Abstract Nelson Goodman notoriously distinguished between autographic works, whose instances should be identified by taking history of production into account, and allographic works, whose instances can be identified independently of history of production. Scholars such as Jerrold Levinson, Flint Schier, and Gregory Currie have criticized Goodman’s autographic/allographic distinction arguing that all works are such that their instances should be identified by taking history of production into account. I will address this objection by exploiting David Davies’ distinction between e-instances and p-instances of a work, thereby restating the autographic/allographic distinction. Then, I will show that this way of restating this distinction leads us to effective accounts of relevant related phenomena such as forgery, plagiarism, and digital technologies. Finally, I will show the theoretical advantages of my account of the autographic/allographic distinction in comparison with an alternative account recently proposed by Jason D’Cruz and P.D. Magnus. This paper aims at addressing an objection that scholars such as Jerrold Levinson, Flint Schier, and Gregory Currie have raised against Nelson Goodman’s distinction between autographic and allographic works.1 I will address this objection by drawing on David Davies’ distinction between e-instances and p-instances of a work. 2 I will show that this way of addressing this objection sheds some light on a variety of phenomena related to the autographic/allographic distinction. In Sections 1 and 2 I will introduce the autographic/allographic distinction and the e-instance/p-instance distinction. In Section 3 I will use the latter distinction in order to restate the former in a way that addresses the issues raised by its critics. In Sections 4 and 5 I will reconsider Goodman’s distinction between one-stage and two-stage works as well as his notion of a notation. In Sections 6 and 8 I will exploit the restatement of the autographic/allographic distinction in order to clarify the specificity of two-stage allographic works, the practices of forgery and plagiarism, and the ontological novelty introduced by digital technologies. In Section 9 I will compare my account of the autographic/allographic distinction with that of a competitor, namely Jason D’Cruz and P. D. Magnus’s account, arguing that the former can solve a problem that afflicts the latter.3 In Section 10 I will draw my conclusions. 1. The autographic/allographic distinction Goodman originally introduces the autographic/allographic distinction drawing on the practice of forgery: Let us speak of a work of art as autographic if and only if the distinction between original and forgery is significant; or better, if and only if even the most exact duplication of it does not thereby count as genuine. If a work of art is autographic, we may also call that art autographic. Thus painting is autographic, music nonautographic, or allographic.4 Here, I prefer to re-phrase the autographic/allographic distinction in a way that does not make reference to forgery. I will come back to the relation between this distinction and the notion of forgery in Section 7 thereby vindicating the logical connection that Goodman establishes between them. For now, I prefer to draw on what Goodman writes in a later work: What distinguishes an allographic work is that identification of an object or event as an instance of the work depends not at all upon how or when or by whom that object or event was produced.5 More specifically, I will focus on this formulation of the autographic/allographic distinction: (AA1) We are in front of a particular X and we wonder whether X is an instance of a certain work W. If the connection between the history of production of X and that of W matters in this respect, then W is an autographic work, otherwise W is an allographic work. As Schier puts it: . . . whether S is the Flagellation by Piero depends upon whether Piero had a hand in producing it, whereas whether P is a performance of the Symphonie Fantastique depends solely on its conformity to Berlioz’s score. Similarly, whether this marble object is Bernini’s David turns on Bernini’s role in shaping it, while a book is an instance of Joyce’s Ulysses so long as it is letter-for-letter identical to the original.6 Schier criticizes the autographic/allographic distinction contending that ‘aetiology is also important in allographic art. Whether P is an instance of the Symphonie Fantastique depends on its complying with and originating from Berlioz’s score’.7 In a similar vein, Levinson argues that all works are autographic since there is ‘no existing art forms in which historical factors are wholly irrelevant to the question of genuineness of work or instances’.8 Likewise, Currie argues that ‘Goodman’s distinction fails to distinguish any kind of art from any other kind’, since ‘being a correct instance of a work of any kind is always partly a matter of history of production’.9 In particular, the autographic/allographic distinction seems incapable of dealing with works that are perceptually indiscernible and yet distinct. Schier points this out drawing on Arthur Danto’s reflections on Jorge Luis Borges’ short story Pierre Menard, Author of the Quixote: if Menard had succeeded in his mad ambition, his work would not be Don Quixote by Cervantes, but rather the emanation of a peculiar symbolist poet of the early twentieth century. Similarly, it is logically possible that a musical performance which complies with a score of Berlioz’s isn’t a performance of a work by Berlioz: perhaps it is a performance of a work by a Twin Earth Berlioz.10 If this is right, all works are autographic since any work has identity conditions that depend not only on its manifest features but also on its history of production. As Richard Wollheim points out in his criticism of the autographic/allographic distinction ‘across the whole range of the arts, or for all works of art, history of production is essential’.11 Remember (AA1): if the connection between the history of production of a particular X and that of a work W bears upon whether X is an instance of W, then W is autographic, otherwise W is allographic. Yet, the argument from indiscernibility claims that the connection between the history of production of X and that of W always bears upon X being an instance of W, and thus the distinction between autographic and allographic works reveals itself to be unhelpful. It is nothing but the distinction between a predicate (i.e. autographic) that applies to all works and another (i.e. allographic) that applies to no work. I argue that this conclusion is too hasty. Even if one accepts that any work has identity conditions that depend on both its manifest features and its history of production, one is not forced to give up the autographic/allographic distinction. As I will argue in what follows, Davies’ distinction between e-instances and p-instances gives us a way to preserve the significance of the autographic/allographic distinction. 2. The e-instance/p-instance distinction Davies shows that the term ‘instance’ when used for works can be employed in two different senses, which he calls ‘e-instance’ and ‘p-instance’.12 An e-instance (epistemic instance) is a particular exhibiting the manifest properties of a work that are relevant for its appreciation, while a p-instance (provenential instance) is a particular that is appropriately connected to the history of production of that work. Being an e-instance depends on what a particular looks like, whereas being a p-instance depends on where a particular comes from. Furthermore, in Davies’ account, an e-instance is strict if it exhibits all the relevant manifest properties of the work, and a p-instance is strict if it is also a strict e-instance.13 Thus all strict p-instances are also strict e-instances, but there may be strict e-instances that are not p-instances. I will call the latter freestanding e-instances. These exactly reproduce the appearance of a work even though they are not properly connected to the history of production of that work. For example, a strict p-instance of Menard’s Don Quixote (henceforth understood according to Schier’s reading) is also a strict e-instance of Cervantes’s Don Quixote, even though it is not a strict p-instance of the latter. In sum, a strict p-instance of Menard’s Don Quixote is a freestanding e-instance of Cervantes’s Don Quixote; that is, an instance that exhibits all the manifest properties of that work that are relevant for its appreciation, even though it is not properly connected to the history of production of that work. 3. Restating the Autographic/Allographic Distinction The e-instance/p-instance distinction allows us to reformulate the autographic/allographic distinction in a way that can avoid objections arising from the argument from indiscernibility: (AA2) We are in front of a particular X and we wonder whether X is a strict e-instance of a certain work W. If the connection between the history of production of X and that of W matters in this respect, then W is an autographic work, otherwise W is an allographic work. In short, autographic works are such that they can have freestanding e-instances. We are facing an instance of Menard’s Don Quixote and we wonder whether this is a strict e-instance of Cervantes’ Don Quixote. We can safely give a positive answer without committing ourselves to the controversial claim that Menard’s Don Quixote and Cervantes’ Don Quixote are the same work. In treating these works as allographic we are just claiming that we can use a strict p-instance of Menard’s Don Quixote as a freestanding e-instance of Cervantes’ Don Quixote (or the other way round). We can do so provided that we associate the freestanding e-instance with the proper stock of information concerning the history of production of the relevant work.14 If we consider an alleged instance of an allographic work, the history of production of the instance can be inappropriate, provided that our information about the history of production of the work is appropriate. That is to say that the information concerning the history of production of the work can be wholly disentangled from the information concerning the history of production of its instances. I do not need to buy a copy of Cervantes’ Don Quixote and a copy of Menard’s Don Quixote. I can save money by just buying a copy of Menard’s Don Quixote and then using it also as an instance of Cervantes’ Don Quixote. I can read two books for the price of one. This is an advantage provided by allographic works. In order to turn my copy of Menard’s Don Quixote into an impeccable instance of Cervantes’ Don Quixote, I just have to couple this copy with the proper stock of information about the history of production of Cervantes’ Don Quixote. I can do so since this stock of information does not lie in the copy itself, but rather in shared memories and practices (for example, the entry ‘Cervantes’ in a reliable encyclopaedia). Autographic works do not allow us to do so. I cannot replace Piero’s Flagellation or Bernini’s David with some doppelgangers to whom I have associated the proper stock of information concerning the history of production of Piero’s or Bernini’s masterpieces. This would not count as a proper appreciation of those masterpieces, whose appearance cannot be wholly disentangled from their history of production. The appearance carries some bits of history that no encyclopaedia entry could contain. Goodman clearly makes this point, as we will see by examining his discussion of two-stage autographic works in what follows. 4. The One-Stage/Two-Stage Distinction The connection between appearance and history applies not only to what Goodman calls one-stage autographic works such as paintings or marble sculptures. This applies also to what Goodman calls two-stage autographic works such as cast sculptures, art prints, photographs or films. The distinction between one-stage and two-stage works can be figured out in the following terms. A one-stage work is such that a beholder can directly appreciate what has been produced by a maker, whereas a two-stage work requires the mediation of a process (e.g. screening in cinema, performance in theatre) that makes the work accessible to the beholder. As Goodman puts it: One notable difference between painting and music is that the composer’s work is done when he has written the score, even though the performances are the end-products, while the painter has to finish the picture. No matter how many studies or revisions are made in either case, painting is in this sense a one-stage and music a two-stage art.15 The one-stage/two-stage distinction is orthogonal to the autographic/allographic distinction. Works of literature and works of theatre are both allographic, and yet the former are one-stage whereas the latter are two-stage. Likewise, marble sculptures and cast sculptures are both autographic, and yet the former are one-stage whereas the latter are two-stage. Two-stage autographic works differ from one-stage autographic works in that they make room for multiple instances, as far as the process that makes the work accessible to the beholder is a repeatable process. Nevertheless, such instances can count as strict e-instances of a certain work only if they are also p-instances of that work. In other words, two-stage autographic works differs from one-stage autographic works in that they make room for multiple instances, but not for freestanding e-instances. For what concerns the latter, two-stage autographic works behave just like one-stage autographic works. If you want to properly appreciate a two-stage autographic work you should pay attention to an instance having the proper causal history, namely a strict p-instance. This is the only way at your disposal to warrant that this instance exhibits all its relevant manifest properties. Autographic works, indeed, are such that their manifest features cannot be disentangled from their history of production. Even if an autographic work is two-stage and thus can have multiple instances, the latter can be strict e-instances only if they are properly placed on the causal path that springs from the creation of that work. Why? Because, as Goodman puts it, autographic works are such that minute discrepancies may always go unnoticed; and there is no basis for ruling out any of them as inessential. The only way of ascertaining whether a print is genuine is by finding out whether it was taken from a certain plate.16 Autographic works are such that we cannot disentangle the totality of their relevant manifest features from their history of production. We can do so only if a work is allographic. In autographic works, the manifest features are drenched with the history of production. Therefore, the only way to warrant the appreciation of all the relevant manifest features of a certain autographic work is to pay attention to an instance having the proper history of production. 5. The Role of Notation What makes a work allographic? What gives it the capacity to have freestanding e-instances? From Goodman’s perspective, there is a clear answer to this question: the availability of a notation, that is, a system of symbols and rules whose ‘primary function’ consists in disentangling the appearance of a work from its history of production.17 I will call this the notational function. More specifically, given an alleged strict e-instance X of W, a notation allows one to establish whether X really is a strict e-instance of W regardless of any knowledge concerning the connection between the history of production of X and that of W. For example, the linguistic notation allows one to establish whether an alleged strict e-instance X of Ulysses really is what it alleges to be simply by comparing X with a strict p-instance of Ulysses. In the fourth chapter of Language of Art (‘The Theory of Notation’), Goodman builds up a formal account of notation, which rests upon syntactic and semantic requirements (namely: disjointedness, differentiation, and non-ambiguity). Yet, for the autographic/allographic distinction, which Goodman discusses in the third chapter of his book (‘Art and Authenticity’), a functional account of notation is sufficient. We can content ourselves with characterizing a notation as a system of symbols and rules that fulfils the notational function, that is, the function of allowing us to assess alleged strict e-instances of a work regardless of any knowledge concerning their connection to the history of production of that work. From this perspective, Goodman’s syntactic and semantic requirements for a notation can be seen as nothing but a formal way of implementing the notational function. 6. The Specificity of Two-Stage Allographic Works The capacity of a notation to allow one to assess alleged strict e-instances of a work regardless of history of production is evident in the case of one-stage allographic works such as poems or novels. Yet, things seem to be more complex in the case of two-stage allographic works such as musical works, since in this case the alleged strict e-instances to be assessed are performances, not mere sequences of symbols to whom the formal rules constituting the notation can be easily applied. In this sense, assessing an alleged strict e-instance of Joyce’s Ulysses is much easier than assessing an alleged strict e-instance of Mozart’s Jupiter Symphony. The latter task requires a specific musical competence that allows one to compare the performance with the score, whereas the former task requires nothing but a comparison of two series of homogeneous symbols (namely, those constituting the alleged strict e-instance of Ulysses to be assessed and those constituting the strict p-instance of Ulysses, which plays the role of yardstick). Still, the complication that characterizes two-stage allographic works does not prevent them from being genuinely allographic. Although assessing an alleged strict e-instance of a two-stage allographic work requires a specific competence, such an assessment remains independent of knowledge concerning the connection between the history of production of the alleged strict e-instance and that of the work. In order to assess whether a performance P counts as a strict e-instance of a musical work W having S as its score, one should just check whether the sequence of sounds constituting P complies with the sequence of symbols constituting S. The connection between the history of production of P and that of W does not bear upon this assessment. Even though P has been produced by means of a score S* that is completely unrelated to W, if S* is notationally identical to S and P is a correct performance of S*, then P is a strict e-instance of W in spite of lacking the proper historical connection to W. In this sense, if one can trust the performers’ capacity to correctly play the score, the easiest way to assess P as an alleged strict e-instance of W consists in comparing the score S* used by the performers with a correct score S of W. Here is a way of bringing the procedure for assessing instances of two-stage allographic works back to that for assessing instances of one-stage allographic works. Comparing two musical scores is exactly the same kind of procedure as comparing two literary texts. The only difference lies in the fact that in the case of two-stage allographic works such as symphonies one should trust the performers whereas in the case of one-stage allographic works such as novels trust plays no role. More generally, a two-stage allographic work involves symbols that are specific to the performance and therefore different from the symbols involved by the score. In the case of music, the score involves written notes whereas the performance involves sounds. In the case of theatre, the score involves inscriptions whereas the performance involves utterances and other actions and events. The most straightforward way to check the correctness of a performance (i.e. to check whether that performance is a strict e-instance of the work) consists in ‘translating’ the performance symbols into the corresponding score symbols (or vice versa), thereby comparing the translated performance with the original score. This operation requires some ‘translation competence’ that can be more or less difficult to acquire depending on both the cultural context and the specificity of the symbols. For example, the translation competence required by theatre (from utterances to inscriptions) seems to be much easier to acquire than the translation competence required by music (from sounds to notes)—at least in the context of our culture, in which literacy is much more widespread than musical education. However, even when one lacks the relevant translation competence, one can still assess an alleged strict e-instance of a two-stage allographic work by limiting oneself to compare the original score with that used by the performers, provided that one trusts the performers. In sum, trusting the performers is not necessary to assess the correctness of performances of two-stage allographic works, but it becomes necessary if one lacks the relevant translation competence.18 7. Forgery and Plagiarism In the wake of Goodman, the distinction between autographic and allographic works can be expressed by claiming that only autographic works can be forged.19 Schier developed this point claiming that while only autographic works can be forged, only allographic works can be plagiarized.20 Davies’ distinction between e-instances and p-instance allows us to clarify the relationship between the autographic/allographic distinction and the practices of forgery and plagiarism. We can conceive of a perfect forgery as a particular to which a fake history of production is ascribed with the aim of treating it as a strict e-instance of a certain work. In this sense, there may be not only forgeries of one-stage autographic works such as paintings but also forgeries of two-stage autographic works such as photographs. A forgery of Piero’s Flagellation is a painting X that pretends to be a strict p-instance of Flagellation even if Piero did not have a hand in producing X. Likewise, a forgery of Robert Capa’s The Falling Soldier is a photograph X that pretends to be a strict e-instance of The Falling Soldier even if X is not properly connected to The Falling Soldier’s history of production. For example, X may be a photograph I took of actors playing the scene portrayed by The Falling Soldier. Still, it does not make sense to forge an allographic work W since in order to ascertain whether a particular X is a strict e-instance of W the history of production is not relevant. There is no reason why to ascribe a fake history of production to a text X in order to turn it into a strict e-instance of Joyce’s Ulysses, since the history of production of X plays no role in X being a strict e-instance of Ulysses. Thus, we can characterize perfect plagiarism as the reverse of perfect forgery. The latter consists in pretending that a particular X is a strict e-instance of a work W even though, in fact, X is not a strict p-instance of W. Conversely, perfect plagiarism consists in pretending that a particular X is not a strict p-instance of a work W even though, in fact, X is a strict e-instance of W. For example, a perfect plagiarism of Joyce’s Ulysses is a text that is identical to the text of Ulysses and yet is pretended to have a history of production that traces back to an author who is not Joyce. From this perspective, the imaginary Don Quixote written by Menard can be seen as an attempt to turn plagiarism into a new form of art. As it does not make sense to make a perfect forgery of an allographic work, so it does not make sense to make a perfect plagiarism of an autographic work. In fact, one cannot plausibly claim that a particular X is not a strict p-instance of an autographic work W (i.e. X is not properly connected to W’s actual history) even though, in fact, X is a strict e-instance of W (i.e. X exhibits W’s actual appearance). That is because the only way in which one can show that X is a strict e-instance of W consists in showing that X is a strict p-instance of W. Of course, I can show you a strict e-instance of The Falling Soldier saying that I am the true author of this photograph. Yet, if I want to show you that this really is a strict e-instance of The Falling Soldier, I should reveal its real history of production, which traces it back to Capa’s creation, and this would inevitably defeat my claim. 8. The Digital Extension of the Allographic Domain Restating the autographic/allographic distinction in terms of the e-instance/p-instance distinction allows us to shed some light on the novelty introduced by digital technologies in the ontology of works such as photographs and films. In fact, digital technologies introduce notations that move these works from the autographic domain to the allographic one. As showed above, analogue photographs and films are two-stage autographic works. They can have multiple instances but in order to ascertain whether an instance is a strict e-instance one should take the history of production into account. For example, the only way of ascertaining whether a certain screening of Orson Welles’ film Citizen Kane is genuine is by finding out whether it was taken from a negative that is properly connected to Orson Welles’ creation. Still, digital technologies overcome this dependence of assessment of instances on history of production. For example, digital technologies turn a picture into a digital score constituted by a series of digital symbols. John Zeimbekis has argued that such a score does not satisfy all the formal requirements that Goodman imposed on notations.21 Yet, what matters, here, is just that the digital notation fulfils the notational function, which allows one to assess an alleged strict e-instance X of a work W, regardless of any knowledge concerning the connection between the history of production of X and that of W. Given a digital picture W, one can assess whether an alleged strict e-instance X of W really is what it alleges to be, regardless of any knowledge concerning history of production. One can do so simply by comparing digital scores. Let us suppose that the picture W has a digital score S whereas the alleged e-instance X has a digital score S*. A comparison between S and S* is all we need in order to establish whether X is a strict e-instance of W. If S is identical with S*, then X is a strict e-instance of W, otherwise it is not. That is all. The connection between the history of production of W and that of X plays no role in this assessment. Thus, a digital picture is a two-stage allographic work just as a musical work like Mozart’s Jupiter Symphony. As seen above, assessing an alleged e-instance X of the Jupiter Symphony, once we have checked or trusted the orchestra’s performance, consist in nothing but comparing the musical score of X with the musical score of the Jupiter Symphony. Likewise, assessing an alleged e-instance X of a digital picture W, once we have checked or trusted the displaying mechanism, consists in nothing but comparing the digital score of X with the digital score of W. The coming of digital technologies introduces an ontological divide in practices such as photography, cinema and recorded music. An analogue photograph, for example, is a two-stage autographic work whose alleged strict e-instances can be assessed only by considering their connection to the history of production of the photograph itself. By contrast, a digital photograph is a two-stage allographic work whose alleged strict e-instances can be assessed simply by comparing their digital score to the digital score of that photograph, regardless of any connection to the history of production of that photograph. It is worth noting that the coming of digital technologies does not automatically turn pre-existing analogue photographs into allographic works. Only brand-new digital photographs are automatically placed into the ontological realm of two-stage allographic works. Indeed, from an ontological point of view, the digitization of analogue works is a quite complex matter. We may use a strict e-instance of an analogue photograph, say, Robert Capa’s The Falling Soldier, in order to produce a digital copy D of it with the best technologies at our disposal. Let us call S the digital score that constitutes D. Has this procedure turned The Falling Soldier into an allographic work? Can we, henceforth, just assess an alleged strict e-instance X of The Falling Soldier by comparing its digital score with S regardless of any knowledge concerning the connection between the history of making of X and that of The Falling Soldier? The answer to this ontological question has relevant practical consequences. If we have a way of turning analogue works into digital ones, there do not seem to be any reasons (at least for what concerns the appreciation of the work) to expend energy preserving the original analogue templates (e.g. the negatives of Capa’s The Falling Soldier). Still, the answer is not obvious. One might argue, in a Goodmanian vein, that the only strict e-instances of The Falling Soldier are those analogue prints that are in the proper causal connection to the original negative, since ‘minute discrepancies may always go unnoticed; and there is no basis for ruling out any of them as inessential’.22 If this is right, the above-mentioned digital copy D does not count as a strict e-instance of The Falling Soldier, and neither do the other alleged digital strict e-instances whose score is identical to that of D. At most, D can be treated as the best digital surrogate of The Falling Soldier, that is, the best allographic approximation of an irredeemably autographic work. This seems to be the main reason why the preservation of material templates is still relevant to the appreciation of works such as photographs, sounds recordings and films that were originally produced as analogue works. 9. A Competitor In their controversy with Zeimbekis on the ontology of digital images, D’Cruz and Magnus provide an account of the autographic/allography distinction which has some affinities to the one I have proposed in this paper. But, as I will show in what follows, there is also a crucial difference. D’Cruz and Magnus rely upon Joseph Moore’s distinction between a S-work, ‘which is determined by having a specific structure’, and a P-work, ‘which is determined by having a specific provenance’.23 According to Moore, ‘P-works supervene upon particular actions, including particular psychological and social events’.24 In this sense, ‘we can usefully regard P-works as “tradition-threads”—that is, as individual threads in the broader causal tapestry of musical, psychological, and social events upon which a musical culture and tradition supervene’.25 D’Cruz and Magnus characterize the distinction between allography and autography in the terms of a distinction between works that exist both as S-works and as P-works, and works that only exist as P-works: We can say that an art form is ‘allographic’ if works of that form have, as a matter of ontology rather than as a matter of aesthetics, both a corresponding S-work and a corresponding P-work. If we take this as a definition, then we should say that a form is ‘autographic’ if and only if works of that form exist only as P-works.26 Therefore, music and literature are allographic since their works exist both as P-works and as S-works, whereas painting and artistic print-making are autographic since their works only exist as P-works. One might relate Moore’s distinction between S-works and P-works to Davies’ distinction between e-instance and p-instances. This can be done by conceiving of the S-work as the formal structure that is instantiated by strict e-instances, and of the P-work as the ‘tradition-thread’ to whom p-instances belong in virtue of the proper causal connection. Yet, it is worth noting that Davies’ basically is a pragmatic distinction whereas Moore’s is a fully-fledged metaphysical distinction. According to Davies, a strict e-instance is a particular that we can use in order to properly appreciate the manifest features of a work, without any ontological commitment on whether that work exists both as tradition-thread and as a formal structure. In this sense, the fact that a work can have freestanding e-instances only shows that there is a formal structure whose instances can be used to properly appreciate the manifest features of that work, but this does not entail that the work exists both as a tradition-thread and as a formal structure. Davies’ distinction between e-instances and p-instances has no specific commitment on the metaphysical way in which works exist. Instead, Moore’s distinction leads to works such as piano sonatas existing both as tradition-threads and as formal structures: there are two entities that answer our talk of one musical work ... . It’s surely surprising, then, to learn that when we count the thirty-two individual musical works that comprise Beethoven’s complete piano sonatas, we traffic in sixty-four distinct entities!27 This difference has important consequences for what concerns the criterion that specifies the autographic/allographic distinction. Davies’ distinction between e-instances and p-instance leads one to a Pragmatic Criterion (PC), which specifies the autographic/allographic distinction (AA2) that I have introduced above: (PC) Might we use a particular that lacks the proper causal connection to the history of production of a certain work W in order to properly appreciate the manifest features of W? (I.e. might we use such a particular, coupled with the proper information about W’s history, in order to properly appreciate W?) If yes, W is allographic; if no, W is autographic. Instead, Moore’s distinction between S-works and P-works leads one to a Metaphysical Criterion (MC), which corresponds to the autographic/allographic distinction proposed by D’Cruz and Magnus: (MC) Does a work W exist both as S-work and as a P-work? If yes, W is allographic; if no, W is autographic. In many cases, (PC) and (MC) give the same answer and it might me tempting to conclude that they are just two different ways of articulating the same criterion. But one should resist this temptation, since there is at least one case in which the two criteria can give different answers. This is the case of digital photographs.28 According to (MC), whether digital photographs are allographic like the other digital images or exceptionally autographic depends on the metaphysical conception of digital photography that one endorses. If one holds that ‘the digital photograph is essentially historical’ (since it essentially has the modal property of counterfactual dependence on the subject portrayed) then ‘it exists as a P-work but not as an S-work. It is therefore, by our criterion, autographic’.29 Conversely, if one denies that the digital photograph is essentially historical, then: ‘For a digital photograph, the formal features are fully specified by the pixels. So there is a corresponding S-work, and digital photography is allographic.’30 By contrast, (PC) gives us an answer that is independent of the metaphysical conception of digital photography that one endorses. That is, digital photographs are allographic regardless of whether the digital photograph is essentially historical or not. Even if one endorses a metaphysical conception of digital photography according to which a digital photograph essentially is counterfactually dependent on the subject portrayed, this photograph remains allographic because we can still use a freestanding e-instance of it in order to properly appreciate its manifest features (i.e. we can still use a freestanding e-instance, coupled with the relevant information about history of production, in order to properly appreciate that photograph). It is worth stressing that, following Davies, I conceive of an e-instance as a particular that allows one to appreciate the manifest features of a work regardless of whether this particular is properly connected to the history of production of the work: ‘E-instances, we saw, are defined in terms of a particular role that they play in appreciation in virtue of their manifest properties alone’.31 Thus, whether a particular is an e-instance of a digital photograph or not only depends on how we can use this particular. It does not depend on whether counterfactual dependency is essential to digital photography or not. Digital photographs still have freestanding e-instances even if they exist only as P-works. Certainly, a freestanding e-instance of a digital photograph, unlike a p-instance of it, is not counterfactually dependent on the subject portrayed. If something different had been in front of the camera, the p-instance would have been different in turn, whereas the freestanding e-instance would have remained the same. However, this does not prevent us, as denizens of the actual world, from using a freestanding e-instance of a digital photograph in order to properly appreciate the manifest features of that photograph – and also in order to fully appreciate that photograph, provided that we have the proper information about its history (indeed, information that we also need in order to properly appreciate that photograph when faced with a p-instance of it). In sum, the lack of counterfactual dependency does not bear upon the possibility of using a freestanding e-instance in order to properly appreciate a digital photograph, even if one thinks that counterfactual dependency is essential to digital photography. One can use a freestanding e-instance in order to properly appreciate a digital photograph and yet keep insisting that a digital photograph (qua photograph) must carry information about what was in front of the camera. This is a theoretical advantage of (PC) in comparison with (MC), since the former, unlike the latter, allows us to disentangle the autographic/allographic distinction from metaphysical controversies on the nature of digital photography. In virtue of (PC), people can share the same autographic/allographic distinction even though they have different views on whether counterfactual dependency is essential to digital photography. Arguably, the case of digital photographs is not the only one that raises such modal issues. Consider a work of history as, for example, Thucydides’ History of the Peloponnesian War. One might argue that this work counterfactually depends on the events it portrays just as a photograph does. If these events had been different, the sequence of words that constitute this work would have been different in turn. In such a counterfactual scenario, a p-instance of the History of the Peloponnesian War would have been different in turn, whereas a freestanding e-instance would have remained the same. According to (MC), this would entail that the History of the Peloponnesian War is ‘essentially historical’ (in fact, it is a work of history), therefore it exists as a P-work but not as an S-work, hence it is autographic. If this is right, (MC) makes the autographic/allographic distinction vary depending on metaphysical preferences not only in the domain of digital images but also in the domain of literary works. If one thinks that counterfactual dependency is essential to works of history just as it is to digital photographs, then one is forced to acknowledge that in the domain of literary works some works (i.e. works of history) are autographic whereas others are allographic, just as in the domain of digital pictures some works (i.e. digital photographs) are autographic whereas others are allographic. (PC) effectively avoids such an awkward partition by focusing on the use of particulars as appropriate instances instead of on the metaphysical nature of works. Even though one thinks that works of history are essentially historical (there seem to be good reasons to think so), and therefore they only exist as P-works but not as S-works, one can still use a freestanding e-instance of the History of the Peloponnesian War as a substitute of a strict p-instance in order to properly read this work. Of course, if the events of the Peloponnesian War had been different, the p-instance would have been different in turn, whereas the freestanding e-instance would have remained the same. Yet, this modal truth does not affect the autographic/allographic distinction, since (PC) allows us to establish whether a certain work is autographic or allographic regardless of whether one thinks that modal properties are essential or not to that work (i.e. whether one thinks that this work can exist only as a P-work or also as an S-work). Finally, the dependence of the autographic/allographic distinction on metaphysical conceptions of works of art seems to raise a problem for (MC) also in the case of paintings. D’Cruz and Magnus argue that paintings exist only as P-works: ‘there is no way to elaborate a sense of S-painting’, since ‘there is no separate S-work apart from the individual object in all its particularity’.32 Yet, one might state that a painting exists both as a P-work and as an S-work. Wollheim suggests such a conception of paintings when he writes: in its determinate properties the physical object changes over time, and it is to be explained by the fact that pigment, stone, and wood are eminently corruptible: colour fades, damp loosens the plaster, the atmosphere erodes the carving. But, by contrast, the work of art itself is incorruptible: its character does not alter with time, and it has no history … the aesthetic properties of the physical object change over time whereas those of the work of art do not.33 What Wollheim calls ‘the physical object’ corresponds to the P-work whereas what he calls ‘the work of art’ corresponds to the S-work. From this perspective, our epistemic difficulties in grasping the S-work that corresponds to a certain painting should not be confused with the metaphysical fact that this painting does have a separate S-work apart from the individual object in all its particularity. In this sense, the practice of painting restoration might be described as an attempt to bring as much as possible the physical object, namely the P-work, back to the S-work with which it does no longer comply. I am not arguing that such a metaphysical conception of paintings is right; I am just suggesting that it is not completely implausible and that one could endorse it. Yet, if one did so, according to (MC) one would be forced to state that paintings are allographic, and this surely is an awkward consequence. Instead, (PC) avoids this consequence. Even if one thinks that paintings exist both as S-works and as P-works, one can acknowledge that we cannot actually use a particular that lacks the proper causal connection to the history of production of a certain painting in order to properly appreciate the manifest features of that painting. Indeed, we could do so only if one day we will discover a notation enabling us to represent and replicate all the manifest features of a certain painting. For example, one might conceive of the ‘super Xeroxing machine’, which has been hypothesized by Currie, as a machine that exploits a (chemical-geometrical) notation in order to produces a molecule-for-molecule duplicate of the surface of a certain painting.34 In this case, just as (PC) states, paintings would become allographic because one would have the possibility to use a particular that lacks the proper causal connection to the history of production of a certain painting (but complies with the structure fixed by the notation) in order to properly appreciate the manifest features of that painting.35 In this sense, the super Xeroxing machine would turn painting into an allographic art in a way similar to that in which digital technologies have turned cinema into an allographic art (see Section 8 above). 10. Conclusion Towards the end of the third chapter of Languages of Art, Goodman hypothesizes that ‘initially, perhaps, all arts are autographic’, but then some of them become allographic when one has found a way to disentangle the appearance of a work from its history of production, thereby cutting the umbilical cord that tied this appearance to the individual who created the work and to the time at which it was created.36 As Goodman puts it, ‘a notation may be devised in order to transcend the limitations of time and the individual’.37 Written language allows us to ‘transcend the limitations of time and the individual’ for what concerns literary works, and musical notation allows us to do so for what concerns musical works. Likewise, the digital notation allows us to ‘transcend the limitations of time and the individual’ for what concerns works such as photographs, sound recordings, and films. Ultimately, revisiting the autographic/allographic distinction allows us to rethink Walter Benjamin’s famous claim that mechanical reproduction deprives works of art of their ‘aura’.38 Taking a Goodmanian reading of Benjamin’s statement, I understand the aura as the connection between the appearance of a work and its history of production.39 In one-stage autographic forms of art such as painting or marble sculpture, the aura is completely preserved since the appearance of the work is inextricable from the individual object that the artist created. In this sense Benjamin compares the individuality of a work of this sort to that of a person.40 In two-stage autographic arts such as photography and film, there may be multiple instances of a certain work, and from Benjamin’s perspective this seems enough to claim that works of this sort have lost their aura. Yet, if one applies Goodman’s distinction between two-stage autographic works and two-stage allographic works, one can notice that, the spectrum goes from pure aura to total mechanical reproduction, analogue photographs and films (as well as analogue sound recordings) are not at the end but rather in the middle. Even though these analogue works make room for multiple instances, they still preserve a connection between the history of production on the instances and that of the work. As argued above, one can assess an alleged strict e-instance of a two-stage autographic work only by relating the history of production of this instance to that of the work. In this sense, two-stage autographic works still preserve some bits of aura, inasmuch as the assessment of the appearance of their instances cannot be disentangled from knowledge concerning history of production. The end of the spectrum, namely the pure mechanical reproduction, is reached only by allographic arts, in which the aura, understood as the connection between appearance and history, no longer plays a role in assessing the appearance of instances. Thus, digital technologies deprive photographs, sounds recordings and films of those bits of aura that they still enjoyed in the analogue age. From this perspective, what Benjamin calls ‘the loss of aura’ is the price to pay for having works that can wholly transcend what Goodman calls ‘the limitations of time and the individual’. Footnotes 1 Jerrold Levinson, ‘Autographic and Allographic Art Revisited’, Philosophical Studies 38 (1980), 367–383; Flint Schier, Deeper into Pictures (Cambridge: Cambridge University Press, 1986); Gregory Currie, An Ontology of Art (London: Macmillan, 1989); Nelson Goodman, Languages of Art: An Approach to a Theory of Symbols (Indianapolis, IN: Bobbs-Merrill, 1968). 2 David Davies, ‘Multiple Instances and Multiple “Instances”’, BJA 50 (2010), 411–426; ‘Enigmatic Variations’, The Monist 95 (2012), 644–663. 3 Jason D’Cruz and P. D. Magnus, ‘Are Digital Images Allographic?’, Journal of Aesthetics and Art Criticism 72 (2014), 417–427. 4 Goodman, Languages of Art, 113. 5 Nelson Goodman, Of Mind and Other Matters (Cambridge, MA: Harvard University Press, 1984), 149. 6 Schier, Deeper into Pictures, 27. 7 Ibid., 28. 8 Levinson, ‘Autographic and Allographic Art Revisited’, 375. 9 Currie, An Ontology of Art, 15; 124 10 Schier, Deeper into Pictures, 28; Jorge Luis Borges, ‘Pierre Menard, Author of the Quixote’, in Ficciones (New York: Grove, 1962), 45–56; Arthur Danto, The Transfiguration of the Commonplace: A Philosophy of Art (Cambridge, MA: Harvard University Press, 1981), 34–35. 11 Richard Wollheim, Art and its Objects (2nd rev. edn) (Cambridge: Cambridge University Press, 1980), 114. 12 Davies, ‘Multiple Instances and Multiple “Instances”’, 412. 13 Davies, ‘Enigmatic Variations’, 647. 14 Davies, ‘Multiple Instances and Multiple “Instances”’, 419. 15 Goodman, Languages of Art, 113–114. 16 Ibid., 119. 17 Ibid., 128. 18 Thanks to an anonymous referee for leading me to figure out the different roles of competence and trust in assessing the correctness of performances of two-stage allographic works. 19 Goodman, Languages of Art , 113. 20 Schier, Deeper into Pictures, 29. 21 John Zeimbekis, ‘Digital Pictures, Sampling, and Vagueness: The Ontology of Digital Pictures’, Journal of Aesthetics and Art Criticism 70 (2012), 43–53. Zeimbekis’s argument has been discussed in the following works: Jason D’Cruz and P. D. Magnus, ‘Are Digital Images Allographic?’; Katherine Thomson-Jones, ‘The Philosophy of Digital Art’, in Edward N. Zalta (ed.), The Stanford Encyclopedia of Philosophy (Spring 2015 edn), <https://plato.stanford.edu/archives/spr2015/entries/digital-art/> accessed 27 Nov 2017; John Zeimbekis, ‘Why Digital Pictures are not Notational Representations’, The Journal of Aesthetics and Art Criticism 73 (2015), 449–453; Jason D’Cruz and P.D. Magnus, ‘Preserving the Autographic/Allographic Distinction’, Journal of Aesthetics and Art Criticism 73 (2015), 453–457. However, the discussion of Zeimbekis’ argument goes beyond the scope of my paper, except for D’Cruz and Magnus’s account of the allographic/autographic distinction, which I will discuss in the next section. 22 Goodman, Languages of Art, 119. 23 D’Cruz and Magnus, ‘Are Digital Images Allographic?’, 423; Joseph G. Moore, ‘Musical Works: A Mash-Up’, in Christy Mag Uidhir (ed.), Art and Abstract Objects (Oxford: OUP, 2013), 284–306. 24 Moore, ‘Musical Works: A Mash-Up’, 301. 25 Ibid., 302. 26 D’Cruz and Magnus, ‘Are Digital Images Allographic?’, 424. 27 Moore, ‘Musical Works: A Mash-Up’, 302–303, my emphasis. 28 Thanks to an anonymous referee for leading me to clarify the difference between the criteria I am proposing and that proposed by D’Cruz and Magnus. 29 D’Cruz and Magnus, ‘Are Digital Images Allographic?’, 425. 30 Ibid. 31 Davies, ‘Multiple Instances and Multiple “Instances”’, 417. 32 D’Cruz and Magnus, ‘Are Digital Images Allographic?’, 424. 33 Wollheim, Art and its Objects, 121, my emphasis. 34 Currie, An Ontology of Art, 100. 35 Davies, ‘Multiple Instances and Multiple “Instances”’, 417. 36 Goodman, Languages of Art, 121. 37 Ibid., 121. 38 Walter Benjamin, ‘The Work of Art in the Age of its Technological Reproducibility’, in Michael W. Jennings, Brigid Doherty and Thomas Y. Levin (eds), The Work of Art in the Age of its Technological Reproducibility and Other Writings on Media (Cambridge, MA: Harvard University Press, 2008). 39 For a similar interpretation of Benjamin’s notion of aura, see Wollheim, Art and its Objects, 117. 40 Walter Benjamin, ‘The Work of Art in the Age of its Technological Reproducibility’, 31–32. © British Society of Aesthetics 2017. Published by Oxford University Press on behalf of the British Society of Aesthetics. All rights reserved. For Permissions, please email: email@example.com
The British Journal of Aesthetics – Oxford University Press
Published: Jan 1, 2018
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 12 million articles from more than
10,000 peer-reviewed journals.
All for just $49/month
Read as many articles as you need. Full articles with original layout, charts and figures. Read online, from anywhere.
Keep up with your field with Personalized Recommendations and Follow Journals to get automatic updates.
It’s easy to organize your research with our built-in tools.
Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.
All the latest content is available, no embargo periods.
“Hi guys, I cannot tell you how much I love this resource. Incredible. I really believe you've hit the nail on the head with this site in regards to solving the research-purchase issue.”Daniel C.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud
“I must say, @deepdyve is a fabulous solution to the independent researcher's problem of #access to #information.”@deepthiw
“My last article couldn't be possible without the platform @deepdyve that makes journal papers cheaper.”@JoseServera