they are in principle also the most appropriate to stand the test of time and become classic artworks because they are by definition the most available for indefinite reinterpretation in different historical contexts.
It is important to emphasize, however, that non-universalizable embodied meaning is not a rigid rule that hypostatizes the heuristic and flexible distinction between great art and ordinary art, making it a metaphysical separation of kinds. consider, for instance, the mass distribution of high-quality film directors, like Hitchcock and Kubrick. Similarly, non-universalizable embodied meaning provides an essential hermeneutical tool for coping with and trying to limit the market forces that have, for example, since the famous chimpanzee Congo that was instructed by Morris in the 1950s, created an entire submarket dedicated to animal art, with 'artworks' made by dogs, elephants, dolphins, and horses. 37

IV. DEFINITION AND IDENTIFICATION: FROM ONTOLOGY TO COMMON SENSE
Ontological questions are usually considered to be related to definitional ones but not identical with them. The former concern the manner in which artworks exist, whereas the latter concern what they are. But the proposals of nonuniversalizable embodied meaning as the ontological common core of artworks allow me to directly consider the definitional questions. Firstly, it is possible to derive from non-universalizable embodied meaning a strong definition that identifies, from the theoretical point of view, the necessary and sufficient conditions for an artefact to be judged an artwork. The ontology outlines a precise taxonomy. In Aristotelian terms, the 'genus' of art is the embodied meaning.
The 'specific difference' is the non-universalization. The 'kind' of art can then be further divided into several subclasses, both according to the principles of artistic genres 38 and to the value criterion concerning the degree of realization of the principle of non-universalization.
But, in order to complete the description of non-universalizable embodied meaning as the constitutive rule of art, it is indispensable to specify how such a rule, as I have described it from the ontological point of view, actually explains the common sense operations related to art. In fact, it is important to keep in mind that common sense is typically characterized by the lack of sophisticated and detailed concepts. Moreover, compared to ontology, common sense relates to art with a different purpose. Whereas ontology as a rational reconstruction aims at defining the unity and the specificity of art in theoretical and abstract terms, common sense aims at identifying and experiencing the single artworks. In order to address this question properly, it is important to stress that, as conceived here, the contrast between ontology and common sense is not coextensive with the distinction between art experts and non-experts. Many experimental data show deep differences in processing between experts and non-experts. 40 Whereas non-experts tend to classify on the basis of what is depicted, that is, content, experts tend to classify on the basis of prototypes of artistic schools and artists. Thus, non-experts refer mainly to the intrinsic content of the work, elaborate in an intuitive way, by virtue of association with their immediate daily experience. By contrast, experts process perceptual, formal, and structural properties together with content on the basis of style, correlated artworks, the context that gives meaning to the artwork, and art history. These processing dissimilarities result in an essential difference concerning the implementation of non-universalizable embodied meaning. Whereas experts experience genuine multi-interpretability, actually situated in the history of effects, non-experts experience it as an uncertainty difficult to understand on the basis of their ordinary personal experience. It goes without saying, however, that an art expert is not necessarily interested and/or competent in the ontology of art.
Moreover, it is essential to clearly point out that identification and definition are absolutely not the same process. 41 There can be identification without explicit or implicit reference to a comprehensive and exhaustive definition. From this point of view, art has nothing specific about it. In recent decades, a growing body of evidence has shown that a large number of human concepts are applied without relying on a definition. 42 Rather, recognition and identification take place on the basis of a widely shared, internally stratified framework of tacit knowledge.
The elements that constitute this tacit knowledge are various. Firstly, there is a basic and rudimentary ontological knowledge, embodied in the background practices concerning art. The conditions of artistic artefacts' identity, individuation, existence, and survival are determined in line with this ontological knowledge.
To give an example provided by Thomasson,43 whether we refer to an abstract or to a concrete individual depends on whether we consider it appropriate to go to a particular place at a particular time to see the artwork. Secondly, according to Danto,44 this ontology is enriched by familiarity with the artworld, conceived as an atmosphere of artistic theories, traditions, and art history. This knowledge, not necessarily considered in the sense of professional expertise, allows interpreters to recognize the difference between things and artworks. Thirdly, the previous experience with the institutional context, the procedures for presenting artworks, and the reference to the authors, their collectively accepted status, their intentions expressed in artworks' titles, and so on, play an important role in identification and recognition.
Within this framework, I propose that familiarity with great art plays a particularly relevant role. As it has been identified by 'grounded cognition' , in a large part of categorization processes, knowledge is organized around 'exemplars' , that is, individual occurrences of a specific category with which subjects have had direct contact in their previous personal experience. Subjects immediately access these exemplars by quickly recalling them as the best examples of the category, that is, the most typical. For this reason exemplars are usually used as mental reference points to establish whether an entity represents a member of a class, a genre, or a category. For instance, the concept 'car' is often categorized by recalling a particular model of car. 45 It is not difficult to assume by analogy that the same is also true for common sense knowledge related to art. 46 From this perspective, artworks that are commonly judged to be exemplars constitute the standard of comparison for the evaluation of the other entities that are candidates for artistic status. The evaluation is carried out on the basis of the relationships of dis/similarity among exemplars and candidates.
Tacit knowledge and especially the relations of dis/similarity centred on exemplars provide both a heuristic function and a normative one. In the former sense, they enable interpreters to see very different entities (such as a building, a poem, or a play) as part of the same kind -namely, art. In the normative sense, they enable interpreters to exclude an entity from the 'art' kind even if surface similarities with artworks exist (for instance, a photograph of the same object depicted in a painting). But, it should be emphasized that the process by which sense is not necessarily systematic, even among experts. As a result, the selection of exemplars may be more or less subjective because it is based on personal experience, sensitivity, expertise, and skills. Similarly, exemplars' properties judged as relevant to attribute artistic status may be selected in an incomplete, imprecise, and approximate mode, especially by non-experts.
Contrary to popular conception, 47 however, it is essential to note that these selective processes are not purely casual, neither in principle nor in fact. Even in this case, non-universalizable embodied meaning represents the mechanism that warrants a rational selection of dis/similarity relations. Obviously, in the tacit Non-universalizable embodied meaning is therefore not an odd, counterintuitive proposal, distant from the ordinary art-related practices.
While, from the ontological point of view, non-universalizable embodied meaning represents the determining principle of art as such, in daily practices it is decomposed into a set of mental sub-properties and in this form makes an essential contribution to warranting, at least to some degree, the rationality and adequacy of the identification of artworks.

V. THE CREATIVITY AND HISTORICITY OF ART
It is now possible to argue convincingly that non-universalizable embodied meaning is compatible with artistic creativity, the historical nature of art, 48 and the current complex state of artistic phenomena.
In Search of the Ontological Common Core of Artworks It is well known that the sceptical perspective denies the possibility of defining art, arguing that it represents a concept with an open structure -a concept that can indefinitely be adapted to new works. 49 On this view, non-universalizable embodied meaning is, by definition, a condition that in no way constrains the creativity and production of new artworks, even extremely original ones with a relevant amount of novelty and violations of expectations. Its realization, in fact, is not limited to fixed and predetermined modes. On the contrary, it is, in principle, an open condition that refuses to be crystallized into a standard formula. From this perspective, in order to be fully satisfied, the normative criterion of non-universalizable embodied meaning requires continuous experimentation.
Without this, candidate artworks will evoke only habituation and boredom.
So, if we agree with Danto that after Andy Warhol's Brillo Boxes art becomes free to be whatever anyone wants it to be, 50 non-universalizable embodied meaning absolutely does not limit this freedom. On the contrary, it represents the frame of reference that allows artistic freedom to be meaningful, innovative and original works to be recognized as artworks, interpreters' frustrated expectations to be restored in new shared stipulations, and novel canons to be formed and diffused. That is, non-universalizable embodied meaning functions as the framework that allows experts to bring in the historical narratives that explain how new artworks are derived from previous ones.
In particular, in the current multifaceted state of artistic production, nonuniversalizable embodied meaning allows us to distinguish different meanings of the word 'art' , especially a prototypical meaning and a peripheral one.
The former refers to artworks that realize non-universalizable embodied meaning in a complete form and can, as such, be subject to a strong definition. The latter refers to peripheral instances of art in which the principle of non-universalization is implemented only in part and disjunctively. For instance, in ordinary art, surprise, produced by the frustration of expectations, can initially often attract attention, but then quickly decreases because interpreters can easily assimilate the violations, and the artwork consequently ceases to be interesting. Examples of this kind of processing include less successful works of conceptual art. 51 It is also possible that the intense experience prompted by the artwork is only a simple emotional reaction, not involving any cognitive elaboration. Examples of this kind of processing include movies produced by the film industry for mass This second subset of artworks can be delineated by a weak definition. From a disjunctive perspective, in this kind of definition non-universalizable embodied meaning is decomposed in a list of properties. As I have argued, these properties are related especially to the mental processes typically activated in the previous experience with high-quality artworks, that is, artworks that fully realize nonuniversalizable embodied meaning. This set of properties is sometimes applied in different combinations, in relation to the kind of processing required by the single artwork. Thus, from this point of view, the reference to the mental processes determined by non-universalizable embodied meaning is essential not only for common sense to identify artworks and make intuitive and non-analytical aesthetic judgements, but also for the ontology of art in order to find appropriate definitions concerning the different kinds and levels of artworks.
With the idea that non-universalizable embodied meaning can be broken down into a set of criteria applicable in different disjunctions, I am following the lead of the 'cluster theories' , in which art is a 'cluster concept' involving a plurality of criteria and disjunctions. However, both in their essentialist version (in which proponents see the disjunctive explanation as a real definition) 53 and in the antiessentialist version (in which proponents see the disjunctive explanation as a description), 54 the main limitation of these theories is that in selecting the criteria by which art is defined/described, they produce lists of properties in which connections are left unspecified because they lack a common reference point. 55 In my proposal, by contrast, that common reference point is represented by nonuniversalizable embodied meaning.
In any case, it is important to emphasize that neither the strong definition nor the weak disjunctive one are algorithms that would make possible a deterministic matching, thus involving a metaphysical distinction between exemplar art, peripheral art, high-quality art, low-quality art, and non-art. On the contrary, both definitions are vague: not because they are somehow confused, but because their application is devoid of sharp and fixed boundaries. 56 If this consideration applies to the first type of definition, because, as I have argued, the detection of non-universalizable embodied meaning is always subject to interpretation, it is even more appropriate for the second one. Outside exemplar art, there is an indeterminate set of hybrid, mixed-media, and borderline cases that, even if they present a number of relevant criteria, make it difficult to define whether they are art, even from the point of view of ontology and rational reconstruction. This is what happens, for instance, in design, one of the most representative examples of so-called 'diffuse aesthetics' . 57 The question is as follows: these kind of tokens realized outside the domain of exemplar art might be (and often are) more creative, innovative, and valuable than many mediocre artworks -but could we still consider them art (in the prototypical sense I have described)?
From this perspective, the first type of definition, if its application is justified, a conceptual, indispensable tool for understanding both the various artistic products and the pseudo-artistic ones, such as the phenomenon concerning so-called 'everyday aesthetics ' . 59 A final consideration to complete my proposal: in recent years there has been an important debate on the adaptive value of aesthetic experience. 60 Regarding this, the central point to emphasize here is that even the few people who deny this value acknowledge that aesthetic experience is a universal feature of the human species and that it is intrinsically characterized by the constant drive to innovate. 61 The available findings show that, at least since the Upper Paleolithic, so-called 'trans-historical' and 'trans-cultural' art, that is, the art realized first in archaic and then in pre-modern societies, emerged as a deliberate technical extension of aesthetic experience and its internal drive to innovate. 62 It is undoubtedly true that the style of cave art remained static over millennia and that the traditional art of present-day hunter-gatherers tends to be deeply conservative. 63 Nonetheless, the absence of stylistic innovation did not prevent our ancestors from developing mental flexibility through the use of artistic products and activities, devoted to re-enact schemata previously entrenched in memory and to work creatively on them in a way that defies ordinary and routinized processing. 64 Even if this attitude to innovate is not universal in art history (consider, for instance, the periods in which the goal was to mimic previous masters and preserve their iconography), it has developed its modern form within the eighteenth-century system of fine arts -a system that gradually underwent profound changes, which I will not go into here. 65 In relation to modern art, great art, as the exemplar manifestation of non-universalizable embodied meaning, must, in principle, be an extremely creative and original work of imagination.
This tendency to innovate, typical of art, has reached its most radical acceleration in contemporary art. The trend that judges innovation as a value in In Search of the Ontological Common Core of Artworks 59 itself has institutionalized innovation as such. Moreover, market pressures force artists to develop a highly distinctive, recognizable style. Thus, at least since the avant-gardes in the early twentieth century, there has been a proliferation of codes and styles so wide and so fast that none of these has been sufficiently stabilized and sufficiently diffused to provide a common language or a common canon.
In the theoretical framework I have articulated, this situation has profound consequences for the status of art -at least as we have known it so far. As I have argued, art is a social entity, established by the stipulations of collective intentionality. The primary condition for the objectivity of this type of entity is the collective assignment of functions based on continuous recognition and constant acceptance by a sufficiently large community over long periods. From my perspective, the unity and the specificity of art are typically given by the stability and sedimentation of the constitutive rule of non-universalizable embodied meaning. Thus, the crucial point is as follows: if the contemporary state of art does not have a shared code and is largely characterized by hybrids aimed at removing not only the difference between the various artistic sub-genres, but also the distinction between art and non-art, do exemplar artworks still typically and continuously show this constitutive rule? Is exemplar art still made today?
Whatever the most persuasive answer to these questions, one must acknowledge that, as a social entity, art (in a general historical sense and in its particular exemplar instances) is devoid of those biological warrants that instead make aesthetic experience a cross-cultural universal of the sapiens brain. From this perspective, it is possible that historical art is currently constituting (or has already constituted) an ontological kind whose unity is exclusively situated in the past, because its consensual centre, the class of art exemplars, ceases (or has already ceased) to be actually produced. Accordingly, it is inherent in the ontology of art as an historical ontology that it might constitute (or already has constituted) an ontology without art, without the exemplar instances of art being produced, and therefore a mere rational reconstruction of a past heritage that still survives in the common sense only as a residue of past art canons. Similarly, the strong definition might be reduced (if it has not already been reduced) to a simple conceptual simulacrum without any present matching object, because only peripheral practices are experimented with far from the traditional centre of art.
How experts respond to these questions largely depends on where they decide to look for in the current state of art and how they interpret art history. 66