Why Joseph Margolis Has Never Been an Analytic Philosopher of Art

In this paper, we support a continuistic reading of Joseph Margolis’ philosophy, defending the claim that in the 1970s, Margolis tackled the issues suggested by the analytic philosophy of art from an original theoretical perspective and through conceptual tools exceeding the analytical framework. Later that perspective turned out to be a radically pragmatist one, in which explicitly tolerant realistic claims and non-reductive naturalism converged with radical historicism and contextualism. We will endorse this thesis by focusing on two important concepts appearing in Margolis’ aesthetics essays from the late 1950s to the 1970s: the type-token pair and the notion of cultural emergence. On the one hand, we will emphasise Margolis’ indebtedness to Peirce’s first formulation of the type-token distinction, involving a strong interdependence between the two elements of the pair, as well as an anti-essentialistic, historicised, and contextualised notion of type. On the other hand, we will delve into Margolis’ exploration of the concept of emergence and cultural emergence, involving a genuinely pluralistic view of ontology, as well as a non-reductive, continuistic form of naturalism. Finally, we will connect the criticism of the so-called closure of the physical world with Margolis’ anti-autonomistic stance in defining artworks.


JoLMA
e-ISSN 2723-9640 3, 2, 2022, 333-364 1 Introduction: A Continuistic Narrative When it comes to reconstructing the history of the so-called analytic philosophy of art, Joseph Margolis (Newark 1924-Philadelphia 2021) is almost invariably included in the list of authors to whom one must refer.Indeed, it is indisputable that the debate that took place within that specific philosophical current was shaped by a number of essays written by Margolis between the late 1950s and the 1970s.His article "The Identity of a Work of Art" (1959), published in Mind, together with "Works of Art as Physically Embodied and Culturally Emergent Entities" (1974) and "The Ontological Peculiarity of Works of Art" (1977), published in the Journal of Aesthetics and Art Criticism, proved deeply influential and provided a crucial contribution to the discussion on the definition and ontology of art.The philosopher's thinking appears to be analytical both in terms of the theoretical tools he employs -the analysis of ordinary language as well as the use of logical or semiotic categories -and in terms of the topics he deals with -the definition of the concept of art, the ontology of artifacts, and the role of intentionality in the constitution of the meaning of an artwork.More generally, his approach to the subject matter also appears analytical: in opposition to classical aesthetic theories, Margolis, like many of his analytic colleagues, believes that a good philosophy of art should be assigned a descriptive rather than evaluative task.
The purpose of this article is to demonstrate that the narrative just proposed, while broadly true, is both simplistic and misleading.Indeed, it does not help us to grasp the fact that Margolis tackled the issues suggested by the analytic philosophy of art from an original theoretical perspective and through conceptual tools exceeding the analytical framework.Later that perspective turned out to be a radically pragmatist one, in which explicitly tolerant realistic claims and non-reductive naturalism converged with radical historicism and contextualism.The pragmatist tradition to which he had been exposed in the early years of his academic life (Margolis 2014) was combined with the influence of Marjorie Grene's philosophy of biology, the so-called later Wittgenstein, and a progressive reading of Hegel, giving rise to an approach to the arts that, in our view, challenged the implicit autonomistic claims of mainstream analytical aesthetics from the very beginning -in a few words, the idea that an answer to the question of what art is, and what kind of entity an artwork is, can and must be found within the artworld and/or art, assumed to be a self-standing institution.Moreover, we will argue that Joseph Margolis embraced a form of radical historicism in his view of the arts and did away with any residue of Platonism that was still present in the analytical philosophy of art.We will defend these claims by focusing on some key conceptual tools he put in his tool-JoLMA e-ISSN 2723-9640 3, 2, 2022, 333-364 will return to Margolis' characterisation of artworks as culturally emergent entities by illustrating the consequences of his use of the concept.More specifically, we will suggest that it involves the claim that the artworld cannot be considered a closed system, standing on its own theories; rather, the concept of emergence provides a decisive contribution by assuming artworks to be integral parts of the human world, i.e. to be related -through complex, multi-directional connections, including causal ones included -to other real components of the human world.
If ur two critical accounts of Joseph Margolis' philosophy prove accurate, we believe we can achieve a twofold result.As the title of the essay promises, we first of all aim to demonstrate that the philosopher's whole intellectual journey should be read in the context of a broader pragmatist project: a project that clearly distinguishes him from the analytical tadition with which he has often been associated.Although we are aware that there are internal lines of development in the author's thought, we wish to present a unified interpretation of it.He considered the arts in strict connection with the peculiarities of the human world from the very beginning -we suggest -and this approach crucially shaped the answers he offered in the debate on the definition and the ontology of art.The later development of his philosophy of art as involving an anthropology of culture is, in our view, a coherent development of the claims he initially formulated as responses to the issues raised within the analytical debate on art.Secondly, we wish to describe the figure and thought of Margolis as a bridge stretched between two philosophical currents that engaged in little dialogue until recent years: Pragmatism and analytic philosophy. 2 We believe that this is one of the important contributions that make Margolis a thinker for the 21st century -although he was not properly searching for a dialogue but rather for good arguments and tools to explore his original philosophical interests.

2
Joseph Margolis: An Analytic Ontology of Artworks?
In order to understand the key role that Joseph Margolis' philosophy has played in the analytical debate on art, it is first necessary to reconstruct the context within which it originated and became meaningful.
At the time when the philosopher began writing his first essays on the topic, academic debate seemed to have reached something of an impasse.In the 1950s, post-Wittgensteinian thinkers such as Paul Ziff (1953) and Morrs Weitz (1956) had denied that the concept of art JoLMA e-ISSN 2723-9640 3, 2, 2022, 333-364 could be defined in the traditional sense, that is, through the identification of necessary and sufficient conditions.Rather, to fall under the notion of "art" were disparate objects or phenomena, linked together not by the same defining characteristics but by an uneven web of resemblances.For these theorists human creativity, expressed in ever-changing forms throughout history, represented the very reason why a systematic treatment of the concept of art was to be deemed impossible.Art and its definition thus remained excluded from systematic philosophical discourse.
Acknowledging the post-Wittgenstenian thinkers' arguments thus meant addressing the question of whether there was a strategy for keeping art within the perimeter of ontological inquiry.In many reflections subsequent to Weitz's, the strategy that was identified consisted in avoiding the problem of defining art as a general notion by shifting the focus to the ontological nature of the artifacts identified by that notion.Recognising the impossibility or difficulty of answering the question "what is art?", one turned toward another question that in some ways was preparatory or at least alternative to it: "what kind of entities are works of art?"3In short, asking this question led to the query, "are there at least necessary conditions for a certain object to be considered a work of art?"And further, "are all works of art ascribable to the same ontological category?"And more in detail, "are works of art equivalent to the physical objects with which they seem to be identified?"-and so on.
It was within this theoretical framework that Joseph Margolis began developing his own philosophy.In an essay entitled "The Identity of a Work of Art" (1959) and then in many subsequent papers, the philosopher argued, first and foremost, that every artifact is the token of a type, a concrete occurrence embodying an abstract entity.
The type-token hypothesis was predominant within the analytic philosophy of art from the 1960s to the beginning of the third millennium.Evidence of this is provided not only by the large number of authors who employed the two categories as explanatory tools (Stevenson 1957(Stevenson , 1958;;Meager 1958;Margolis 1958Margolis , 1959;;Khatchadourian 1960;Wollheim 1968;Dipert 1993;D. Davies 2004), but also by the wide time gap separating early attempts at refutation (Bachrach 1971) from more recent ones (Rohrbaugh 2003).Moreover, sections devoted to the topic appear in all the major analytic-oriented aesthetics handbooks published over the last two decades (Levinson 2003;Kivy 2004;D'Angelo 2008;Livingston 2021).2723-9640  3, 2, 2022, 333-364 Although the important contribution that Joseph Margolis has made to the analytical debate about the ontology of artworks is undeniable, the purpose of the next pages is to show that the philosopher's thinking does not perfectly fit with the line of development of analytic philosophy.We will therefore proceed as follows.First we will describe how and for what purpose the categories of type and token were introduced in philosophy through the semiotic theory of Charles S. Peirce.Then we will illustrate the original way in which Margolis draws upon that theoretical context.The hypothesis we will put forward is that, from his early writings on art onward, the philosopher formulates a version of the type-token theory which is not only compatible with but also indebted to Peirce's pragmatist semiotics.Following the thread of our argument, we thus hope to highlight some line of continuity between the early Margolis, who is considered analytic, and the later constructivist one.In doing so, we will also show that there is a significant difference between Margolis' version of the type-token theory and the one adopted by most of his analytic colleagues.

2.1
The Type and Token Categories in the Semiotics of Charles S. Pierce It was in 1906 that the type and token categories were first introduced into the philosophical vocabulary, when Charles S. Peirce published "Prolegomena to an Apology for Pragmaticism".In this essay, which is a brief compendium of the philosopher's semiotic theories, the two categories are used to describe the nature of signs.In a sense, the use of this pair underlies the nine other ways in which a sign can be classified according to Peirce.These further modes describe the different kinds of signs not in and of themselves, but always on the basis of their relation to their reference, to the denoted object. 4 The meaning of each sign is defined by the link between a certain type and a corresponding class of tokens.To understand what this relationship consists of, it is good to start with the class of linguistic signs, which Peirce himself seems to regard as a paradigmatic field of investigation.It is precisely from language that the philosopher draws the example through which he begins to articulate his argument: 4 In Peirce's own terminology, denoted objects are also called dynamical objects and are defined as "the Reality which by some means contrives to determine the Sign to its Representation" (Peirce 1906, 505)  A common mode of estimating the amount of matter in a MS. or printed book is to count the number of words.There will ordinarily be about twenty thes on a page, and of course they count as twenty words.In another sense of the word "word", however, there is but one word "the" in the English language; and it is impossible that this word should lie visibly on a page or be heard in any voice, for the reason that it is not a Single thing or Single event.It does not exist; it only determines things that do exist.Such a definitely significant Form, I propose to term a Type.A Single event which happens once and whose identity is limited to that one happening or a Single object or thing which is in some single place at any one instant of time, such event or thing being significant only as occurring just when and where it does, such as this or that word on a single line of a single page of a single copy of a book, I will venture to call a Token.(Peirce 1906, 505-6) As the example chosen by Peirce perfectly illustrates, there are two distinct yet related ways of considering each word, each linguistic sign.On the one hand, each meaningful expression can be understood in its uniqueness, as something that presents itself concretely to our senses without being repeatable on any other occasion: an inflection or volume of voice, a particular handwriting, a regional accent etc.On the other hand, the same word can be thought of as an abstract entity embodied in concrete occurrences that resemble each other yet are not identical.In this sense, therefore, each sign possesses a twofold nature: in the former case it will be called a token, in the latter case a type.
To say that every sign has a twofold nature is not simply to argue that each of them can be interpreted either as a concrete occurrence or an abstract entity.The twofold nature of signs theorised by Peirce resides, in a far more essential sense, in the co-dependent relationship that exists between a certain type and the class of tokens corresponding to it.As an abstract entity, a type will only exist when it is embodied by some physical form, perceptible by the senses; conversely, an occurrence will acquire a determinate meaning only in its relation to a uniquely identified type.This relation of co-dependence is of the utmost importance, since it has to do with the principles of economy and recursion that govern language as such: precisely because signs exist in the twofold guise of types and tokens, it is possible to express an infinite number of meanings using a limited range of words.
In Peirce's theoretical framework, describing the relationship between type and token as co-dependent is also crucial for another reason.Indeed, co-dependence establishes an ontological asymmetry between the two terms.Tokens all possess a physical nature and thus exist in their own right, although they are formally distinct from the JoLMA e-ISSN 2723-9640 3, 2, 2022, 333-364 objects they denote (dynamical objects); types, with which the occurrences are associated, are instead abstractions that are not part of the sensible world.
The abstract nature of types, however, does not lead Peirce to develop either a Platonic or a mentalist conception of these two categories.Tokens and type are neither identified with mental states nor with universal kinds, but are rather taken to denote rules based on habitual associations.In this sense, they are as real as the occurrences that embody them.
Only through the mediation of habitual association do types establish a relationship with their corresponding tokens and gain ontological weight.Such associations must be thought of, in this context, as a preexisting and acquired background that enables and informs the interpretation of each new sign.They will involve events of a different nature depending on the kind of sign subjected to interpretation.In the case of what Peirce calls indexes, for example, the habitual association is to be understood as natural, i.e. as determined by the qualitative properties of a certain object: the regular presence of smoke caused by the lighting of fires will make the former a sign of the latter.In the case of linguistic signs, the understanding of a word or phrase will be made possible by a set of social or cultural habits.As we are about to see, precisely this communitarian aspect of interpretation will also be central to Joseph Margolis' philosophy of art.
In addition to habit, which should nonetheless be understood as a kind of interpretive framework, the relationship between type and token is thus mediated by a third element that, in Peirce's semiotics, is equally constitutive of signs.This is the tone, which the philosopher defines as "an indeterminate signifying character" (Peirce 1906, 506).One example of it might be the vocal colouring that is given to an utterance while it is being delivered. 5 The presence of this third aspect of the sign makes the identification of a token with its corresponding type extremely complex.Consider the case of irony: when it is used in speech, it is intended to give a certain utterance a meaning opposite to that which the same sentence would have in a normal context.Although identical, the two enunciations cannot be recognised as occurrences of the same type.
It is interesting to note right away that tones and their contextual status have been expunged from all philosophies of art that, inspired by Peirce's semiotics, include the categories of type and token in their toolbox.This fact is rather surprising because, in the very years in which these philosophies were being developed, tones acquired great importance in the field of the pragmatics of language: 5 For further exploration of Peirce's semiotic theory from the perspective of the type-token-tone triad, see Hilpinen 2012.
Finally, before checking how compatible Peirce's theoretical framework is with Joseph Margolis' philosophy of art, two clarifications are in order.First, the extent of the concept of linguistic sign must be considered.Although Peirce constructs the example by which he introduces type and token by using words as units, the two categories can be applied as theoretical tools far beyond this limit.After all, sentences, paragraphs, and whole texts or speeches are signs in their own righ; and they are such not merely as sums of other signs, but also as vehicles of a certain overall meaning.For this reason, all semantic units -independently of their extent and complexity -share the same ontological nature: they are abstract types embodied in concrete occurrences.It is precisely this observation that will allow the categories of type and token to be employed in the ontology of literary works.
Second, it is necessary to bear in mind that Peirce's semiotics is not limited to the narrow field of language.What has been argued about linguistic signs turns out to be true, according to the philosopher, in the case of all other kinds of signs as well.Amatriciana pasta, for instance, can be considered both that specific dish composed of bucatini, guanciale, tomato and pecorino cheese that I now find on my plate (token) and the recipe that establishes the cooking of those ingredients (type).Again, the relationship between the two is characterised by co-dependence: while the appearance and taste of the dish will be determined by a set of abstract rules, those rules will have to materialise in at least one physical occurrence for the amatriciana pasta to exist in full.

Joseph Margolis' Ontology of Artifacts: Intentionality
Having outlined the theoretical framework from which Joseph Margolis borrows the categories of type and token, our goal in this section will be to verify to what extent the philosopher's original position is not only compatible with but also indebted to Peirce's semiotics.Our aim will be to show that, even in essays that have become canonical in the analytic philosophy of art, Margolis adopts a perspective strongly influenced by Classical Pragmatism.If our hypothesis proves true, it would therefore be necessary to interpret his theoretical trajectory in a markedly continuist sense.
A good way to introduce Margolis' conception of the categories of type and token is to examine them in relation to the rival conception in opposition to which it was first formulated.The argument advanced in the essay in which the philosopher started to use the pair, "The Identity of a Work of Art" (Margolis 1959), begins with a JoLMA e-ISSN 2723-9640  3, 2, 2022, 333-364 critique of the theories of an analytic thinker: Charles Stevenson.The latter was the first to become aware of the possibility of applying the two categories to a specific area of aesthetics: the philosophy of literature.
In "On 'What Is a Poem?'" Stevenson (1957) proposed extending the principles of Peirce's semiotics to textual units so as to explain how the identity of literary works is individuated.Each poem, Stevenson noted, consists of a certain sequence of words that is physically manifested in a plurality of sensible expressions.Leopardi's L'Infinito, for instance, can be appreciated both on the page of a book and through an actor's performance, regardless of the substantial differences between one form of expression and the other.From a semiotic point of view, then, the reference to poetry lends itself to ambiguities.
Starting from this observation and following Peirce's footsteps, Stevenson concluded that what emerges in the analysis of literary works of art could only be resolved and explained in the following way: the identity of any textual artifact is identified by the relationship between a type -equivalent to the norm by which an order of succession is attributed to a set of meanings -and a class of tokens -which make that type intelligible to the senses.Thus it is again thanks to the twofold nature of signs that one can refer to the same poem in the sense of both a specific physical occurrence and the abstract entity that identifies it.Conversely, to deny that literary works are individuated by types embodied into tokens leads to an absurd conclusion: lacking the principle that traces each repetition back to the same abstract entity, one would be forced to consider each occurrence of the same verses an entirely new artifact.This would not only crowd the world with literary works that are all identical, but would contradict the way these are experienced in all societies around the world.
Stevenson has a semantic conception of the type-token relationship.Indeed, the identity between one term of the pair and the other is based on the sharing of a sequence of meanings.It is because of its purely semantic-textual nature that the philosopher's theory remains confined to the narrow sphere of literature without extending to the realm of the other arts.
It is precisely against this conception that Joseph Margolis will begin to articulate his own version of the type-token theory and, more generally, his own philosophy of art.The fallacy which Margolis (1959, 39) imputes to Stevenson is failing to recognise the peculiar trait that distinguishes a work of art from other forms of linguistic expression.Appealing only to the order of meanings does not allow us to establish any difference between a genuine poem and a simple series of words randomly lined up by a gust of wind.On the contrary, recognising a cultural expression as such always requires that a certain aesthetic design be imputed to it.In short, what Ste- 2723-9640  3, 2, 2022, 333-364 venson fails to grasp in his own theory is the intentional quality of all artistic phenomena.
Margolis also return to the topic of intentionality in a later essay, "The Ontological Peculiarity of Works of Art" (1977), where he once again applies the categories of type and token to the ontology of artifacts.In his article, the philosopher imagines an artist trying to attribute the property of "being a work of art" to all the objects randomly brought by the sea to the shoreline of a beach: although the imaginary artist seeks a ploy to create non-intentional artifacts, his unavoidable activity of selection will only make the purpose fruitless.The conclusion of the thought experiment thus demonstrates, by absurdity, that intentionality is a necessary condition for a work of art to be recognised as such: this is true even in the case of ready-mades, where intentionality is expressed to the lowest imaginable degree. 6 The realisation that artistic expressions are not only signifying forms but also entities oriented by an aesthetic design is of fundamental importance in the framework of Margolis' philosophy.This allows the philosopher to overcome both the assumption proposed by Morris Weitz that art is an unconditioned phenomenon and the semiotic perspective adopted by Charles Stevenson.By considering artworks not in terms of signifying forms but as objects invested with an aesthetic design Margolis is able, on the one hand, to give his theory an actual ontological status and, on the other, to extend the use of the categories of type and token to the philosophy of all artistic genres, not only literature.Even the identity of a piece of music, for example, can now be identified by the relationship between an abstract entity (type) and its physical occurrences (token).This is certainly a first point of contact with Peirce's semiotics.Indeed, the concept of sign on which it is based has far more extensive boundaries than mere linguistic signs.
The notion of aesthetic design employed in Margolis' early essays is itself very vague.It can be applied as much to works of art as to any other form of cultural production: industrial objects, advertisements, amateur or folk artifacts, and so on.This is therefore not a philosophy of art in the strict sense, but rather a philosophy of culture understood in the broadest possible terms.It does not provide conditions within which to circumscribe the totality of artistic phe- 6The examples mentioned so far concern the voluntary acts of a single subject.However, it would be wrong to assume that the concept of intentionality is limited to this in Margolis' thought.As we shall see later on in this section and in section 3.1, the philosopher identifies the notion of Intentional -written with a capital letter -with that of cultural.This broad conception of intentionality is clearly stated in the essays of his maturity, but it also seems to be present, albeit embryonically, in his writings on art from the late 1950s.2723-9640  3, 2, 2022, 333-364 nomena, but rather offers a theoretical framework for understanding the material products of the human mind from an ontological point of view.The difficulty in isolating art from other cultural forms should not be regarded here as a weakness of the theory, but as one of its deliberate features: it reflects the way in which Margolis believes human culture operates.

JoLMA e-ISSN
According to the philosopher, the distinction between disciplines or forms of expression does not reside in any specific property but has to do exclusively with the collective activities within a social community.This assumption becomes crystal clear and fully developed in the essays that follow Margolis' so-called analytic phase.Indeed, in these writings he adopts a notion of Intentionality -written with a capital 'I' to distinguish it from the subjectivist, transcendental conception derived from Brentano's and Husserl's phenomenology-that is dependent on the activities and relations that occur in participating in the same form of life.In a paper from 2000 he writes: Artworks characteristically possess representational, expressive, symbolic, semiotic, stylistic, genre-bound, traditional, and historic properties.I call such properties "Intentional", meaning by that to equate the Intentional and the cultural (or, the culturally meaningful --or, intrinsically interpretable).(Margolis 2000, 112) Although only expressed in a nutshell, the same notion of Intentionality is certainly present in the writings from the 1970s. 7In "The Ontological Peculiarity of Works of Art", Margolis defines the properties that make an artwork what it is by using almost the same words.The development of the philosopher's thought in this time frame must therefore be considered homogeneous: Broadly speaking, those properties are what may be characterised as functional or intentional properties and include design, expressiveness, symbolism, representation, meaning, style, and the like.
[…] Be that as it may, a reasonable theory of art could hold that when physical materials are worked in accord with a certain artistic craft then there emerges, culturally, an object embodied in the former that possesses certain orderly array of functional properties of the kind just mentioned.(Margolis 1977, 49) Regarding Margolis' early writings on art, in contrast, Russell Pryba (2021) observes that the philosopher employs a rather different terminology.In the essays written in the 1950s, he argues that art- works are such because they can be appreciated through a specific kind of perception: imaginative perception (Margolis 1958, 32;1959, 38).Thus, it seems here that Margolis comes close to the theory of art as aesthetic experience formulated by Monroe Beardsley (1958) in the same years.
Pryba is certainly right that Margolis' philosophy undergoes considerable evolution over the course of two decades between "The Identity of a Work of Art" (1959) and "The Ontological Peculiarity of Works of Art" (1977).Any reference to a special perceptual mode by which art would be experienced is slowly dropped.However, one further aspect should be noted.The imaginative perception of which the philosopher speaks in his early essays is not triggered by any property or characteristic that a special class of objects would possess in and of itself.Rather, this perceptual mode, which lies somewhere between sensory perception and the imagination, is possible only on the condition that the subjects involved in the experience of an artwork -the artist and their audience -have already learned the perceptual and imaginative habits of the society in which they live.In an essay just prior to "The Identity of a Work of Art" Margolis writes about a painter's activity: And the imagination of a culture can inspire the perception of a supervening work of art -itself sufficiently steady and clearly enough organised so that public discussion may range about it.[…] The original artist himself perceived such a work of art emerging as he applied paint to canvas, but he left only the canvas behind.The habits of perception and imagination that captured his society and himself and have proved sufficiently like those of our own society provide both for his attending and our attending to the same canvas as a work of art.(Margolis 1958, 33) If we interpret the passage quoted above correctly, then we should admit at least two consequences.On the one hand, we should admit that, despite the substantial difference in terminology, Margolis advocates a notion of Intentionality that is immediately social and cultural even in his very early essays on art.From this point of view, it is therefore possible to draw a line of continuity that binds together the different stages in which the philosopher's thinking evolves.On the other hand, we must recognise, for the second time, the influence of Peirce's semiotics and Pragmatism on Margolis' philosophy of art.As in Peirce's case, the social sharing of a habit seems to be a necessary condition for the existence, experience, and interpretation of any artifact.

Joseph Margolis' Ontology of Artifacts: Works of Art as Tokens-of-types
The culturalist conception of the Intentional properties that characterise artifacts also affects the way in which Joseph Margolis describes the ontology of the categories of type and token.A first point that must be emphasised is the fact that the philosopher, just like Peirce, considers the two entities as interdependent.A cultural product, he observes, only exists when it is actually embodied in an occurrence that can be perceived by the senses; conversely, such an occurrence will be perceived as an artifact only when it is identified by an abstract entity, which in turn will be determined by an organised set of social habits.A musical work, for example, will not exist unless it is actually performed by someone or recorded using notation; at the same time, a series of sounds will be identified with the token of a song only if these sounds are included in a network of culturalised behaviours.
From an ontological point of view, Margolis' insistence on the co-dependent relationship between type and token fulfills a double function.First, the philosopher uses it in the context of a refutation of idealist philosophies derived from the aesthetic theories of Benedetto Croce, an author who, although only partially and poorly translated into English in the interwar period, was highly influential through the work of Robin Collingwood (1938). 8Whereas these theories identify the essence of an artwork in the pure imaginative act occurring in the mind of an artist, by establishing the link between type and token, Margolis asserts the need to recognise art as both a corporeal and intersubjective ensemble of phenomena.Second, the interdependence between type and token functions as a distinctive criterion for differentiating the ontology of artworks from that of other generic entities.The relationship between an artwork and its occurrences cannot be the same as the relationship between a class and its members: while we can imagine a class that counts no members, we cannot imagine an artwork that is not embodied by any occurrences.Similarly, the relationship between a work of art and its occurrences cannot be the same as that between a kind and its examples: for universal kinds exist in the atemporal dimension of eternity; instead, a work of art can always be created and destroyed at given moments in time.
As in the case of Peirce's semiotics, the co-dependent relationship that exists between a type and the corresponding tokens also deter-8 Croce himself (1929) chose Collingwood as the translator of his Aesthetica in Nuce.
On the misunderstandings that occurred in the interpretation of Croce's theses in the United States, see Simoni 1952.

Dreon, Ragazzi Why Joseph Margolis Has Never Been an Analytic Philosopher of Art
JoLMA e-ISSN 2723-9640 3, 2, 2022, 333-364 mines an ontological asymmetry between the two terms.Beginning with the description of types, we shall say that Margolis identifies them with abstract particulars which have heuristic status and are embodied in physical occurrences.In "The Ontological Peculiarity of Works of Art", the wording is particularly clear: It must be possible to instantiate particulars (of a certain kind or of certain kinds) as well as to instantiate universals or properties.I suggest that the term "type" -in all contexts in which the type/token ambiguity arises-signifies abstract particulars of a kind that can be instantiated.(Margolis 1977, 45) And shortly thereafter: Types are actual abstract particulars in the sense only that a set of actual entities may be individuated as tokens of a particular type.(1977,47) Lastly: There are no types that are separable from tokens because there are no tokens except tokens-of-a-type.The very process for individuating tokens entails individuating types, that is, individuating different sets of particulars as the alternative tokens of this or that type.[…] What may mislead is this: the concept of different tokens of the same type is intended, in the arts, to accommodate the fact that the aesthetically often decisive differences among tokens of the same type (alternative performances of a sonata, for instance) need not matter as far as the individuation of the (type) work is concerned.
[…] This is simply another way of saying that works of art are culturally emergent entities […].(1977,49) Types are thus real entities according to Margolis, but their reality is bound, on the one hand, to the existence of no less than one physical occurrence for each of them and, on the other hand, to the existence of a homogeneous, at least implicitly shared and historically contingent cultural context.
First, it should be noted that the notion of type as a particular contrasts with the understandings of the term adopted by most analytic philosophies of art.Wolterstorff (1975) defines these entities as normative kinds consisting of all predicates attributable to the well-formed occurrences of an artifact: these predicates are established once and for all by the author of the work.This opinion seems to be shared by Jerrold Levinson (1990, 78-82), who defines the types identifying musical works as "structures-as-indicated-bya-composer-at-a-time".A Platonic conception of types is held by phi-JoLMA e-ISSN 2723-9640 3, 2, 2022, 333-364 losophers of music such as Kivy (1983Kivy ( , 1987) ) and Dodd (2000Dodd ( , 2002Dodd ( , 2007Dodd ( , 2008Dodd ( , 2010Dodd ( , 2012)): both argue that musical compositions are atemporal structures that are discovered and not invented by their authors.In a similarly Platonistic perspective, Currie (1989) defines all artworks as the causal chains of action-types necessary to produce a certain concrete artifact.Finally, David Davies (2004) identifies each work of art with the chain-of-action-token that produces the artifact embodying it and from which the relevant types can be abstracted as patterns of repetition.
In short, while Margolis defines types as abstract yet real historical (because culturally individuated) particulars, many analytic philosophers identify them as sets of necessary and sufficient conditions.These conditions are mostly thought of either as lacking temporal flexibility or determined by the intentional act of a single subject. 9One could hardly imagine an idea more distant from the philosophical system that we have sketched so far.
Margolis also reiterates the same view of the abstract entities identifying cultural products in later essays, where he consciously and critically abandons the terminology of the type-token theory: Sentences, artworks, selves, histories are ascribed determinate meanings, or meaningful structures, only in the way of suitable abstractions made within the shifting milieus of similar assignments made (or already made) of other such denotata.(Margolis 1999, 97) Much like Peirce, also Margolis seems to conceive of artworks as individuated by culturally determined and mutable rules of association.The continuity that, at least from this point of view, ties together his early writings on art and those from later stages is substantial.It must therefore be concluded that Margolis does not gradually stop using the type-token pair because his own theory has changed too much.Rather, he does so to avoid misunderstandings: to distance himself from the uses of those categories that do not correspond to his own but are widespread among many of his contemporary analytic colleagues.
Having clarified the ontological nature of types, let us now move on to consider the nature of tokens.According to Margolis, occurrences should not be immediately identified with and reduced to the physical objects with which they also coincide from a conceptual point of view: tokens are entities logically distinguishable from mere things.
9 For a critique of this notion of 'type', see Rohrbaugh 2003.Rohrbaugh himself seems to hold a very similar position to that of Margolis: both define artworks as non-physical historical individuals.Since he (2003,21) considers his own to be a metaphysical innovation but never compares it with Margolis' thinking, it would be interesting to know his opinion on the latter's philosophy of artifacts.

Dreon, Ragazzi Why Joseph Margolis Has Never Been an Analytic Philosopher of Art
JoLMA e-ISSN 2723-9640  3, 2, 2022, 333-364 Evidence of the ontological difference between physical things and tokens can be directly observed by considering ready-mades, a genre of art that includes works having properties different from those of the objects with which they coincide.Take the example of In Advance of a Broken Arm, by Marcel Duchamp: although the work consists solely of a snow shovel, the French artist's sculpture does not enjoy the same properties as all other snow shovels. 10The artwork possesses not only an economic value, but also aesthetic and cultural attributes that are quite different from those possessed by the tool.For example, everyone will be able to say of a shovel that it is useful and sturdy, but nobody will be able to say the same thing about Duchamp's sculpture: one may say instead that it is intellectually refined and Dadaist.This happens not because some obscure metaphysical force is added to Duchamp's chosen shovel, but rather because the object, as it is, is placed in a network of relations different from and ulterior to that within which it is usually understood.It is the relationships with other objects, events, and behaviours in which the artist places the work that allow the latter to acquire new properties.These acquired characteristics always transcend the medium on which the artwork is based, although its existence always depends on the matter of which it is made. 11 In considering Margolis' conception of tokens, we are prompted to note, for one last time, the philosopher's closeness to Peirce's Pragmatism.As mentioned before, even in Peirce's semiotics the dynamical object is never presented in and of itself, but always through the mediation of a complex and situated system of signs, beliefs, and knowledge -what is called the "ground" in Peirce's jargon.It is from this preexisting and shared system of habitual associations that the interpretation of a certain cultural entity derives, and thus also the validity of the relationship between the type that identifies it and the embodying tokens.
Margolis points out that the relations established between already existent features of reality cause new properties to emerge from and transcend them.Such emergent properties are real, although their existence is dependent on and conditioned by matter itself.A change in the apparatus of relations existing between a given set of physical elements produces an evolution that increases the properties and functions of which matter itself is capable.9640  3, 2, 2022, 333-364 Types, tokens, and the relationship between the two sets, then, are not conceived by Margolis as either ideal and mental or independent; instead, they depend on the complex and ever-changing cultural background within which they are always inscribed.Given what has been said so far, we can claim that Margolis' philosophy of art has always involved a form of radical historicism which is alien to the versions of the type-token theory formulated within the analytical tradition.Compatible with the pragmatist project, the philosopher's aim has always been to find a third way between idealism and materialism that can account for the ineradicable social and historical dimensions within which works of art take on meaning and are interpretable.
Notwithstanding the lines of evolution that we have touched upon in these pages, Margolis' theoretical aim remains essentially unchanged over time.The philosophical evolution uniting the early writings on art and the later essays can thus be described as the inexhaustible attempt to explain the following definition: Works of art are physically embodied and culturally emergent entities.(Margolis 1974, 187) The previous paragraphs, we might say, offer some guidance for interpreting Margolis' use of the categories of type and token in light of this definition.Our purpose has been to show that, although the philosopher participated in central debates shaping the so-called analytic philosophy of art, he did so in the wake of Charles Peirce's pragmatist semiotics.
In the following pages, we instead wish to pursue a different goal.We will focus on the second part of the definition quoted above, and in particular on the notion of emergence.Through the study of this concept we will see how Margolis articulated a system of thought such that the theory of art is always proved to be dependent on a broader philosophical anthropology.
While so far we have analysed Margolis' philosophy of art from the point of view of its theoretical roots, in the next few pages we will illustrate it from the point of view of its long-term goals.

3
On Cultural Emergence and Its Consequences sic continuity is that Margolis assumed the peculiar properties characterising works of art and broadly cultural phenomena to provide an answer to the major ontological questions troubling the analytical philosophy of art after Morris Weitz's (1956) famous paper "The Role of Theory in Aesthetics" and, more importantly, the disrupting developments of contemporary arts.As Margolis himself always argued (although with increasing emphasis over the years), the very idea of culturally emergent properties implied a more extensive enterprise than the analytical ontology of art was prepared to concede, namely a theory of the emergence and evolution of culture out of pre-existing biological resources and an account of the emergence of human beings as cultural artifacts out of primates and the fortuitous yet irreversible appearance of language. 12Even though the concept of emergence is needed in order to develop a full-fledged form of naturalism without any reduction in Margolis' later thought (2010), we would argue that it already plays a crucial role in his analysis of the ontological peculiarities of artworks and paves the way precisely for those further developments.
In what follows, we will support a continuistic reading of Joseph Margolis' work; 13 more specifically, we will defend the thesis that his account of cultural properties as emergent ones involved a strong challenge to what we suggest calling the "causal closure of the artworld" -namely, the claim that an ontology of artistic entities and a definition of art could be formulated and stand on its own resources.
In order to do so we will integrate Margolis' first characterisation of cultural emergent properties ( § 3.1) through his reflections on the very concept of emergence even beyond the cultural world and cultural phenomena ( § 3.2), in order to return to the issue at stake in his first paper and evaluate the consequences of this conceptual category within the philosophy of art ( § 3.3). 14

Emergent, Cultural, and Intentional Properties
In his 1974 article, Margolis seems to address the issue of the ontological status of artistic entities according to the modes of approach and even the stylistic features belonging to the analytic philosophy of art.His declared purpose is to provide "a theory of the actual ontological standing of a work of art" (Margolis 1974, 187), where the first part of his definition -works of art are "physically embodied" -rep-12 See Margolis 2017 for a comprehensive summary of this view.
14 For a different analysis of Margolis' characterisation of artworks in term of emergence and embodiment, see Jacquette 1986.
JoLMA e-ISSN 2723-9640 3, 2, 2022, 333-364 resents the extension, which is to say the solution to the question of identifying and fixing the reference, while the second part -works of art are "culturally emergent entities" -identifies or intensionally specifies something as a work of art.Even the characterisation of emergence is to a certain extent minimal: works of art are emergent entities because they exhibit properties that cannot be ascribed to the objects in which they are incorporated (Margolis 1974, 189; for a similar definition see Margolis 2000).Michelangelo's David is an emergent entity because it expresses strength and fierceness in a sense that cannot be shared by the white marble it is made of.
However, Margolis' statement is much more complex.On the one hand, it involves a form of pluralistic, already tolerant realism, assuming that works of art are real parts of this world, although they are not reducible to physical entities.Speaking of emergence allows him to maintain a form of consistent materialism while resisting reductionism, but the implicit issue at stake is none other than calling into question the very idea that physical entities represent the paradigm of what it means to be real, as well as the assumption that only physical things can be causally effective.These points will become explicit a few years later, when Margolis (1987) will criticise the claim about the unity of science, as we will see in the following section.On the other hand, it implies an extension of the ontological focus far beyond artistic products to any kind of cultural entity or person.For the moment, the consequences of such an ontological assimilation have not yet been completely laid out, but later it will become clear that this is the first step toward Margolis' claim that works of arts as well as speeches and other cultural practices form the kind of cultural utterances through which animals of peculiar sort became (phylogenetically) and still become (ontogenetically) human selves through the feedback action of language and cultural practices with respect to the organic resources they are made of and the ways these resources are organised. 15 For the moment, the author's emphasis is on artworks as culturally emergent entities, namely as ones that display a basic connection with a "cultural tradition", a "cultural context", and "contextual assumptions" (Margolis 1974, 191-2).Margolis is not simply defending the need of an interpretation in order for something to be considered a work of art.His claim is more radical, insofar it implies that the emergent properties of an artwork cannot be perceived as such apart from a form of life quite exceeding the boundaries of ar-15 The process of humans' emergence from primates is conceived of as double, as noted by Pryba 2015, 229;Pihlström 2015, 101;and Hildebrand 2015, 42: in other words, according to Margolis 2015, 27, it involves both an internal Bildung and an external Bildung.The point, to us, is that these processes should not be considered as ordered in terms of premise and consequence, but as mutually constitutive Dreon 2017.tistic theories and institutions, in contrast to Danto's and Dickie's approaches.Here we need to return to Margolis' conception of "Intentionality" and consider it in greater depth.
In his 1974 article, Margolis is also introducing the idea that artworks are "Intentional objects" (Margolis 1974, 192).In this paper, the word "Intentional" -written with a capital I -means that artworks are significant only with reference to "human society", namely a complex set of practices and the conventions of a given culture, occurring through human bodily actions, such as rising an arm to greet someone or nodding in approval.It even involves a criticism of Danto's claim that an artwork is always intentional -written in lowercase -in the sense of being about something: carpets and landscape gardens are not about anything, Margolis says polemically. 16Some years later, Margolis will provide a mature definition of Intentionality: in The Cultural Space of the Arts, he states that Intentional means "culturally significant and/or significative" (Margolis 2010, 49).It includes intentionality in the phenomenological sense of the term, albeit an essentially revised one, which excludes any kind of mentalism and/or internalism, because it assumes Intentionality to be a function of the cultural practices taking place among people rather than in the alleged internal theatre of one's mind.

Refining the Concept
It is over ten years later that Margolis (1986;1987) comes to deal extensively with the concept of emergence in the field of science and beyond the sphere of cultural phenomena, even if the core issue he wants to tackle is precisely Intentional properties and their explicability within the unity of science program.In 1986, he publishes an article entitled "Emergence" in The Philosophical Forum and devotes a whole chapter to "Emergence and the Unity of Science" in his book Science Without Unity, published in 1987.Intentional or broadly cultural properties are the major novelty with respect to the physical and organic world, which could be dealt with by means of the concept of emergence when adopting a realistic yet non-reductive stance, as happens in Margolis' case.By contrast, they represent an obstacle when espousing a physicalist program in the field of natural sciences or pursuing the so-called naturalisation of epistemology in philosophy.In his chapter in the 1987 volume, Margolis engages with the po-16 Even though it must be said that Danto would not have considered gardens and carpets to be works of art in the absence of a theory including those kinds of artifacts within the artworld.For this reason, he did not hesitate to exclude cave paintings from the realm of art (Danto 1964).
JoLMA e-ISSN 2723-9640 3, 2, 2022, 333-364 sition of Herbert Feigl, an early member of the Vienna Circle, as well as with the theories of Mario Bunge, the influential physicist and philosopher.Bunge's conception of emergence is interesting for Margolis because it seems to allow a form of emergentism that is compatible with reductionism.Against this background, Sellars, Davidson, and Putnam are also frequently mentioned by Margolis, confirming Cahoone's (2021) view that Margolis's naturalism did not came from the so-called Columbia Naturalists -from Frederik Woodbridge to Ernest Nagel and Justus Buchler -but was nourished through his engagement in the debate on the philosophy of mind.In any case, it is clear that the output of his research is not analytic at all, being committed to a form of non-reductive naturalism that includes emergence at its core and is consequently already in line with the explicit endorsement of Pragmatism he expressed in later years.This probably occurred through a complex series of influences: his strong interest in the peculiarities of cultural entities, as we have already seen; his early exposure to Pragmatism at Columbia (Margolis 2014); even his closeness to Marjorie Grene's philosophy of biology and philosophical anthropology, as proven by the warm dedication in a previous (and almost complementary) book, Culture and Cultural Entities.Toward a New Unity of Science (Margolis 1983).
Margolis' focus on emergence in these works lies at the heart of his criticism of the established idea that it is possible to explain nature on the basis of a unitary and all-encompassing order that can be detected scientifically.Emergent properties represent a challenge to this idea and Margolis believes that, if we honestly recognise the reality of cultural phenomena, we must abandon the whole project, opting for a more pluralistic, inclusive, and complex conception of nature.In addition to his previous works, Margolis focuses here not only on cultural phenomena, but also on biological entities and processes, being conscious that living organisms already present emerging properties that cannot be exhaustively explained in physical terms.The reason for this is that he already assumes as paradigmatic the emergence of living organisms from the inorganic world, the emergence of Homo sapiens from primates, and the emergence of language from physical nature -topics that will be at the centre of his later philosophical anthropology of the human being as a "natural artifact" (Margolis 2016).17 Without delving into the details of Margolis' analysis of Feigl's and Bunge's concepts of emergence, it is important to point out a couple of aspects.Firstly and in contrast to his previous, minimal definition of emergent properties, here Margolis (1987) explicitly connects the idea of emergence with the notion of system, whose emergent proper-JoLMA e-ISSN 2723-9640 3, 2, 2022, 333-364 ties do not appear to be analysable and/or explainable through their reduction to pre-existing components, although there is evidently a pertinent relation between the former and the latter.In a few words, emergent properties are properties that cannot be explained as the result of the mere association of the original system's components.One problem is the lack of a directly causal and/or generative link between pre-existing features and an emergent property: for example, between neurological processes and the obsessive nature of certain mental states, or between brush strokes and, say, the aggressive vividness of a Blaue Reiter painting.Another crucial problem is that, from a physicalist point of view, the notion of a system includes the idea of its closure and the denial of the so-called downward causation exercised by new properties (say, mental states or social practices) on previous components of the system (say, brain processes).For the physical reductionist, only physical events and entities are real and physical events can only have physical causes (Davidson 1970;Kim 1998): the boundaries of the realm are sharp and any physical change must be explained separately, on exclusively physical grounds.This point will prove important in order to shed light on the way in which Margolis used cultural emergence to characterise artworks some years before, as we will see in the next section.
Secondly, the divergences between the two main conceptions of emergence we have considered regard their being compatible with the physicist program, namely the claim for a unity of science based on reducing different non-primarily physical phenomena to physical explanations.While Feigl's view leaves no room for emergence within reductive naturalism, Bunge's conception of emergence is compatible with a form of reductionism because it admits a plurality of interconnected systems, while maintaining that they are hierarchically ordered.According to this view, inferior systems are open to superior ones, but their relations are linear and univocal; consequently, a weak unity of science is still guaranteed, if not strong physicalism -namely, a view of science that Margolis rejects as still conservative and dogmatic.
In addition to offering a reminder of the historicity of the very concept of 'physical' entities or processes, Joseph Margolis espouses a view of emergence involving causal connections and the efficacy of newly emergent properties with respect to already existing resources -mental events can have an impact on neurological processes, just as literary works can have a disrupting effect on the reader's habits of action and beliefs.Consequently, his conception of emergence involves the exclusion of the so-called closure of the physical world, and the denial of the view of nature as a unitary system or a hierarchically ordered system of systems, ultimately regulated by homonomic laws.Margolis emphasises that not only do broadly linguistic and cultural phenomena resist hierarchical organisation, but so do JoLMA e-ISSN 2723-9640 3, 2, 2022, 333-364 biological systems, which appear to be complex, non-hierarchically ordered systems equal to psychological and cultural systems.Ultimately, we cannot deny that much of nature is complex and escapes a univocal logic.Margolis has no hesitation in assuming that linguistic and cultural phenomena are natural developments of previous physical and organic resources and that we do not have any need to refer to extra-natural interventions to explain them -as he will explicitly affirm when endorsing the pragmatist view of culture as continuous with nature (Margolis…).Entities and events within the cultural world are as real as those in the physical world; they are causally efficient and do not give raise to a closed system, because they cannot exist apart from cognitive and social practice and a form of life.Moreover, complex systems -both biological and cultural systems -cannot be explained in purely functional terms, i.e. independently from the material means through which they work.Certainly, even this last insight has important consequences on the view of works of art as culturally emergent entities, as we will make clear in what follow.

3.3
Getting Rid of the "Causal Closure of the Artworld" Hence, what consequences can we derive from Margolis' development of the concept of emergence that might be relevant for his view of the ontological status of artworks?It is clear that his idea of emergence came to play a crucial role in his philosophical anthropology -leading him to regard humans as natural artifacts, produced through the feedback actions of language and transformed through cultural practices. 1818 See, for example, Margolis' succinct overview of the human in his Prologue to The Arts and the Definition of the Human.Toward a Philosophical Anthropology: "The human is artifactual; socially constituted; historicised; enlanguaged and encultured; 'second natured'; real only within some culture's collective life; embodied through the cultural transformation of the infant members of Homo sapiens; originally or externally gebildet; sui generis; emergent through mastering a first language and whatever aptitudes such mastery makes possible; indissolubly hybrid, uniting biological and cultural processes and powers; capable therefore of hybrid acts or "utterances" (speaking, making, doing, creating) incarnate in the materiae of any part of physical nature; self-transforming or internally gebildet through its second-natured powers; empowered and constrained by the collective history it shares with similarly emergent creatures; capable, thus, of functioning as a self, a person, a subject, an agent, within an aggregate of similarly formed selves, that is, free and responsible, capable of causally effective (incarnate) initiatives, capable of self-reference, of reporting its inner thoughts and experience in a public way, of understanding the utterances and acts of similarly endowed selves; inherently interpretable and subject to change through being interpreted; not a natural-kind entity but a history, or an entity that has a history rather than a nature, or a nature that is no more than a history-a history determinable but not determinate.All in all, the human is a unique sort of being, you must admit, but an individuated being nevertheless: emergent in part by natural (biological) means and in part However, it is less evident how this idea affected the ontology of art he had developed.It is precisely this aspect that we now wish to make the focus of our attention.
In addition to what has already been stated in the previous section, let us recall a useful definition of emergence provided by Margolis: By an emergent order of reality […] I mean any array of empirical phenomena that (i) cannot be described or explained in terms of the descriptive and explanatory concepts deemed adequate for whatever more basic level or order of nature or reality the order or level in question is said to have emerged from, and (ii) is causally implicated and cognitively accessible in the same "world" in which the putatively more basic order or level is identified.(Margolis 1995, 257) 19   We are now in a position to clarify the details of Margolis' thesis that works of art are culturally emergent entities.
First of all, works of art are real, although their properties -say, the meaning of a psalm within a religious ceremony -cannot be causally explained by simply referring to the material conditions on which they rely.Second, works of art (can) exert a causal action on the components of the material world from which they have emerged: for example, reciting a psalm in a community of believers can have a calming effect on bystanders, it can strengthen mutual bonds or excite more suggestible individuals.This means assuming a complex view of causality, as constituted by multiple concomitant factors, and positing a feedback action by cultural practices on socio-cultural features -e.g.reinforcing bonds between believers -as well as on physiological processes -e.g.relaxing muscles and producing endorphins.Thirdly, the alleged "basic order of nature", "the first level of reality", or the "zero-system" from which the emergent properties of artworks emerge is not a merely physical world, but a Lebenswelt consisting of the various forms of human life: a complex system of relations and systems that can be locally ordered and distinguished, but are not hierarchically ordered, as they frequently overlap and are mutually connected.Consequently (and fourthly), artworks cannot be taken to constitute an autonomous, self-standing realm or system.For example, Warhol's Brillo Box is not supervenient upon a physical level of merely physical entities by virtue of a theory of art and the attribution of the "is" of artistic identification (Danto 1964).Indeed, by artifactual or cultural transformation-possibly, then, a conceptual scandal or even the living refutation of many a convention of canonical philosophy" (Margolis 2008, 19).
19 I am drawing this quotation from Sami Pihlström's essay, an extensive section of which is devoted to Joe Margolis' conception of emergence (Pihlström 2015).
JoLMA e-ISSN 2723-9640 3, 2, 2022, 333-364 even the formulation of artistic theories belongs to a specific cultural context and tradition -a secularised context where many efforts have been made in relation to a wooden Madonna to isolate the work of art from the worship object, as well as from the handcrafted artifact (Shiner 2003).Although in late modern times and within Western culture we have witnessed a process of partial artistic autonomisation, the artworld is far from being a closed, self-standing system -it cannot stand "to the real world in something like the relationship in which the City of God stands to the Earthly City" (Danto 1964, 582).On the contrary, it "is causally implicated and cognitively accessible in the same 'world' in which the putatively more basic order or level is identified" (Margolis 1995, 257), as Margolis states in a passage already quoted. 20 Finally, the means by which artworks are embodied are not indifferent and merely functional, which is to say that they cannot be replaced with other means without making any difference: similarly to biological systems, works of art are complex systems where the ways in which a specific function is realised contribute to determining the function itself -to put it with the Pragmatists, means contribute to making ends (Dewey …, Hickman…).To take an extreme example, Kosuth's famous One and Three Chairs claims to show a chair independently from how we grasp it (through the direct perception of the object itself, through a picture of the object, or by means of a definition), yet in this case too the peculiar assemblage in which the work of art is embodied enters into its very constitution and is decisive in shaping and fixing the artwork's identity and capacity to have an impact on the surroundings.

Conclusions
Between the late 1950s and the 1970s, Joseph Margolis substantially contributed to debates that were to prove crucial for the development of the analytical philosophy of art.His ontology of artifacts was commented upon and reworked by authors such as Wollheim (1968) 20 For a more extensive treatment of the Danto-Margolis debate, see Pryba (2015).
Pryba interprets Joseph Margolis' critical claim that Danto's theory prevents him from recognising the reality of paintings according to his conception of perception.The problem for Margolis is that Danto considers the perception of artworks to consist of purely sensory, organic or biological processes, immune to historical change, and thus denies perception any role when it comes to tracing a distinction between works of art and ordinary things.In Margolis' view, Danto is incapable of seeing that human perception is permeated by cultural concepts, i.e. he fails to appreciate that the cultural features produced by human practices and societies affect perception itself -that they exercise a kind of downward causation on previously existent forms of animal perception, to put it in emergentist terms.and Wolterstorff (1975), among many others belonging to that specific philosophical tardition.However, our argument is that the philosopher's theoretical trajectory should be interpreted in a different way.Indeed, the pragmatist perspective that permeates Margolis' more mature essays can also be identified -albeit in a less articulate form -in his early writings on art.
To prove our point, we set out in two opposite and complementary directions.On the one hand, we studied the way Margolis introduced the categories of type and token into his own ontology of art.In this way, we retrospectively highlighted the filiation between the latter and Charles Peirce's semiotics.On the other hand, we retraced the way in which the concept of emergence employed in Margolis' definition of artifacts was gradually clarified and specified over the years, until it became the pivot of a broader philosophical anthropology.
From an ontological point of view, from the late 1950s onward Margolis defined all artifacts as tokens of types.In doing so, he adopted an understanding of the two categories that in many respects mirrored Peirce's.Like the latter, Margolis holds that tokens are linked to their types by a relationship of inescapable co-dependence: whereas a type cannot exist unless it is embodied in at least one physical object or event, a token cannot be identified as such unless it is recognizable as the occurrence of an abstract entity.In addition to this, there are two further elements of similarity.Firstly, Margolis and Peirce define types as abstract particulars: they are real individual entities, although they are dependent on the matter in which they are embodied.Secondly, both philosophers believe that tokens are not simply equivalent to physical objects: the two members of the categorical pair are only identified as such if they occur within the framework of at least implicitly shared and historically determined behaviours and habits.
Insofar as Margolis understands the type and token pair in this sense, his theoretical system differs in at least two respects from those of his analytical colleagues who employ the same categories.A first difference concerns the nature of the types that identify each artifact.In all analytical ontologies, these abstract entities are thought of as sets of conditions or attributes that identify specific classes of objects or events.Consequently, works of art are conceived of as properties possessed by certain physical phenomena.To this attributive model Margolis opposes the idea that each artifact is the token of an abstract particular to which different properties can be attributed depending on the cultural contexts within which its occurrences are produced.A second difference concerns the way in which intentionality is acknowledged to be a constitutive characteristic of all works of art.Whereas in Margolis' understanding the notion of Intentional coincides with that of cultural, and thus has boundaries located in the sphere of the collective, in analytical philosophies of art the intentional properties of an artifact are either determined by the JoLMA e-ISSN 2723-9640 3, 2, 2022, 333-364 action of a single subject oriented by a certain theory or institution, or are properties of thought.
These two differences combined produce considerable effects from an ontological point of view.Regardless of the specific cases, analytical philosophies of art do not allow for the recognition of temporal flexibility as an inherent characteristic of all artifacts.Theories such as Wolterstorff's (1975) assign individual authors or artists the responsibility of defining the essential properties of their works; Platonistic theories such as Currie's (1989) situate the principle of individuation of each artifact outside of time.By contrast, by conceiving of types as culturally determined, Margolis' ontology of artifacts is characterised by a historicism that is as substantial as it is radical.
Considering the role that the categories of type and token play in Margolis' philosophy of art has allowed us to show that a pragmatist perspective is present in it from the very beginning.It has also allowed us to grasp a first sense in which we can understand the philosopher's definition of artifacts as physically embedded and culturally emergent entities.
The concept of emergency was at the core of our second inquiry.Margolis used it consistently and in an increasingly specific way from the 1970s to the end of his career.In his early aesthetic writings, the notion is used to highlight the fact that works of art exhibit Intentional properties that cannot be ascribed to the mere physical objects in which they are embodied.In later writings, the scope of the term expands to become the pivot of a genuine philosophical anthropology and ontological theory.In this sense, the emergence of artifactual properties from objects is taken as a model to explain other events, such as the emergence of mind from brain mass and, more generally, of life from matter.
Within the framework of his own ontology, Margolis uses the notion of emergence to describe a system in which properties appear that are connected to matter by relevant relations, yet cannot be explained in terms of the mere association of pre-existing features and involve novel forms of organisation.Since the theory rules out the presence of any additional metaphysical forces beyond the material world, we might say that the philosopher embraces a position of non-reductive naturalism.In this article, we have tried to show how Margolis' model differs from other, similar forms of naturalism.
The main difference between Margolis' theory and, say, Donald Davidson's anomalous monism concerns the organisation of causal relations.Physicalist naturalism assumes that lower-level events can cause higher-level events, but not the other way round.This implies the closure of any natural system.In contrast, Joseph Margolis argued with increasing conviction that causal relationships between higher and lower levels are possible.In the context of his philosophy of mind, for example, the philosopher held that there exist not only JoLMA e-ISSN 2723-9640 3, 2, 2022, 333-364 causal relationships from the physical realm to the mental one, but also feedback effects in the opposite direction.In this framework, the concept of emergence can be associated with that of a complex system in which each level may be causally relevant for all others.
This difference in the way naturalism is understood entails a further difference in the way the scientific project as a whole is conceived.Physicalist philosophies advocate the unity of the sciences based on the possibility of explaining non-primarily physical phenomena or at least identifying them as purely physical events.Margolis' emergentism instead implies a denial of the idea that nature is a system of systems hierarchically regulated by homonomic laws.
As we have tried to show, this view of nature and science also has important implications for Margolis' understanding of artifacts.Firstly, consistently with his own way of explaining causal relations, the philosopher believes that cultural objects can have an effect on components of the material world.Conversely, the lowest level of reality from which the properties that characterise artworks emerge does not coincide with causal chains located in the physical realm, but rather with a form of life already constituted by a complex web of relations between different levels of reality -the material, the mental, and the cultural.Works of art, although dependent on physical objects, are therefore real in their own right.
An investigation of the general ontology proposed by Margolis finally led us to point out a substantial difference in the way the philosopher solves the problem of defining art compared to his analytical counterparts.Indeed, authors such as Danto and Dickie seem to suggest that art too is a closed system, fully definable by referring to elements that are allegedly internal to it and constitutive of it, such as art theories or institutions.Margolis, by contrast, believes that this is not the case and definitely rejects the idea of the autonomy of the realm of art, the artworld, or art institutions.The premise that no system is actually closed, but that all systems are mutually linked within a historical flux, is coherent with the philosopher's claim that a definition of art and its objects can only be given in Intentional terms, that is, by participating in a form of life whose boundaries and peculiarities are mutable over time and continuously made to fit practical purposes.
Non-reductive naturalism and historicism are two perspectives that Margolis set in an ever-closer alliance, in both his ontology and his philosophy of art.In this article, we have attempted to highlight how the seeds of this alliance were sown by the philosopher in his early writings on art and then sprouted into comprehensive ontological and epistemological theories.If our argument is correct, we will then have to abandon the idea that there was an analytical Margolis, who only later approached Pragmatism.Instead, the trajectory of his thought will have to be envisaged as a line of development that is as continuous as it is coherent.

10
Marcel Duchamp, In Advance of the Broken Arm, wood and galvanised iron snow shovel, 132 x 35 cm, 1915.11 For a detailed account of how Margolis conceives of the categories of type and token and the codependent relationship existing between the two terms, see Jacquette 1986; 1994.JoLMA e-ISSN 2723-