Are Some Perfumes Works of Art?

What more do we need to fully appreciate perfumes, beyond considering them objects for aesthetic appreciation? My contention is that our appreciation of some perfumes would be largely incomplete, unless we acknowledged them as works of art. I defend the claim that some perfumes are works of art from the point of view of different definitions. Nick Zangwill’s aesthetic definition makes it easy to defend the proposed claim, but is not very informative for the purposes of fully appreciating some perfumes. On the other hand, Jerrold Levinson’s intentional-historical definition and Dominic Lopes’s approach to defining art make it more challenging to defend the proposed claim. I show that, even so, the challenge may be met. Moreover, the challenge is well worth engaging with, since tackling it uncovers features of some perfumes that are essential to their full appreciation.


i. introduction
Are some perfumes works of art? A natural response to this question is to wonder who could possibly raise it. In other words, who cares whether perfumes are works of art? Here is an example of who and why. New regulations about admissible perfume ingredients have been issued at several stages by the European Commission (Wendlandt and Denis 2014). These regulations restrict the concentration of a number of ingredients amply used in the perfume industry, on the grounds that they might result in skin irritation in some individuals. The regulations imposed in 2014, for instance, restricted the use of oakmoss, which is present in the original version of classic perfumes such as Chanel N°5 and Miss Dior. As a result of introducing these regulations, according to Olivier Maure, head of a supplier of concentrates for major perfume brands such as Dior, "Many perfumes have had to be reformulated"-i.e., some of their ingredients have had to be either removed or reduced in concentration-"even though they were considered masterpieces." Maure compares the reformulation of these perfumes to "changing the colours of the Mona Lisa" (Euractiv.com 2014). So, here is someone who cares about whether perfumes are works of art, and who thinks that some of them are. The reason why this is important to him (as well as to other like-minded people about some classic perfumes) is that, if a certain perfume is a work of art, then it is unacceptable to reformulate it, just as it would be unacceptable to change the colors of the Mona Lisa. This is one reason, but not the only reason, for being interested in the question as to whether some perfumes are works of art. After all, there might be independent considerations for objecting to the reformulation of classic perfumes that do not hinge on those perfumes being works of art. So, what is at stake with considering some perfumes works of art? This is best clarified by contrasting the idea that some perfumes are works of art with the idea that they are simply objects of aesthetic appreciation. 1 I explore this contrast in the next section.

ii. beyond aesthetic appreciation
Suppose that we consider perfumes objects of aesthetic appreciation, namely objects to which a number of aesthetic categories, ranging from the beautiful to the ugly, apply. That these categories The Journal of Aesthetics and Art Criticism 78:1 Winter 2020 C 2020 The Authors. The Journal of Aesthetics and Art Criticism published by Wiley Periodicals, Inc. on behalf of American Society for Aesthetics apply, I believe, is quickly evident: some perfumes are aesthetically pleasing, others offensive, some refined, others unrefined. The question is: what more do we need to fully appreciate perfumes, beyond considering them objects of aesthetic appreciation? My contention is that our appreciation of some perfumes would be largely incomplete if it were limited to their being objects of aesthetic appreciation. As I illustrate in due course, these perfumes have much more to offer for appreciation, which could not be captured without considering them works of art from the point of view of specific definitions of art. Now, there is a substantive ongoing debate on the necessary and sufficient conditions for something to be a work of art (for example , Danto 1964;Dickie 1974;Levinson 1979Levinson , 1989Stecker 1990;Davies 2015). Some theorists even deny that these conditions can be provided (Ziff 1953;Weitz 1956;Gaut 2005). Still, this article works on the assumption that attempting to define art is a worthwhile endeavor (Stecker 1997). Specifically, as far as perfumes are concerned, I suggest that the full appreciation of some of them will only be possible if we are aware of what makes them works of art according to some definitions, which this article gradually illustrates. In sum, exploring the reasons why some perfumes are works of art will lead to their full appreciation. This is what is at stake with considering them works of art, and why it is worth engaging with the question as to whether they really are.
How am I going to show that some perfumes are works of art? First of all, note the restriction of my claim to the effect that only some perfumes are works of art. This is critical. I do not believe that all perfumes are works of art and, as will become apparent in the rest of this article, the perfumes I choose to argue for my thesis have a claim to the status of works of art in virtue of characteristics that are not shared by many other perfumes available on the market. I evaluate some perfumes in the light of some influential attempts at defining art. 2 I proceed in an order of increasing difficulty. I begin by considering Nick Zangwill's aesthetic definition of art. As I show, this makes it relatively easy for perfumes to be works of art, but obscures what is important for the full appreciation of some of them. I then consider Jerrold Levinson's intentionalhistorical definition, which I take as the central case for evaluating the claim that some perfumes are works of art. On the one hand, this definition poses a much harder challenge to the proposed claim. For example, a version of it calls for an art form to which perfumes belong, whose existence is not obvious, and which I discuss in conjunction with Lopes's approach to defining art (2014). On the other hand, addressing this challenge brings to light the features of some perfumes that are essential to their full appreciation.
iii. beyond intentions to determine aesthetic properties I begin by showing how perfumes satisfy Zangwill's definition of art. In order to spell out this definition, let me first say something about what nonaesthetic properties and aesthetic properties are. Nonaesthetic properties include physical properties (for example, shape, size) and what are sometimes called secondary qualities, such as color. Olfactory properties, such as smelling like peppermint, can be considered nonaesthetic properties insofar as they are ascribed on the basis of discriminative abilities-for example, correctly identifying a certain odor as peppermint rather than thyme. 3 Aesthetic properties, by contrast, are those for which, in Frank Sibley's words, "it would be natural to say that aesthetic sensitivity was required to see, notice, or otherwise perceive" them (1965,135). Among these, Zangwill distinguishes what he calls substantive aesthetic properties, examples of which are "elegance, daintiness, balance, or frenzy," and verdictive or evaluative ones, which are aesthetic merit or demerit (Zangwill 1995b, 307; see also Sibley 1959).
According to Zangwill's definition, something is a work of art if, and only if, someone acquires the knowledge that certain aesthetic properties will be determined by a certain configuration of nonaesthetic properties, and, therefore, intentionally endows something with certain aesthetic properties in virtue of that certain configuration of nonaesthetic properties (Zangwill 1995b).
Note that, from the point of view of this definition, any perfumer who has an insight that a certain combination of olfactory properties will result in aesthetic properties will create a work of art. So, most perfumes on the market turn out to be works of art according to this definition. This is something that Zangwill would most likely not find problematic, as illustrated by the following statement: "An advantage of my view is that for it fine art . . . lies on a continuum with everyday artistic activities. . . . If cake-decoration falls out as art, that's fine by me" (1995b, 316).
While Zangwill's aesthetic definition makes it easy for many perfumes to be works of art, it also obscures features that only some of them possess, and whose awareness is essential for their full appreciation. These are gradually uncovered as I evaluate whether some perfumes satisfy another definition of art: Levinson's intentional-historical definition.
iv. introducing levinson's intentional-historical definition: intention for regard-as-a-work-of-art According to Levinson's definition, something is a work of art if, and only if, it was "intended for regard-as-a-work-of-art, that is, regard in any way preexisting works of art are or were correctly regarded" (Levinson 1989, 21;see also 1979, 2002. There are two variations on this definition: an intrinsic and a relational one. According to the intrinsic definition, regard-as-a-work-of-art can be spelled out as a cluster of attitudes, such as "with close attention to form, with openness to emotional suggestion, with awareness of symbolism," without reference to the appreciation of any specific previous artworks-or, indeed, without reference to the appreciation of any specific art form, such as music or sculpture (Levinson 1989, 21). According to the relational definition, on the other hand, regard-as-a-work-of-art can be understood by reference to how specific preexisting works of art are or were correctly regarded, without any indication of the attitudes involved in this regard. It seems harder to defend the idea that some perfumes are works of art from the point of view of either version of Levinson's definition than from the point of view of Zangwill's definition. Still, I show how this idea can be defended, beginning with the intrinsic version.
v. the intrinsic definition: openness to emotional suggestion and awareness of symbolism The regard for which a work of art is intended, according to Levinson's intrinsic definition, includes openness to emotional suggestion and awareness of symbolism. So, the first question to tackle is whether any perfumes were intended to be regarded with openness to emotional suggestion or awareness of symbolism. 4 An affirmative answer to these questions has already been given by Shiner (2015). In what follows, I partly rely on his case.
Let me begin with emotional suggestion. Were any perfumes intended to be regarded with openness to emotional suggestion? An apt example in support of an affirmative answer to this question is Jacques Guerlain's L'Heure Bleue-literally, "the blue hour"-launched by his own perfume house, Guerlain, in 1912. 5 The inspiration for this perfume is said to have been given to Jacques Guerlain while taking a walk along the Seine in Paris, one summer evening of 1911. This was after sunset, when daylight was already gone, but the stars were not out yet. Guerlain reported that this hour evoked in him a feeling extreme calm and peace, and declared his attempt to create a perfume that would express it (Guerlain n.d.). This is an example of a perfume intended to be regarded with openness to emotional suggestion. 6 Note that, in order for Guerlain's intention to be of the correct kind, according to Levinson, it should not be a "passing" intention, that is, a temporary, fleeting one (Levinson 1979, 236). Rather, it should be a stable intention-for example, one that informs the creative process, rather than one that is temporarily entertained about the finished product (more on this in Section xii). Now, are any perfumes meant to symbolize and be regarded as such? 7 First, let me say something about a widespread phenomenon that is not an instance of symbolization. It is to be noted that odors in general, and not just perfumes, may give rise to imagery by way of association of ideas. Scruton (2009) calls our attention to this phenomenon by providing the following example: In Proust's novel Remembrance of Things Past, the flavor of a madeleine brings back some childhood memories to the protagonist's mind by way of association. 8 In this case, we would not say that the flavor of the madeleine symbolizes these childhood memories. 9 I will show that perfumes, however, may genuinely symbolize by way of exemplification, defined by Goodman (1968) as follows: Something exemplifies a given property if, and only if, (1) it possesses that property and (2) it refers to that property. 10 How does a perfume exemplify a property? For example, how does a perfume exemplify the olfactory property of smelling like a rose? First, this perfume will possess this property, that is, it will smell like a rose, thus satisfying (1). Second, condition (2), to the effect that the perfume refers to the smell of a rose, may be satisfied thus: the perfumer will find ways to call our attention to this property-for example, he or she will name the perfume in a way that alludes to a rose (for example, Ralf Schweiger's Lipstick Rose, launched by perfume house Frederic Malle in 2000), or mention "rose" in the description of the perfumefor example, listing it as one of its notes (that is, different identifiable odors-more on these in Section viii).
Olivia Giacobetti's Dzing!, launched by perfume house L'Artisan Parfumeur in 1999, exemplifies the olfactory properties of the leather of harnesses and saddles, of hay, of the fur of animals, and of the caramel of toffee apples. The description of the perfume draws our attention to these olfactory properties (L'Artisan Parfumeur n.d.). The exemplification of these properties contributes to the more abstract aim of evoking a circus-an aim that is also made explicit in the perfume description, as well as by the illustration on the bottle label, where a lady wearing an ostrich feather is riding a rearing up tiger.
It is a further and very interesting question how to characterize the relation holding between Dzing! and a circus. Since doing justice to this question in the context of the present article would be impossible, from now on I will use the term evoke to indicate this relation, without any commitment to the idea that it is symbolic. 11 It should be now clear that, although this most likely does not hold for all perfumes available on the market, there are many perfumes that were intended to be regarded with openness to emotional suggestion or awareness of symbolism. Moreover, reviews written by perfume admirers evaluate perfumes as more or less successful in these respects. 12 From the point of view of symbolism, for example, the case of Giacobetti's Dzing! is very divisive. Some reviewers think that Dzing! is a pleasant perfume, while not recognizing some of the olfactory properties pointed out by Giacobetti. Others express disgust precisely because they do recognize these olfactory properties. Here is an example of the latter reaction: Upon spraying, I was immediately hit with "petting zoo on a hot, summer day." It has it all: sweaty barnyard animals, muddy hay, . . . Awful! . . . I enjoy taking my kids to the petting zoo, but I definitely don't want to smell like it. (Basenotes 2017) vi. a proviso to the intrinsic definition: relatively complete ways of regard I have just argued that some perfumes were created with the intention that they be regarded with openness to emotional suggestion or with awareness of symbolism. I should now point out that, for all I have said about Levinson's intrinsic definition so far, it may seem as though a perfume could qualify as a work of art simply if it was created for either openness to emotional suggestion alone or awareness of symbolism alone. If that were the case, many more perfumes than I am putting forward as candidate works of art would qualify as works of art-for example, simply in virtue of exemplifying a given olfactory property, such as smelling like a rose. This result is blocked by a proviso on Levinson's part about the intrinsic definition: something is a work of art if, and only if, it was intended for "relatively complete" ways of regard, and not just single ones (Levinson 1989, 24). In order for Levinson's intrinsic definition to be satisfied, therefore, a perfume should be intended for relatively complete ways of regard, which do include openness to emotional suggestion and awareness of symbolism, but are not limited to them. Rather, complex ways of regard include also other attitudes, for example, attention to form.
In the course of the previous section, I have said that Guerlain's L'Heure Bleue was intended to be regarded with openness to emotional suggestion, and that Giacobetti's Dzing! was intended to be regarded with awareness of symbolism. However, a case can be made that each of these perfumes was intended for relatively complete ways of regard. For example, some descriptions of L'Heure Bleue also call our attention to specific olfactory properties-for example, the smell of iris and heliotrope. The description of Dzing!, on the other hand, suggests that this perfume expresses a playful state of mind, and also provides indications about the diachronic structure (a notion that is explained in Section viii) of this perfume: "[a]s the book closes so do the last notes of the paper waft gently" (L'Artisan Parfumeur n.d.). Thus, some perfumes satisfy the intrinsic version of Levinson's definition, and not on superficial grounds. 13 vii. the relational definition and lopes's proposal: does art perfumery exist?
Now, do some perfumes qualify as works of art according to the relational version of Levinson's definition? According to it, regard-as-a-work-ofart can be understood by reference to how specific preexisting works of art are or were correctly regarded, without any indication of specific regards.
The relational definition raises the most difficult challenge for the thesis that some perfumes are works of art. This is because the correct kind of regard has to be spelled out by reference to how previous works of art within that specific art form are or were correctly regarded. This calls for an art form to which perfumes belong. As we will see in due course, it is not obvious that such an art form exists.
The existence of a specific art form to which perfumes belong is also essential for a recent approach to defining art, namely Lopes's (2014). Lopes suggests that, rather than attempting to define art in general, we should rather attempt to define specific art forms. He provides a formula according to which x is a work of art if, and only if, "x is a work of K, where K is an art" (2014,14). 14 So, in order to defend the thesis that some perfumes are works of art from the point of view of the relational version of Levison's definition, as well as by the lights of Lopes's proposal, we need an art form within which perfumes can be regarded as works of art-let us call it art perfumery.
Providing evidence for the existence of such an art form may seem like an insurmountable obstacle. Indeed, Shiner, who has addressed the question as to whether some perfumes are works of art, maintains that an established art form to which perfumes belong is currently lacking, and that this is the main reason why perfumes, which are created for commercial purposes, are not works of art (2015,389). To be more specific, according to Shiner, while there exists commercial perfumery, art perfumery is yet to emerge as a distinctive way of producing perfumes that are works of art. 15 Assuming for the sake of argument that art perfumes do not yet exist, what would it take for them to come into existence? According to Shiner, Instead of norms related to wearability or a focus on pleasure, harmony, and beauty, art perfumes might favour scents intended to be appreciated for their combination of innovative structural complexity with an expressivity and symbolism that challenged the receiver's expectations of what a perfume should be like. (2015, 391, my emphasis) On the basis of these criteria, however, I argue that art perfumery already exists, although it is likely only in its infancy.

viii. innovative structural complexity
First, what is structural complexity in perfumes? To answer this question, something needs to be said about what the structure of a perfume amounts to. Perfumes can be said to have structure along two dimensions. A first dimension is a synchronic one: at any given point in time, upon smelling a perfume, one can distinguish different notes, namely different identifiable odors, such as rose or vanilla (see Richardson 2018). The chemical composition of a perfume is responsible for the possibility of smelling one or more odors at any given point in time, although the way in which different chemicals may give rise to different odors is far from simple: the same molecule can give rise to different odors, and the same odor might be produced by different sets of molecules (see, for example, Smith 2012; Poivet et al. 2016). 16 So, at any given point in time (that is, from the synchronic point of view), upon smelling a certain perfume, we might be capable of discerning multiple notes: for example, both rose and vanilla. Already from the synchronic point of view, there is room for innovative structural complexity in the creation of a perfume by way of, for example, producing unusual notes or unusual sets of notes. For example, Christopher Sheldrake's Douce Amère, launched by Serge Lutens in 2000, reproduces the smell of absinthe. This is what Sheldrake says about it: The idea of Douce Amère is back in the Twenties with the absinthe. Obviously it's narcotic. It became an illegal product. . . . But it didn't just get you high drunk, it actually damaged the nervous system and in this fragrance I wanted to put as much absinthe as possible, which I did-well, as much as I could legally! And why Douce Amère? Because the idea was to pour absinthe on a cube of sugar before you eat it-so it was bitter and sweet. . . . It's not very well known but challenging and a great fragrance. (Osborne 2013) Another dimension along which structure may be conferred to perfume is a diachronic one: due to the properties of the molecules composing perfumes and the different timing with which they evaporate, many perfumes undergo three discernibly distinct phases in their temporal development, such that different notes or sets of notes may be recognized at each of these different stages. The first set of discernible notes are commonly referred to as top notes, the second as middle notes, and the last as base notes. Innovative structural complexity may be obtained by exploring contrasts between top, middle, and base notes. For example, Jérôme Epinette's 1996 Inez & Vinoodh, launched by Byredo in 2013, features a stark contrast between very harsh top notesjuniper berry and black pepper-and soothing base notes, where vanilla is predominant.
Here is another interesting way in which some perfumes exhibit innovative structural complexity. Geza Schön is a perfumer who has created perfumes based on a single molecule. This is the principle behind his own brand, aptly named Escentric Molecules. For example, Geza Schön has designed the perfume Molecule 01, launched in 2006, composed of the artificial molecule Iso E Super only. One way in which this perfume challenges the receiver's expectations of what a perfume should be like is by being as comforting as it is elusive: from the point of view of the wearer, due to the properties of Iso E Super, Molecule 01 intermittently disappears-that is, occasionally the wearer cannot smell it. 17 Thus, Molecule 01 exhibits innovative structural complexity by means of an especially surprising diachronic development. Incidentally, this characteristic of Molecule 01 also reflects little concern with norms related to function: why wear a perfume that you yourself cannot smell?

ix. symbolism that challenges expectations of what a perfume should be like
Let me move on to the idea that the norms related to wearability-now with a concern for pleasantness-might be relaxed in favor of expressivity and symbolism that challenge the receiver's expectations of what a perfume should be like. Fulfilling this condition might seem like an impossible obstacle for any perfume, since perfumes are, by their definition, made to be worn and, relatedly, smell nice. Yet, I will show that some perfumes are indeed created with the explicit intent to heavily relax, and in some cases nearly break, the link with this function, through a symbolism that challenges the receiver's expectations of what a perfume should be like.
A case in point is Antoine Lie's Sécretions Magnifiques, launched by Etat Libre d'Orange in 2006. Here is the idea behind this perfume in Lie's words: The concept was what is happening inside your body when you're getting an emotion, when you are with a woman or a man and you are desiring her or him. The mechanism of the fluids, what would it smell like if it were outside your body? When you get an emotion, your blood runs faster and you get more adrenaline, you sweat because you are excited . . . The transition of this emotion into a fragrance, what would it smell like? (Matos 2013) The idea behind Sécretions Magnifiques is to effectively and realistically evoke this state of the body, through the expression of tension and excitement and the exemplification of bodily fluids, such as blood and sweat. The result is sufficiently convincing that a significant number of people find this perfume unbearable: they refuse to wear it and would much rather not to smell it on anyone around them. 18 This is by no means a mistake on its creator's part. In designing Sécretions Magnifiques, Lie intentionally created something as realistic as it is shocking, meant to challenge people's expectations, to the detriment of ease of wearability. In Lie's words, Sécretions Magnifiques is a reaction from the very conceptualized and safe universe of fragrances. For me it's not a fragrance per se. It's more an olfactive construction, trying to push the limits to the point where we say that it's not possible to go further than that. (Matos 2013) Moreover, it can be easily shown that many other perfumes by Lie-for example, Red+MA (launched by Blood Concept in 2012), which smells like blood-use symbolism that challenges a receiver's expectations of what a perfume should be like. Another interesting example is XX Latex, launched by UERMI in 2014, which, in keeping with its name, smells like latex. But should you be under the impression that Lie, who has gained the fame of "subversive perfumer," is an isolated case in the exploration of expressivity and symbolism that challenge the receiver's expectations of what a perfume should be like, think again of Olivia Giacobetti's Dzing!. The previously reported reaction to it shows that the sort of olfactory properties that Dzing! exemplifies are sometimes neither expected nor especially welcome.
x. are some perfumes regarded as works of art?
In sum, so far I have provided a few examples of perfumers who create perfumes with a concern both for innovative structural complexity (for example, Jérôme Epinette and Geza Schön) and for expressivity and symbolism that challenge the receiver's expectations of what a perfume should be like (for example, Antoine Lie and Olivia Giacobetti), with relaxed adherence to norms related to function. This, however, may not yet be enough to show that the perfumes created with this kind of sensitivity are works of art. A doubt that may now legitimately arise is that the perfumes I have described, although created with certain characteristics in mind, are not appreciated for these characteristics. Here is an example of how this could happen. Suppose, on the one hand, that a perfumer designs a given perfume with an eye to innovative structural complexity. Suppose, also, that the audience appreciates this perfume as, for example, pleasant and well made, while being unaware of its innovative structural complexity. Thus, the audience would not appreciate this perfume for its innovative structural complexity. So, are perfumes ever appreciated for their innovative structural complexity or challenging symbolism?
At least some of the time, they are. As pointed out by Shiner, quoting perfume designer Jean-Claude Ellena, some perfumes are evaluated by some perfume admirers in terms of their expressive and structural characteristics, independently of any concern for how pleasant they are (Ellena 2011;Shiner 2015). To this, let me add that a very important element for a perfumer to gain recognition among perfume admirers is that his or her creation should be original: if it is a, however masterful and aesthetically pleasing, repetition of what has already been done in the past, it is deemed uninteresting. The following review of Bertrand Duchaufour's Tralala, launched by Penhaligon's in 2014, illustrates this idea, in that it shows some form of appreciation of this perfume in purely aesthetic terms while condemning it as lacking originality: Fear not: it isn't terrible. In fact, it isn't even bad. . . . As we'd expect from Duchaufour, Tralala is an expertly put-together piece of work. The trouble is: it is precisely the sort of work we'd expect from Duchaufour, almost to the point of self-parody. . . . So yes, it is quirky . . . but it also feels like an uninspired rehash of old ideas. . . . ultimately it fails to make an impact. (Alavi 2014) Another example of the kind of evaluation that some perfumes undergo is Luca Turin's review of Quelques Fleurs L'Original, originally created by Robert Bienaimé and launched in 1913 by Houbigant, and recently reformulated: It was a 1912 composition, and it is pretty clear that easily half of the materials in the current version did not exist in 1912. I would be prepared to forgive everything in the name of progress if the fragrance were remotely interesting. But it is as dull as a floral can be. (Turin and Sanchez 2010, 299, my emphasis) Now, so far I have given reasons for thinking that some perfumers design perfumes with a concern for innovative structural complexity and expressivity and symbolism that challenge the receiver's expectations of what a perfume should be like. Moreover, I have given reasons for thinking that some perfume admirers also evaluate perfumes from the point of view of these characteristics, and not only (and sometimes not at all) from the point of view of how aesthetically pleasing these perfumes are. However, for all I have said so far, these two phenomena-that of perfumers creating perfumes with attention to certain characteristics, and that of perfume admirers evaluating perfumes in the light of these characteristicsmight be unrelated. In other words, I have said that some perfumers create perfumes with a concern for, for example, innovative structural complexity. But do they have any expectation that their audience will appreciate their perfumes specifically in the light of these characteristics? 19 In other words, as part of their creative process, do they also intend their perfumes to be appreciated for these characteristics? I try to show that, at least as of very recently, this is indeed the case.
xi. are some perfumes intended for regard-as-works-of-art? the infancy of art perfumery Thanks to the phenomenon of publishing perfume reviews on the internet, the attention that some perfume admirers pay to challenging symbolism and more subtle characteristics of perfumes, such as innovative structural complexity, is becoming increasingly well known. More specifically, it is becoming increasingly well known to perfumers, and there are reasons for thinking that this has started to influence some perfumers' creative process. A proof of this is given by the following interview to perfumer Christopher Sheldrake. He says: What I like about the internet is the fact that there is obviously a real interest, and a desire to know more, and to understand, and I think that's great-curiosity is a wonderful thing. I think it keeps the industry on its toes, encouraging more interesting creations. (Osborne 2013, my emphasis) However, one may worry that the sort of attitude reported by Christopher Sheldrake is shared by very few perfumers, rather than by a wider community. For example, Shiner points out that the main obstacle for the existence of art perfumery is that the production of perfumes for commercial purposes is, by its nature, subject to the norm of wearability: perfumers have to avoid scents that consistently give rise to aversive reactions because of a necessity to respect the needs of the average consumer (2015, 386). Shiner does admit that, occasionally, some perfumers heavily relax adherence to the norm of wearability, but he considers this phenomenon too isolated and occasional to claim that even a subset of commercial perfumery has the status of art (387).
I want to challenge this idea. I have consistently mentioned the perfume houses within which the perfumes that are candidate works of art were created. Many of these, such as L'Artisan Parfumeur and Etat Libre d'Orange, can be classified as instances of niche perfumery. Niche perfumery is a section of commercial perfumery that privileges creativity over mass distribution. This is reflected in the fact that niche perfume houses typically invest less in advertising than do mainstream brands. 20 Of course, niche perfume houses sell their own perfumes just as mainstream brands do, but, since the former target a smaller audience, they are under less pressure to create something that will please a high number of customers, compared to brands that mass-distribute their perfumes. This makes it possible for niche perfume houses to be less concerned with norms related to function (specifically, wearability) and more with creativity. More generally, the restricted audience to which their perfumes are directed makes niche perfume houses the most natural terrain for perfumes as works of art within the world of perfumery. By contrast, many perfumes available on the market-typically those that are mass-produced-are created without much, if any, regard for characteristics such as innovative structural complexity or challenging symbolism. Rather, the creative process behind them concentrates on the purely aesthetic appreciation that these perfumes will receive, in accordance with the need to meet the expectations of the average consumer (see also Shiner 2015, 386).
What is interesting for the purposes of my thesis is that niche perfume houses are increasingly less an isolated phenomenon. Moreover, in line with the proposal that I am making, note that Christopher Sheldrake alludes to niche perfume houses as those that are more likely to try to create more interesting perfumes, partly as a response to people demanding more interesting perfumes: Why are there so many niche brands today? It's because people are fed up with the big brands that are all making the same thing, because it wasà la mode, because it was successful, and the niche brands have come up with alternative ideas, and that's all to do with communication and people putting their point over. (Osborne 2013, my emphasis) Therefore, the increasing amount of niche perfume houses and the creative process they adopton the one hand, less subject to the needs of the average consumer strictly in terms of wearability, and, on the other hand, receptive to the demands for more interesting and challenging creationscan be taken to support the very recent emergence of an art form concerned with the creation of perfumes. In sum, I suggest that art perfumery exists, though it might be fair to say that it is in its infancy.

xii. the relational definition within art perfumery
Having made a case that a specific art form involving perfumes-art perfumery-does exist, I am now in the position to provide concrete examples of how some perfumes meet the relational version of Levinson's definition of art. According to it, something is a work of art if, and only if, someone intended it for regard-as-a-work-of-art, in the way specific preexisting works of art are or were correctly regarded, and intended that it is correct that this is so regarded (Levinson 1990). I now show that some perfumes meet this definition in the context of art perfumery.
Let me first explore and reject a way in which perfumes may satisfy Levinson's relational definition on superficial grounds. I do this by considering what Stecker called "the case of the stupid relational intention" (1997,93). 21 Suppose that a shop assistant calls a customer's attention to some random perfume on the shelf, and intends that the customer regards it as the art perfume L'Heure Bleue by Guerlain is correctly regarded, and intends it to be correct that the perfume on the shelf is so regarded. However, the shop assistant is unable to say anything more about his intention. Has the shop assistant turned the random perfume on the shelf into a work of art? We would like to say that he has not: his intention concerning the perfume on the shelf is too superficial. How can we rule out this intention as sufficient for producing art on the basis of Levinson's definition?
I consider the following lines of response that are available to Levinson. First, as mentioned in Section v, the relevant kind of intention according to Levinson should be nonpassing, that is, stable. By contrast, the shop assistant's intention seems to be a fleeting one: it comes and goes with the shop customers (see Levinson 1993b, 417). Moreover, Levinson has suggested that the person with the art-constituting intention should have "proprietary right" over whatever counts as an artwork (1979,236). This condition has been further spelled out by Levinson in terms of a "right to transform materials or determine their identity conceptually" (Levinson, personal communication reported in Stecker 1997, 91; see also Levinson 1993aLevinson , 1993b. This condition is not without controversy, but for my purposes it is enough to say that the creative process behind art perfumes is such that perfumers may be credited with proprietary right over the perfumes they create, since they literally bring them into existence by composing a certain formula. 22 By the same token, since perfumers' intentions inform the creative process behind the resulting perfumes, their intentions are also nonpassing. By contrast, the shop assistant in the previous example lacks proprietary right over the perfume on the shelf, and his intention is a passing one. Let us now consider Levinson's relational definition, amended to include the proprietary right condition. In support of the claim that some perfumes satisfy this definition, let me point out that some perfumes are explicitly declared by their creators to be inspired to, and to be regarded in the same way as, previous classics: this is the case for Passportà Paris by Dawn Spencer Hurwitz (launched in 2013 by perfume house Dawn Spencer Hurwitz), advertised by its creator as paying tribute to both Aimé Guerlain's Jicky and Paul Parquet's Fougère Royale (launched in 1882 by Houbigant). 23 But this is not the only way in which art perfumes could be intended for regard by reference to previous perfumes. Ways of correctly regarding artworks may include the acknowledgement that a certain work of art repudiates a certain tradition. 24 This kind of regard is correctly given to the aforementioned Sécretions Magnifiques, which was created in a conscious effort on Antoine Lie's part to distance himself from the previous perfume tradition. Moreover, Etienne de Swardt, the founder of perfume house Etat Libre d'Orange, within which Sécretions Magnifiques was created, ambitiously calls 2006, the year of the foundation of Etat Libre d'Orange, "Year Zero of Perfumery," showing at once an awareness of past tradition, as well as the intent of breaking free from that tradition. Thus, any perfumes that are similarly revolutionary may be regarded as Sécretions Magnifiques is correctly regarded: as repudiating a certain tradition. In conclusion, many perfumes can be shown to satisfy Levinson's relational definition.
xiii. conclusion I started out with the question as to whether some perfumes are works of art. I showed that Zangwill's aesthetic definition makes it relatively easy for many perfumes to count as works of art, simply in virtue of having been created to determine aesthetic properties. I endeavored to show that there is more than this to some perfumes. I have shown that at least some perfumers create their perfumes with the intention that they will be regarded in ways preexisting works of art are or were correctly regarded, namely with a complex of attitudes that include openness to emotional suggestion and awareness of symbolism. Thus, some perfumes satisfy the intrinsic version of Levinson's intentional-historical definition of art.
I have also shown that an emerging art form specifically concerned with the creation of perfumes, namely art perfumery, exists, although it is in its infancy. This is supported by the existence of increasingly widespread practices of perfume design-in particular, within niche perfumerythat display a greater concern for creativity than for norms related to function. Many perfumers in this growing tradition explore innovative structural complexity and symbolism that challenges the receiver's expectations of what a perfume should be like. Thus, some perfumes may be considered works of art insofar as they belong to art perfumery, in line with Lopes's approach to defining art by means of defining art forms.
Lastly, a closer look at the history of perfumery reveals a practice of creating some perfumes that are intended for regard as previous art perfumes are or were correctly regarded, thus satisfying the relational version of Levinson's intentionalhistorical definition. In sum, even from the point of view of Levinson's definition of art and Lopes's approach to defining art, which could have initially seemed rather challenging to meet, some perfumes are indeed works of art. In the light of the reasons why they are works of art, they can be fully appreciated and acknowledged to be worthy of a special kind of regard. 25 CHIARA BROZZO Department of Philosophy Durham University Durham DH1, United Kingdom internet: chiara.brozzo@gmail.com references experience that could be involved in the appreciation of art. Shiner and Kriskovets address this point, noting that the rapid visceral reaction that certain perfumes elicit does not exhaust many people's experience of these perfumes (2007). Shiner and Kriskovets point out that experiencing perfumes also involves our cognitive abilities: for example, experiencing perfumes may recruit the appreciation of the sequential ordering of different notes (as I previously illustrated with Epinette's 1996 Inez & Vinoodh). While cognitive abilities may not be exercised by every person experiencing a perfume, they are exercised by perfume admirers, as witnessed by perfume reviews. I am grateful to an anonymous referee for raising this point.
19. What I have pointed out is that some perfumers call attention to the expressive and symbolic aspects of their perfumes. But this leaves it open whether they also call attention to other features, such as innovative structural complexity. Also, it is possible that the audience of these perfumes might spontaneously appreciate perfumes for characteristics such as innovative structural complexity, without it being part of the perfumer's intention that they should thus appreciate them.
20. For example, you are very likely to have, however accidentally, sometime stumbled upon the advertisement of a mainstream brand perfume, but hardly, if ever, will you have accidentally stumbled upon the advertisement of a perfume by Etat Libre d'Orange or L'Artisan Parfumeur. It is, in fact, very likely that you will never have heard of these brands, unless you are a perfume admirer.
22. This terminology is used by perfumer Ellena (2011). 23. Incidentally, if it is true that art perfumery is only in its infancy, the reason why perfumes, such as Guerlain's Jicky, would be works of art is because their status as works of art would be retroactively acknowledged. See, for example, Carney (1994).
24. This idea derives from Carroll (1988, 155). Carroll puts forward his own proposal (rather than a definition) about how to identify works of art, and suggests how to integrate features of his proposal into Levinson's definition.
25. I am grateful to Hong Yu Wong for encouraging me to write on this topic and for feedback on a previous version of this article; to him, Gregor Hochstetter, Marco Santambrogio, and Krisztina Orbàn for helpful guidance, as well as to the audiences of the Philosophy of Neuroscience Research Group Forum at the University of Tübingen (including Katia Samoilova, Susanna Schellenberg, and Maarten Steenhagen) for feedback and support on an early version of this work. I am very thankful to Bence Nanay for insightful suggestions on previous versions of this article, which steered it in the direction it eventually took, and to Steve Butterfill for feedback on a previous version of this article. I am indebted to Elisa Caldarola for invaluable extended discussion on many aspects of this article, including the topic of representation and expression. I also want to thank James Nguyen and Ulrich Schlösser for extremely helpful input on that topic. I am grateful to Barry C. Smith for giving me the opportunity to present this article at the CenSes Seminar at the Institute of Philosophy, University of London, for raising important questions, for his feedback and that of other participants to the seminar (including John Behan, Anna Drozdzowicz, Raphaël Millière, and Harry Sherwood). I am also grateful to Clare Mac Cumhaill for her feedback and the opportunity to present this article at Durham University. I am thankful for discussion to the audience of the talk I gave there, especially Cain Todd for his thought-provoking response, and Bob Kentridge and Mohan Matthen for their feedback and support. I am grateful to the editors of this journal and two anonymous referees, whose clear-sighted feedback greatly improved this article. I owe much of my knowledge of niche perfumery to Maxime Bocxtaele of Necessities (Antwerp) who also encouraged me to write this article. I happen to be a great admirer of the work of both Bertrand Duchaufour and Olivia Giacobetti, including Tralala and Dzing!, despite the reviews that I reported. My research was supported by the John Templeton Foundation (ACT Fellowship awarded to Hong Yu Wong) and by the Fritz Thyssen Foundation.

this is an open access article
This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.