WHAT IS AN INSTANCE OF AN ARTWORK?

The expression ‘an instance of an artwork’ is often used in philosophical discourse about art. Yet there is no clear account of what exactly this expression means. My goal in this essay is to provide such an account. I begin by expounding and defending a particular definition of the concept of ‘an instance of an artwork’. Next, I elaborate this definition – by providing definitions of the main derivatives of the concept of ‘an instance of an artwork’, namely the concepts of ‘a well-formed instance of an artwork’ and ‘a non-wellformed instance of an artwork’. Finally, I examine the relation of the foregoing definitions to the existence and identity conditions of artworks and make some additional remarks concerning these definitions.

expression mean? This question has received scant attention in the philosophical literature. 2 Meanwhile, the question is doubtless important.
By answering it, we would improve our understanding of philosophical discourse about art. Furthermore, an acceptable answer to it would provide a philosopher of art with a tool that, at least in theory, would enable her to establish whether an object is, or is not, to be categorized as an instance of an artwork. Also, such an answer might have non-trivial implications concerning certain ontological properties of artworks -primarily, the properties of existence and identity. 3 Finally, it might provide a means to resolve some philosophical issues, such as the issue of whether all artworks are multiple in nature and the issue of whether the novel can be treated as a performing art. 4 My goal in this essay is to answer the question posed above. In particular, I aim to define the concept of 'an instance of an artwork' as well as its constituents -the concepts of 'a well-formed instance of an artwork' and 'a non-well-formed instance of an artwork' . I begin by defining 'an instance of an artwork' (Section II). Next, I defend this definition against potential objections and provide reasons to adopt it (Section III). Then I turn to defining the concepts of 'a well-formed instance of an artwork' and 'a non-well-formed instance of an artwork' (Sections IV and V). Finally, I make a few additional remarks regarding the concept of 'an instance of an artwork' . In particular, I define 'a token of an artwork' , which is closely related to 'an instance of an artwork' , specify certain ontological implications of the definition of 'an instance of an artwork' , and provide an alternative formulation of this definition (Section VI). For an illustration of this point, see Section VI, Remark 2. 4 For an explanation of how an account of 'an instance of an artwork' could be used to resolve the issue of whether all artworks are multiple in nature, see Davies, 'Multiple Instances'; 'Enigmatic Variations'; 'Varying Impressions'; for an explanation of how such an account could be used to resolve the issue of whether the novel can be treated as a performing art, see Alexey Aliyev, 'What Instances of Novels Are' , Philosophia 45 (2017): 163-83. A remark concerning the existing alternatives to the account presented in this essay. So far as I am concerned, the only explicit account of the nature of the concept of 'an instance of an artwork' has been offered by David Davies. In his view, something is an instance of an artwork just in case it (a) 'makes manifest to the receiver some or all of the experienceable properties bearing upon its appreciation' 5 or (b) stands in a particular 'kind of historical-intentional relation to the work's history of making' 6 -the relation that grounds the fact that the work is the logically first product of the artist's creative activity. 7 However, this account has certain shortcomings. First of all, it appears insufficiently informative, as some of the expressions it employs are not quite clear. Thus, what exactly is meant by 'some' in clause (a)? Davies does not provide an answer to this question. At the same time, in this case, possible conventional answers are unsatisfactory. We do not want to understand 'some' as 'at least one' , since an entity manifesting just one experienceable property bearing upon the appreciation of an artwork does not have to be an instance of this work (thus, a black cat makes manifest one of the experienceable properties that bear upon the appreciation of Wassily Kandinsky's Black Relationship -the property of blackness -but, obviously, such a cat is not an instance of this painting). And, for similar reasons, we do not want to take the meaning of 'some' to be equivalent to 'a few' . So, given what has been said, the meaning of 'some' in the given case is unclear. Meanwhile, without an acceptable clarification of this meaning, Davies's account of 'an instance of an artwork' can hardly be called sufficiently informative.
Furthermore, this account does not seem to fully accord with our linguistic practice -in particular, with the actual use of the expression 'an instance of an artwork' . 8 As is clear from what has been said in the previous paragraph, according to Davies, something is an instance of an artwork if it stands in a particular 'kind of historical-intentional relation to the work's history of 5 Davies, 'Varying Impressions' , 81. 6 Davies, 'Multiple Instances ' , 414. 7 For a detailed description of this relation, see ibid. 8 The expression 'instance of an artwork' (as well as related expressions: 'well-formed instance of an artwork' , 'genuine instance of an artwork' , 'instance of a novel' , and so on) is technical: It is used primarily by philosophers and is normally absent from ordinary, non-philosophical discourse. Thus, if you go to a bookshop and ask the salesman for an instance of War and Peace, he or she most likely will have a hard time trying to figure out what it is exactly that you want. Similarly, many of those who are not familiar with philosophy will be puzzled if they hear you say that you have never listened to a well-formed instance of The Rite of Spring. In light of this, in this essay, it is assumed that the actual use of the expression 'an instance of an artwork' amounts to the actual use of this expression by philosophers. making' . 9 As a result, his account entails that there can be instances of artworks that fail to make manifest to receivers many or even all of the experienceable properties bearing upon the appreciation of these works. For example, an artefact cast from the mould of a work of cast sculpture can be an instance of this work even if it 'has lost all of its detail through erosion and therefore no longer has any capacity to play the experiential role in the appreciation of [the] work' . 10 However, this consequence seems to go against the actual use of the expression 'an instance of an artwork' . An overwhelming majority of theorists, including Stephen Davies, Nelson Goodman, Jerrold Levinson, Guy Rohrbaugh, and Nicholas Wolterstorff, would agree that to be an instance of an artwork, an entity must provide at least some (and, presumably, sufficient) experiential access to this work; if an entity fails to do that, then it cannot even be called 'a non-well-formed instance of this work' -it should be categorized as something that is not an instance of this work at all. 11 The account I present in this essay, while similar to Davies's account in some respects, does not have the foregoing shortcomings. And, at the same time, it does not seem to have any other shortcomings. In light of this, I am inclined to think that it can help us substantially improve our current understanding of the concept of 'an instance of an artwork' and, as a consequence, achieve the results specified at the beginning of this essay.

II. DEFINING 'AN INSTANCE OF AN ARTWORK'
As is clear from what has been said above, the central question of this essay is the question 'What is meant by "an instance of an artwork"?' , or, in other words, 'How can "an instance of an artwork" be defined?' To answer this question, let us specify some characteristic features of an instance of an artwork.
What seems common to all instances of artworks is that they manifest various properties of these works. Take, for example, an instance of Beethoven's Symphony No. 5 -for example, one of its correct performances. It manifests certain properties of this symphony -primarily, the properties related to its sound. Or consider an instance of the Mona Lisa -say, its original canvas. It manifests certain properties of this painting, namely the properties concerned Tolstoy's War and Peace -such as one of its correct readings 12 -manifests certain properties of this novel, including its sonic properties and properties related to the semantic content.
Here, one might ask: What exactly is meant by saying that x manifests a property? This question can be answered as follows: For all x, x manifests a property just in case this property is apprehensible by directly perceiving xthat is, by perceiving x with the help of one or more of our sensory faculties, such as sight (vision), hearing (audition), taste (gustation), smell (olfaction), and touch (somatosensation). Regarding this answer, two remarks are worth making. First, the only thing that determines whether an entity manifests a property is whether this property can be apprehended by perceiving this entity with the help of at least one sensory faculty. Thus, a red apple manifests the property of being red -since this property can be apprehended by perceiving this apple with the help of the faculty of sight. At the same time, an apple produced in Florida does not manifest the property of being produced in Florida -because this property cannot be apprehended by perceiving this apple with the help of any sensory faculties.
Second, manifesting a property is not equivalent to having this property.
For an object can have a property without manifesting it. Thus, an apple produced in Florida has the property of being produced in Florida but, as mentioned above, does not manifest this property. Similarly, a musical score has the property of sounding a particular way but does not manifest this property. (If it did, then at least some sonic properties could be apprehended by perceiving it with the help of a sensory faculty. However, no sonic property can be apprehended that way. In order for this to be possible, it must be possible to hear a musical score. But no musical score can be heard, since (a), strictly speaking, only sounds can be heard, and (b) a musical score is not a sound -rather, it is a concrete sequence of notes and other symbols.) 13 Thus, given what has been said, one characteristic feature of an instance of an artwork is that it manifests certain properties of this work. But what exactly are these properties? It is reasonable to suppose that they are such that if they are not experienced by us, then we will not be able to fully appreciate this work. 12 By 'a reading' here is meant a sequence of sounds generated as a result of reading a text. For an argument in favour of the view that readings are instances of novels, see Aliyev, 'What Instances of Novels Are' . 13 Of course, we can meaningfully say that a musical score can be heard. But when we say this, we do not mean that this score can literally be heard; what we mean is that the sounds generated with its help (or perhaps the sounds it encodes) can be heard.
In other words, they are properties that must be experienced for a complete appreciation of the work.
But now the following question arises: Is it necessary for an instance of an artwork to manifest all the properties that must be experienced to fully appreciate this work? It may seem as though the right answer is 'Yes' . But, in fact, that is not the case. Consider Alexander Ivanov's painting The Apparition of Christ Before the People. It is reasonable to assume that to fully appreciate this painting, it is necessary to learn about the process of creating this painting. 14 Meanwhile, learning about this process is impossible without engaging experientially with at least some of the properties of Ivanov's preparatory sketches (or so we can assume). 15 Thus, experiencing these properties is requisite for a full appreciation of The Apparition of Christ Before the People. But they cannot be possessed by any instance of The Apparition -the original canvas (and perhaps certain very good copies of this canvas) -and, hence, cannot be manifested by an instance of this painting.
Or, consider Beethoven's Symphony No. 5. It can be plausibly argued that a complete appreciation of this symphony requires an experiential engagement with those experienceable properties that enable to grasp the history of its composition. At the same time, these properties cannot be possessed by any of the instances of Symphony No. 5 -particular musical performances -and, hence, cannot be manifested by an instance of this symphony.
Thus, the set of properties manifested by an instance of an artwork does not necessarily involve all the properties that must be experienced to fully appreciate this work. But what, then, determines whether a property that must be experienced to fully appreciate an artwork must be manifested by an instance of this work? Presumably, the right answer is this: What determines this is whether such a property falls under the category of properties through which the primary content 16 -that is, the set of 'those contentful properties that may be the ground of other contentful properties but which are not What Is an Instance of an Artwork?

14
This assumption as well as its generalized version -that knowing the history of an artwork's creation is required to fully appreciate this work -are a direct consequence of aesthetic contextualism -the view that 'artworks are essentially historically embedded objects, ones that have neither art status, nor determinate identity, nor clear aesthetic properties, nor definite aesthetic meanings, outside or apart from the generative contexts in which they arise and in which they are proferred' . Jerrold This is not to say, of course, that learning about the creative process leading up to an artwork -all the stops and starts, all the false turns, and so on -is always requisite to full appreciation. 16 The term 'content' is used here in a broad sense -to refer to the overall artistic content (and not just the semantic content). themselves grounded in contentful properties' 17 -of the work is articulated: If the property falls under this category, then it must be manifested by an instance; otherwise, the property does not have to be manifested by this instance.
So, in light of what has been said so far, an instance of an artwork has the following characteristic features: (a) it manifests certain properties of the work, (b) these properties are such that they must be experienced to fully appreciate the work, and (c) through these properties the primary content of the work is articulated. Now, with the help of these features -and assuming that the expression 'primary properties' denotes 'properties through which the primary content of an artwork is articulated' -'an instance of an artwork' can be defined as follows: Definition of 'an Instance of an Artwork' (hereafter: 'DI e '): For all x, x is an instance of an artwork if and only if x manifests certain primary properties that must be experienced to fully appreciate this work. 18 Note that by DI e , the only thing that matters for being an instance of an artwork is the manifestation of the relevant primary properties that must be experienced to fully appreciate this work: As long as an entity manifests such properties, it is an instance of the corresponding work. Thus, consider, for example, the original canvas of Kazimir Malevich's Black Square and its indiscernible counterpart that was created by someone completely unfamiliar with Malevich's works. 19 According to DI e , since both the canvas and the counterpart manifest the same properties, they are both instances of Black Square. Likewise, DI e entails that both a correct recitation of Robert Louis Stevenson's 'To Friends at Home' and its indiscernible counterpart produced by someone who has never encountered Stevenson's poetry (including 'To Friends at Home') manifest the same properties and so are instances of this poem. 20 17 Davies, 'Multiple Instances' , 411. Alternatively, the primary content of an artwork can be characterized as the set of the basic properties that determine the content of this work.  20 It is assumed that the original canvas of Black Square and the recitation of the poem 'To Friends at Home' manifest the relevant primary properties that must be experienced to fully appreciate the corresponding works. Does DI e , at least partially, reflect the existing meaning of the expression 'an instance of an artwork'? 21 A number of philosophers of art use this expression according to DI e with regard to at least some artworks. For example, Julian Dodd uses it this way when he talks about instances of musical works, and Nelson Goodman and Catherine Elgin use it this way with regard to instances of notational artworks (such as literary and musical works). 22 Given this, the question posed above can be answered in the affirmative. 23 However, although DI e reflects one of the existing senses of 'an instance of an artwork' , it does not reflect all of the senses of this expression there are. For, besides the sense of 'an instance of an artwork' specified by DI e -which might be termed 'epistemic' -in the relevant discourse (primarily, discourse concerning the ontology of artworks), there exists another sense of this expression, which DI e fails to capture. According to this latter sense -call it 'relational' -an instance of an artwork is not just what manifests certain primary properties that must be experienced to fully appreciate this work; rather, such an instance has an additional characteristic feature -that of standing in an appropriate historical-intentional relation to the work. 24 Now, in the relevant discourse, the relational sense is rather common: It is widely employed by a considerable number of philosophers, including Gregory Currie, Arthur C. Danto,Stephen Davies,Andrew Kania,Jerrold Levinson,Ted Nannicelli,and Richard Wollheim. 25 Given this, it is clear that to be complete, What Is an Instance of an Artwork? According to the terminology used in this essay, the fact that a definition reflects an existing sense of an expression does not entail that this definition does that correctly. Hence, one might ask: Does DI e reflect the corresponding sense of 'an instance of an artwork' correctly? An answer to this question is provided in Section III. 24 The nature of this relation is elucidated later in this section. 'Singular works must have one and only one instance. Multiple works specified through an exemplar must also have at least one instance, though they may have many more' (Davies,'Ontology of Art' ,160). 'Music is perhaps the art that presents the most philosophical puzzles. Unlike painting, its works often have multiple instances, none of which can be identified with the work itself' (Kania, 'Philosophy of Music'). any account of 'an instance of an artwork' must provide a definition of this expression used in the relational sense. What could such a definition be?
Given what has been said above, 'an instance of an artwork' in the relational sense (hereafter: 'an instance r of an artwork') has all the characteristic features of an instance of an artwork in the epistemic sense (hereafter: 'an instance e of an artwork') and, in addition to that, stands in an appropriate historicalintentional relation to the work. Taking this into account, 'an instance r of an artwork' can be defined as follows: It is worth stressing that unlike being an instance e of an artwork, being an instance r of an artwork requires not only manifesting certain primary properties that must be experienced to fully appreciate this work but also standing in an appropriate historical-intentional relation to this work. Thus, an indiscernible counterpart of Malevich's Black Square that was created by someone completely unfamiliar with Malevich's works is not an instance r of Black Square, since this counterpart does not stand in any appropriate historical-intentional relation to this painting. Similarly, an indiscernible counterpart of a recitation of Stevenson's 'To Friends at Home' produced by someone who has never encountered Stevenson's poetry (including 'To Friends at Home') is not an instance r of 'To Friends at Home' because this counterpart does not stand in any appropriate historical-intentional relation to this poem.
Regarding DI r a natural question arises: What exactly is the appropriate historical-intentional relation to an artwork in which an instance r of this work must stand? An answer to this question can vary depending on what kind of art Note that neither Davies nor Kania could use 'an instance of an artwork' in the epistemic sense here. If Davies did that, he would not be able to state that singular art kinds must have just one instance (as in this sense, an artwork can always have more than one instance). Likewise, if Kania used 'an instance of an artwork' in the epistemic sense, then it would be impossible for him to assume that there is only one instance in the case of a painting (for in this sense, a painting can have many instances). 26 Comp. Davies, 'Multiple Instances'; 'Enigmatic Variations'; 'Varying Impressions' , in particular the definition of 'a provenential instance of an artwork. ' Note a crucial difference between DI r and Davies's definition of 'a provenential instance of an artwork': Unlike the latter, DI r does not entail that something can be an instance of an artwork without being an instance e of this work. is under consideration. In the case of classical music, the historical-intentional relation to an artwork in which an instance r of this work stands is usually understood as the relation of being identical to a performance of this work generated with the help of either the original score (that is, the score directly created -say, written or typed -by the composer) or an entity that stands in the 'copy' relation to this score (where x stands in the 'copy' relation to y just in case x is a copy of y, or x is a copy of a copy of y, or x is a copy of a copy of a copy of y, and so on). In the case of photographic art, the historical-intentional relation to an artwork in which an instance r of this work stands is usually taken to be the relation of being identical to a print derived from the original photographic film or some other appropriate source (say, a photographic plate or bitmap) created by the author. In the case of painting, the historicalintentional relation to an artwork in which an instance r of this work stands is typically treated as the relation of being identical to the original canvas. 27 And in the case of literature, the consensus is that the historical-intentional relation to an artwork in which an instance r of this work stands is the relation of being identical to the work's original manuscript or an entity that stands in the 'copy' relation to this manuscript. 28 What Is an Instance of an Artwork? Assuming that the foregoing treatment is correct, the concepts of 'an instance r of a work of classical music' , 'an instance r of a photographic work' , 'an instance r of a painting' , and 'an instance r of a literary work' can be defined, respectively, as follows: For all x, x is an instance r of a work of classical music if and only if x (a) manifests certain primary properties that must be experienced to fully appreciate this work and (b) stands in the relation of being identical to the work's performance generated with the help of a score that is either the original score (that is, the score directly created by the composer) or an entity that stands in the 'copy' relation to this original score. For all x, x is an instance r of a photographic work if and only if x (a) manifests certain primary experienceable properties that must be experienced to fully appreciate this work and (b) stands in the relation of being identical to a print derived from the original photographic film or some other appropriate source (say, a photographic plate or bitmap) created by the work's author. For all x, x is an instance r of a painting if and only if x (a) manifests certain primary properties that must be experienced to fully appreciate this painting and (b) stands in the relation of being identical to the original canvas. For all x, x is an instance r of a literary work if and only if x (a) manifests certain primary properties that must be experienced to fully appreciate this work and (b) stands in the relation of being identical to the work's original manuscript or an entity that stands in the 'copy' relation to this manuscript.
It should be noted that the foregoing interpretations of the appropriate historical-intentional relation to an artwork in which an instance r of this work must stand are not claimed to be correct. In fact, although these interpretations are widely accepted, they may well be misguided. The interpretation provided in the case of classical music implies that an instance r of a work of classical music must be a performance. But why can't an instance of such a work be something other than a performance -say, a playing of a recording of a performance or a reproduction of a performance (that is, a particular sequence of sounds generated by some electronic device with the help of a musical score)? Similar questions arise with regard to the interpretations given in the case of painting and photographic art. According to the interpretation given in the case of painting, a painting has only one instance, namely the canvas. But why can't a painting be instanced not only by its canvas but also by something else -say, certain copies (for example, molecule-formolecule duplicates) of this canvas? The interpretation provided in the case of photographic art assumes that the only instances of photographic artworks are prints. But why can't photographic artworks be properly instanced by things other than prints -say, copies of prints? Finally, it can be questioned whether the interpretation given in the case of literary works is right in identifying instances of literary works with either original manuscripts or their copies. 29 (What has been said here is not intended to show that the mentioned interpretations are, in fact, misguided. The goal is to suggest that these interpretations could be misguided.)

III. EVALUATING THE DEFINITION
Let us now turn to the question of whether DI r/e is acceptable. As is generally agreed, to define x is to explicate the meaning of x by specifying a set of conditions that (a) are satisfied by all entities that actually fall under x and only by such entities and (b) are sufficiently informative -in particular, (1) are not enumerative (that is, do not amount to a disjunctive list of objects that fall under x), (2) do not involve a vicious circle (that is, do not explicitly or implicitly contain the concept being defined), and (3) do not contain meaningless expressions or expressions whose meanings cannot be understood. In light of this, Di r/e could potentially be rejected on the grounds that it is insufficiently informative or on the grounds that it does not cover all of those entities that fall under 'an instance of an artwork' and only such entities. But there is no real reason to question the sufficiency of the informativeness of DI r/e . One could 29 In 'What Instances of Novels Are' , I provide an argument showing that this interpretation is wrong in doing that.

Alexey Aliyev
Estetika: The Central European Journal of Aesthetics, LVI/XII, 2019, No. 1, 00-00 object, of course, that DI r/e is insufficiently informative because the meaning of the word 'certain' is unclear. But this objection can easily be defused by referring to the material of the next two sections -in particular, the definitions of 'a well-formed instance of an artwork' and 'a non-well-formed instance of an artwork' -where the sense of 'certain' is precisified. It could also be objected that DI r/e is insufficiently informative, since the definiens contains part of the concept being defined -namely, the concept of 'an artwork' . Yet this objection fails, as it is based on a false assumption -that the goal of DI r/e is to define 'an instance of an artwork' (that is, both of the concepts involved). In fact, however, its goal is to define 'an instance of an artwork' , or, in other words, 'an instance' as it is used in the context of being 'of' an artwork -in particular, (a) an artwork in general, or (b) some kind of artwork, such as a novel, a musical work, or a sculpture, or (c) some specific artwork, for example, the Mona Lisa or War and

Peace.
A more pressing objection to DI r/e draws upon the idea that it does not cover all of those entities that fall under 'an instance of an artwork' . Suppose some entity E makes available for experience the primary properties that must be experienced to fully appreciate some artwork A (where y makes x available for experience just in case y makes it possible to experientially engage with x either by directly perceiving x or by perceiving x as a result of applying a special skill, that is, a skill that is not acquired in a natural way -such as the skill of reading or the skill of playing a musical instrument). Suppose next that E does not manifest these properties. Then DI r/e entails that E is not an instance of A. Is this entailment true? Given the actual use of the expression 'an instance of an artwork' , (a) to be an instance e of an artwork, it is sufficient to make available for experience the primary properties that must be experienced to fully appreciate this work, and (b) to be an instance r of an artwork, it is sufficient (i) to make available for experience the primary properties that must be experienced to fully appreciate this work and (ii) to stand in an appropriate historicalintentional relation to this work. In light of this, E is an instance e of A. Moreover, assuming that E stands in an appropriate historical-intentional relation to A, E is an instance r of A. Thus, the foregoing entailment is false -and, as a result, DI r/e fails to cover all of those entities that fall under 'an instance of an artwork' . This objection is based on the thesis that the actual use of the expression 'an instance of an artwork' supports the account of 'an instance of an artwork' according to which (a) if x makes available for experience the primary properties that must be experienced to fully appreciate an artwork, then x is an instance e of this work, and (b) if x (i) makes available for experience the primary properties that must be experienced to fully appreciate an artwork and (ii) What Is an Instance of an Artwork? stands in an appropriate historical-intentional relation to it, then x is an instance r of this work. Is this thesis true? Suppose there is a score S of a work of classical music M such that by applying to S a particular special skill -namely, the skill of silent score reading -one can imagine, and, hence, experientially engage with, a performance of S that manifests the primary properties that must be experienced to fully appreciate M. 30 Then S makes available for experience the primary properties that must be experienced to fully appreciate M. Furthermore, it can be assumed that S stands in an appropriate historical-intentional relation to M. Given this, if the foregoing account of 'an instance of an artwork' is true, then S is an instance r/e of M. However, this result does not correspond to the actual use of the expression 'an instance of an artwork' , since according to this use, scores of works of classical music are not instances of these works (the consensus is that instances of works of classical music are solely musical performances). 31 Alternatively, that the foregoing account of 'an instance of an artwork' does not accord with the actual use of the expression 'an instance of an artwork' can be shown as follows. Suppose there is a verbal description V of some very simple drawing D -say, a drawing of a black square. Suppose also that by applying a special skill -namely, the skill of reading -to V, one can mentally form, and, hence, experientially engage with, an accurate image of D -an image that makes it possible to perceptually grasp the primary properties that must be experienced to fully appreciate D. Then V makes available for experience the primary properties that must be experienced to fully appreciate D. Furthermore, it can be assumed that V stands in an appropriate historical-intentional relation to B. Thus, if the definition of 'an instance of an artwork' being discussed is true, then V is an instance r/e of B. However, this does not accord with the actual use of the expression 'an instance of an artwork' , since according to this use, no verbal description of a drawing is an instance of this drawing.
Thus, the foregoing alternative account of 'an instance of an artwork' contradicts the actual use of the expression 'an instance of an artwork' . As a result, the thesis that the actual use of this expression supports this account is 30 Many musical scores are too complex for score reading. So S is not any musical score. 31 Thus, Levinson writes: 'Those familiar with recent reflection on the ontological question for works of art will know of the widespread consensus that a musical work is in fact a variety of abstract object -to wit, a structural type of kind. Instances of this type are to be found in the individual performances of the work' (Levinson,'What a Musical Work Is' ,(5)(6). In the same way, the consensus view on the nature of the entities that serve as instances of musical works is described in Kania, 'Philosophy of Music' . false. Meanwhile, if this thesis is false, then, since, as already mentioned, the objection being discussed is based on this thesis, this objection fails.
Are there any other potentially successful objections to the idea that DI r/e covers all and only those entities that fall under 'an instance of an artwork'?
No -or so it seems. As a result, given that, as already mentioned, there are no successful objections to the idea that DI r/e is sufficiently informative, there seems no reason to reject this definition. At the same time, there is reason to consider it satisfactory. First, as mentioned above, it satisfies one of the criteria of a successful definition -that of being sufficiently informative. Furthermore, there is a strong consideration in favour of the view that it also satisfies the second criterion -that of covering all and only those entities that fall under the expression being defined. As mentioned in Section II, many theorists use the expression 'an instance of an artwork' according to DI r , with regard to all artworks, while some use it according to DI e , with regard to at least some artworks. Thus, there is good reason to hold that DI r/e reflects the actual use of this expression. Meanwhile, if that is the case, then this definition covers all and only those entities that fall under 'an instance of an artwork' .

IV. DEFINING 'A WELL-FORMED INSTANCE OF AN ARTWORK'
Now that we have shown that DI r/e is acceptable, let us elaborate it further. Note that DI r/e does not specify whether an instance r/e of an artwork is capable of manifesting all the primary properties that must be experienced to fully appreciate this work. Can such an instance r/e (in principle) manifest all such properties? Apparently, the only plausible reason to answer 'No' is that there are no entities capable of manifesting all the primary properties that must be experienced to fully appreciate artworks. But this reason is unsatisfactory.
If there are no entities capable of manifesting all the primary properties that must be experienced to fully appreciate artworks, then no artwork can be fully appreciated. But this consequence is doubtless false. A complete appreciation of an artwork may be hard, but in an overwhelming majority of cases, it is, at least in principle, possible. (This is not to say, of course, that every artwork can be fully appreciated. First of all, a full appreciation is normally impossible in the case of artworks that no longer exist. Thus, we cannot fully appreciate the University of Vienna Ceiling Paintings by Gustav Klimt -Philosophy, Medicine, and Jurisprudence -which are believed to have been destroyed by retreating German SS forces in 1945 -as the available information about these paintings, including the information derived from the relevant photographs and preparatory sketches, is insufficient to enable us to do that. Furthermore, besides the case of complete destruction, it is often impossible to fully What Is an Instance of an Artwork? appreciate an artwork in another case -when the work has been considerably damaged. To illustrate, consider the original drawing by Willem de Kooning that forms the basis of Robert Rauschenberg's Erased de Kooning Drawing. This drawing could perhaps be characterized as existing, but it is too deformed to afford a complete appreciation, and the available information about the drawing is too scarce to make it possible for us to fully appreciate it.) So, there seems no real reason to think that an instance r/e of an artwork cannot manifest all the primary properties that must be experienced to fully appreciate this work. At the same time, there is a good reason to uphold the opposite thesis -that an instance r/e of an artwork can, in principle, manifest all of these properties. The reason is that according to the consensus among ontologists of art, for most (though not all) artworks, there, in fact, existed, exist now, or will exist instances r/e that manifest all the primary properties that must be experienced to fully appreciate these works. Thus, taking into account what has been said, it is reasonable to conclude that there can be instances r/e of an artwork that are capable of manifesting all the primary properties that must be experienced to fully appreciate this work.
Given the foregoing result, we are justified in adding to DI r/e the definition of 'an instance r/e that can manifest all the primary properties that must be experienced to fully appreciate the corresponding artwork' -or, in other words, the definition of 'a well-formed instance r/e of an artwork': 32 Definition of 'a Well-formed Instance e of an Artwork': For all x, x is a wellformed instance e of an artwork if and only if x manifests all the primary properties that must be experienced to fully appreciate this work.
Definition of 'a Well-formed Instance r of an Artwork': For all x, x is a wellformed instance r of an artwork if and only if x (a) manifests all the primary properties that must be experienced to fully appreciate this work and (b) stands in an appropriate historical-intentional relation to it.

V. DEFINING 'A NON-WELL-FORMED INSTANCE OF AN ARTWORK'
The only difference between well-formed instances r/e of an artwork and instances r/e of an artwork is that well-formed instances r/e manifest all the primary properties that must be experienced to fully appreciate the work, whereas instances r/e manifest certain primary properties that must be 32 This kind of instance r/e could also be characterized as 'strict ' , 'genuine' , 'perfect' , or 'ideal' . experienced to fully appreciate the work. In light of this, one might ask: Are the class of instances r/e and the class of well-formed instances r/e coextensive?
Put otherwise, are all instances r/e well-formed? Consider a slightly damaged print of a photograph or a musical performance that contains one incorrect note. Clearly, neither the performance nor the print provides access to all the primary properties that must be experienced to fully appreciate the corresponding works. But, at the same time, both the print and the performance provide access to a significant set of such properties. As a result, it seems reasonable to think that both of them can be (a) non-wellformed instances e and -assuming that each of them can stand in an appropriate historical-intentional relation to the corresponding work -(b) nonwell-formed instances r . And, in fact, most of those who employ the concept of 'an instance of an artwork' (primarily, ontologists of art) do think so. Given this, the above question, I think, should be answered in the negative.
In light of the fact that the class of instances r/e is not exhausted by wellformed instances r/e , a natural question arises: How can 'a non-well-formed instance r/e of an artwork' be defined? As is clear from what has been said in the previous paragraph, a non-well-formed instance r/e of an artwork manifests only some of the primary properties that must be experienced to fully appreciate this work. Can this fact alone be used to define 'a non-well-formed instance r/e of an artwork'? No -for there are entities that (a) manifest only some primary properties that must be experienced to fully appreciate an artwork but (b) are not instances r/e of this work. Consider, for example, a black and white image of a typical colour painting (such as Auguste Renoir's Luncheon of the Boating Party). Such an image cannot manifest any colour properties.
Meanwhile, such properties are doubtless among the primary properties that must be experienced to fully appreciate a typical colour painting. 33 So a black and white image cannot manifest all the primary properties that must be experienced to fully appreciate a typical colour painting. But such an image can manifest some of these properties -for example, those that are concerned with the shapes of what is depicted in this painting. Thus, a black and white image can manifest some of the properties that must be experienced to fully appreciate a typical colour painting. At the same time, as is generally agreed, no such image can be an instance r/e of such a painting.
So, what distinguishes non-well-formed instances r/e of artworks from entities that are not such instances? Taking into account the art-ontological context, What Is an Instance of an Artwork?

33
It should be underlined that here, we are talking about typical colour paintings, not all colour paintings, as there may (at least, in theory) exist colour paintings whose colour properties are not among the primary properties that must be experienced to fully appreciate these paintings. this question, I think, can be answered as follows: Unlike entities that are not instances r/e of artworks, non-well-formed instances r/e of artworks (a) manifest sufficiently many, though not all, of the primary properties that must be experienced to fully appreciate these works and (b) can, in principle, manifest all sensory kinds of such properties (where 'a sensory kind of property' denotes any kind of property that is relevantly concerned with a sensory modalityvisual properties [that is, properties related to vision], auditory properties [properties related to hearing], olfactory properties [properties related to olfaction], and so on). Given this, 'a non-well-formed instance of an artwork' can be defined in the following way: Definition of 'a Non-Well-Formed Instance e ': For all x, x is a non-well-formed instance e of an artwork just in case x (a) manifests sufficiently many, though not all, of the primary properties that must be experienced to fully appreciate this work and (b) could, in principle, manifest all sensory kinds of such properties.
Definition of 'a Non-Well-Formed Instance r ': For all x, x is a non-well-formed instance r of an artwork just in case x (a) manifests sufficiently many, though not all, of the primary properties that must be experienced to fully appreciate this work, (b) could, in principle, manifest all sensory kinds of such properties, and (c) stands in an appropriate historical-intentional relation to this work.
Note that one of the features that any non-well-formed instance r/e must possess is the feature of being, in principle, capable of manifesting all sensory kinds of the primary properties that must be experienced to fully appreciate the corresponding artwork. Thus, a performance of a musical work that contains an incorrect note but can, in principle, manifest all sensory kinds of the primary properties that must be experienced to fully appreciate this work can be a nonwell-formed instance r/e of this work. At the same time, given some plausible assumptions, a playing of an audio recording of a live performance of a work of classical music cannot be a non-well-formed instance r/e of at least some classical musical works. Such a playing, being non-visual, cannot, in principle, manifest any visual properties. 34 Meanwhile, the primary properties that must 34 Note that what is said here applies only to playings of 'audio only' (non-video) recordings. Perhaps playings of audio-video recordings -recordings that capture both the sonic and the visual aspects of a performance -can manifest all the primary properties that must be experienced to fully appreciate the corresponding works. For a discussion of this possibility, see Christy Mag Uidhir, 'Recordings as Performances' , British Journal of Aesthetics 47 (2007): 298-314. be experienced to fully appreciate a work of classical music often include certain visual properties. 35 So, in some cases, a playing of an audio recording of a live performance of a work of classical music cannot manifest all sensory kinds of the primary properties that must be experienced to fully appreciate this work.
Likewise, a soundless screening cannot be a non-well-formed instance r/e of a sound film. The primary properties that must be experienced to fully appreciate such a film include particular sonic properties. But no soundless screening can manifest any such properties. So, no such screening can manifest all sensory kinds of the primary properties that must be experienced to fully appreciate a sound film.
Here, one could object as follows: The foregoing definition implies that being, in principle, capable of manifesting all sensory kinds of the primary properties that must be experienced to fully appreciate an artwork is necessary to be a non-well-formed instance of this work. However, this implication is false.
There can be non-well-formed instances that can manifest only some of the sensory kinds of the mentioned properties.
Is this objection successful? It assumes that it is possible for a non-wellformed instance of an artwork to be incapable of manifesting all sensory kinds of the primary properties that must be experienced to fully appreciate this work. This assumption, however, is problematic. According to a widely accepted view, non-well-formed instances of artworks are slightly incorrect well-formed instances of these works. 36 Meanwhile, a slightly incorrect well-formed instance of an artwork is doubtless capable of manifesting all sensory kinds of the primary properties that must be experienced to fully appreciate this work.
So, if the above-mentioned view is true -and there seems no reason to think otherwise -non-well-formed instances of artworks must be, in principle, capable of manifesting all sensory kinds of the primary properties that must be experienced to fully appreciate these works.
What Is an Instance of an Artwork?

35
Here is another consideration against the assumption being discussed. If this assumption is true, then there must be non-well-formed instances that are, in principle, incapable of manifesting all the sensory kinds of the primary properties that must be experienced to fully appreciate the corresponding artworks. What entities could serve as such instances? Here are some possible candidates: (a) a silent screening of a sound film; (b) a playing of the sound of a film; (c) a silent performance of a sound play; (d) a purely sonic performance of a play; (e) a purely sonic performance of a musical; (f ) a silent performance of a musical.
But can (a)-(f ), in fact, serve as non-well-formed instances that are, in principle, incapable of manifesting all the sensory kinds of the primary properties that must be experienced to fully appreciate the corresponding artworks? To be such instances, (a)-(f ) must satisfy two conditions. First, they must be, in principle, incapable of manifesting at least one sensory kind of the primary properties that must be experienced to fully appreciate the corresponding artworks. Second, they must manifest sufficiently many, but not all, of these properties. There is no doubt that (a)-(f ) satisfy the first condition.
But do they satisfy the second one? Prima facie, (a)-(f ) can manifest sufficiently many of the primary properties that must be experienced to fully appreciate the corresponding artworks only if perceiving (a)-(f ) enables us to adequately (though, of course, not fully) appreciate these works. However, we cannot adequately appreciate a sound film solely by watching its silent screening or solely by listening to its sound; likewise, we are unable to adequately appreciate a sound play or a musical just by watching their silent performances or just by listening to the sound of their performances. Thus, there is good reason to hold that (a)-(f ) do not manifest sufficiently many primary properties that must be experienced to fully appreciate the corresponding artworks.
Meanwhile, if this is so, then (a)-(f ) cannot be non-well-formed instances. Now, what has been said about (a)-(f ) can, I think, be said about any other potential candidates for the role of non-well-formed instances that are, in principle, incapable of manifesting all the sensory kinds of the primary properties that must be experienced to fully appreciate the corresponding artworks. So, there are no non-well-formed instances that are, in principle, incapable of that -and, hence, the assumption being discussed is false.

VI. FINAL REMARKS
In closing, a few additional remarks: Remark 1: In the philosophical literature, there is an expression that is closely related to 'an instance of an artwork' -namely, 'a token of an artwork' . This latter expression could be defined as follows: For all x, x is a token of an artwork just in case (a) this work is a type (where a type can be understood as an entity that can have multiple instances r/e ) and (b) x is its instance r/e . Does M exist in this case? Prima facie, the answer is 'Yes' . M exists qua an entity that is, in some sense, 'contained' in that encoding. Thus, a musical work can exist even if there are no instances r/e of this work.
(2) If (a) there is an instance r of some artwork A and an instance r of some artwork B and (b) these instances (i) manifest the same primary properties that must be experienced to fully appreciate A and B and (ii) stand in the same historical-intentional relation to A, then A is identical to B.
Substantiation: Suppose there is an instance r of some artwork A and an instance r of some artwork B. Suppose next that these instances (i) manifest the same primary properties that must be experienced to fully appreciate A and B and (ii) stand in the same appropriate historical-intentional relation to A. Then, since nothing can stand in an appropriate historical-intentional relation to more than one artwork, both of them must be instances r of one and the same work. But if this is so, then A must be identical to B.
(3) The fact that (a) there is an instance e of some work A and an instance e of some work B and that (b) these instances manifest the same primary properties that must be experienced to fully appreciate A and B does not entail that A and B are identical.
What Is an Instance of an Artwork? 37 This does not imply, of course, that any artwork can exist if none of its instances r/e exist. According to a widely accepted view, a painting is identical to its only instance r/e -the original canvas. If this view is correct, then a painting cannot exist if no instances r/e of this painting exist. Substantiation: Consider Brahms's Piano Sonata Opus 2 (1852) and an (imaginary) 'work identical with it in sound structure, but written by Beethoven': Brahms's Piano Sonata Opus 2 (1852), an early work, is strongly Liszt-influenced, as any perceptive listener can discern. However, [the] work identical with it in sound structure, but written by Beethoven, could hardly have had the property of being Liszt-influenced. And it would have had a visionary quality that Brahms's piece does not have. 38 Given what has been said, the foregoing works are not identical. Suppose now that there are some instances, I 1 and I 2 , that manifest all the primary properties that must be experienced to fully appreciate Brahms's work. Then, since, by assumption, Beethoven's work is identical in its sound structure to Brahms's work, I 1 and I 2 also manifest all the primary properties that must be experienced to fully appreciate Beethoven's work. Thus, it is possible (a) for an instance e of an artwork A and an instance e of an artwork B to manifest the same primary properties that must be experienced to fully appreciate A and B and (b) for A and B to be non-identical.
(4) If (a) there is an instance r/e of some work A and an instance r/e of some work B and (b) these instances manifest non-identical sets of primary properties, then it cannot be inferred that A and B are non-identical. Substantiation: Suppose there is an instance r/e of A and an instance r/e of B.
Suppose next that the primary properties manifested by the instance r/e of A and the primary properties manifested by the instance r/e of B are not the same. Then, of course, these instances are not identical. But they can nevertheless be instances r/e of the same artwork, since each of them can manifest different sufficient sets of the primary properties of the same work and stand in an appropriate historical-intentional relation to this work. Thus, in this case, A and B are not necessarily non-identical.
Remark 3: Given what has been said in Sections IV and V, DI r/e can be formulated in a way other than the way it is formulated in Section II. As shown in Sections IV and V, an instance r/e of an artwork can be well-formed or non-well-formed. At the same time, there can be no instances r/e other than the well-formed and the non-well-formed. In light of this, as well as the definitions of 'a well-formed instance r/e of an artwork' and 'a non-well-formed instance r/e of an artwork' , 'an instance of an artwork' can alternatively be defined as follows: 38 Levinson, 'What a Musical Work Is' , 12. -a well-formed instance e of A -an entity that manifests all the primary properties that must be experienced to fully appreciate A; or -a non-well-formed instance e of A -an entity that (a) manifests sufficiently many, but not all, of the primary properties that must be experienced to fully appreciate A and (b) could, in principle, manifest all sensory kinds of these properties.
Definition of 'an Instance r of an Artwork': For all x, x is an instance r of some artwork A if and only if x is either: -a well-formed instance r of A -an entity that (a) manifests all the primary properties that must be experienced to fully appreciate A and (b) stands in an appropriate historical-intentional relation to A; or -a non-well-formed instance r of A -an entity that (a) manifests sufficiently many, but not all, of the primary properties that must be experienced to fully appreciate A, (b) could, in principle, manifest all sensory kinds of these properties, and (c) stands in an appropriate historical-intentional relation to A.
Clearly, the foregoing formulation provides a more detailed account of 'an instance of an artwork' than the formulation given in Section II. It should be underlined, however, that the former formulation does not differ extensionally from the latter one; that is, these formulations cover exactly the same set of entities.