Art Objects and the Experience of Art-Making

In considering the question of machine as artist, the art object can be analytically separated from its making, and its making can be dualistically conceptualized as process on one hand and experience on the other. One of the reasons we value art is that there was an experience of its making. To better understand what is meant by the experience of art-making, this paper presents results from a qualitative, phenomenological study of a group of artists. These results appear in three groups: feeling, intention and lifeworld. Machines cannot experience art-making, at least not in the same way as humans, and thus they cannot create art but only art-like objects. Even so, in the present century, we should not be asking whether machines can be artists, but rather how machines can help more people experience art-making for themselves.


Introduction
Perhaps one of the things that makes us human is our penchant for wondering about what makes us human.Many answers to this question have been proposed, and just as many disproven.The latest of these, it seems, is art.We have long supposed that only humans can make art, and yet now computers seem to be making art.To mention two recent examples: DeepBach, an algorithmic model trained on the chorales of Johann Sebastian Bach, can produce compositions that many people, including experts, cannot distinguish from the actual work of Bach [8]; and images made by Creative Adversarial Network were found to be indistinguishable, including in terms of novelty, from art made by flesh-and-blood artists [4].Such facts have caused celebration and worry.Will artists lose their value?And what are humans for, anyway?In this paper, I seek to assuage these worries.
If we are concerned with the prospect of the machine as artist, we must first ask why it matters and what it means for something to be art.Analytically, we can separate art objects from art-making, and art-making in turn can be considered in terms of process on one hand and experience on the other.I present results from an empirical study of artists' experiences to show what this means.While machines can evidently create art-like objects through some process, even a creative process, they cannot experience, and I argue that experience is one of the reasons we value art.More critically, machines cannot question, and questioning is central to-even constitutive of-art-making.Perhaps that will change in the future.But in the present century, I would suggest that we should not be asking whether machines can be artists, but rather how machines can help more people experience art-making for themselves.

Art Objects
Why do we worry if one thing or another is art?What is at stake with such a question?One way to approach the issue is to ask: What does "art" get us?From the viewer's perspective, answers to that question might include beauty, numinous experiences, and moral knowledge, among others-the bundle of which we can call aesthetic values for simplicity.
Of course, viewers can find aesthetic value in a whole range of phenomena, many of which we don't consider to be art: sunsets, stormy seas, other people, industrial products. . .So things may be art-like without being art.Certainly machine-made objects can be art-like in this sense, whether we are speaking of automated factories for Apple products or algorithm-generated digital images.

Art-making
But what does it take for something to be not only art-like, but art?It seems to be about, in a word, the process.That is, some crucial aspect of any object is to be found in the understanding of how that object came to be.
Any number of cultural theorists have pointed out that deeper insight into a product or artifact can be gained by looking at how it was made-to name a few, documentologist Suzanne Briet (1951Briet ( /2006)), sociologist Howard Becker (2003) and anthropologist Tim Ingold (2012).But, moreover, it seems to me that this view is deeply human.I suspect, for instance, that this was the motivation for our creation myths, perhaps the earliest tool we humans used to understand ourselves and our world, and one that still drives our forays in cosmology and religion alike.Even on smaller scales, we are interested in process.One of the characters in the novel Umami, by Laia Jufresa, a musician, has just this realization: "One day I just said enough is enough to the product mentality, you know?It's not that I'm giving up playing, I just don't need to package it.I devote myself to the music now, not the orchestra.I'm all about the process now." So I suggest that, when we talk about art, one of the things we mean is that it was made.We think of art as a creative endeavor, after all, and create is right in that word.

Art-making: Process and Experience
When we speak of the "process" of art-making, there are two sides to the coin: what can be roughly called the subjective and the objective-that is, looking from within, and from without.
This distinction is an old one in philosophy, going back at least to ancient Greek and Roman attempts to consider human affairs in light of universal reason.In the modern era, the issue has been taken up by such thinkers as Spinoza, Kant and Kierkegaard, whose spirit lives on in contemporary philosophers such as Nagel and Scruton.As such, there is not space here for a thorough discussion.But in short, we humans perceive the world from a situated, limited perspective, and there is something it is like to be us.Still, we sense that there is an external reality, and we endeavor to view the world from that external reality-sub specie aeternitatis-such as in scientific investigations.In these efforts we are tempted to think that the internal view is accounted for by the external one, but according to the philosophers mentioned here, this is not the case: The subjective and the objective are not reducible to each other.For the interested reader, Nagel (1979) gives a succinct and readable account.
When it comes to art, this means we can conceptualize art-making in two distinct and irreducible ways: as the process of art-making, and as the experience of art-making.

Process
On one hand, art-making is describable as a process-that is, series of events, much as a game of chess can be represented as a list of moves.This is the tack that Becker (1982) takes in Art Worlds, in which he seeks to give the lie to the concept of the artist as genius.He writes: Works of art. . .are not the products of individual makers, 'artists,' who possess a rare and special gift.They are, rather, joint products of all the people who cooperate via an art world's characteristic conventions to bring works like that into existence.(p.35) Becker's study shows the complex cooperation of many agents involved in the creation of any work of art, including manufacturers, couriers, librarians, gallery owners, critics, etc.Thus Becker deposes talent, showing it to be a social construct; and art, on Becker's account, is simply "what an art world ratifies as art" (p.156).There is no other criterion than that.Aesthetic judgments come into play only insomuch as they are produced and negotiated by collective social activity.Art objects are simply those that are ratified, supported and preserved by the art world.
The kind of work that Becker describes can certainly be done by machines-indeed, machines have long played various roles in the art world, and recently they have participated "as artist," such as in the examples I mentioned in the introduction.Of course, Becker would deny that we should consider a machine an artist anymore than a human being.

Experience
On the other hand, art-making can be described from the subjective perspective, in terms of an artist's experience.On this view, we look at art-making not in terms of events, but actions.
That is to say, when we talk about art, we mean that there was something it was like for the art to have been made.That intuition, I think, is at the root of our classifications of some things as art and others not, and our ascription of certain qualities to art that we do not ascribe to other objects.Research in cognitive science shows that this holds even if the "art" and the "non-art" are the same objects in different experimental conditions [11].Such results indicate that we consider artworks to be extensions of the people who made them-who effected them physically and intentionally-i.e., that we see artworks more as people than as objects.Related accounts have been developed in the philosophical literature, such as by Ferraris and Scruton: Ferraris discusses how artworks awaken in us the same sort of sentiments that other people do [5]; and Scruton describes our meeting artworks as an act of overreaching intentionality, seeing the world "face to face" [13].The fact that we see artworks as extensions of people points us to the importance of human creative experience in defining art, as we value people because of their capacity for experience.

Aspects of the Art-making Experience: A Study
So we can conceptualize art-making as process on one hand and experience on the other, two irreducible perspectives.I suggest that what we uniquely value in art is the experience of its making-our ability to ascribe an art-making experience to the agent who made it.After all, staplers and fire hydrants are made mechanically through some process, but such products do not awaken in us the sentiments that artworks do.
But what exactly does "an art-making experience" mean?In my research, I have sought to characterize that in some detail, through an empirical, phenomenological study of the work of artists.Here I briefly discuss that work, which was part of a larger study on the information behavior of artists [7], for its relevance to the discussion at hand-that is, those results that outline the subjective quality of the art-making experience.I present thematic results of the art-making experience in three groups: feeling, intention and lifeworld.

Methods
This study followed the guidelines of phenomenology of practice [14], an interpretative methodology for exploring lived experiences of in any domain of practice, including the arts.In phenomenology of practice, a researcher qualitatively analyzes the experiences of a small group of participants (approached primarily through interviewing and observation) in order to understand the structure and essence of those experiences.The goal in phenomenology-of-practice research is not to garner particular generalizable facts, but to generate insights for new ways of being and designing; it "aims to open up possibilities for creating formative relations between being and acting, between who we are and how we act, between thoughtfulness and tact" [14, pp. 69-70].
To carry out this study, I recruited seven artist from Philadelphia to create self-portraits.Purposive and convenience sampling were used to assemble a participant pool that was diverse along the dimensions of age, gender, profession and artistic medium.I received ethics board approval from my university and worked with each artist to attain informed consent.In particular, each artist decided to be identified in this study rather than anonymized.In this paper, the artists are identified using their first names: Brian, Brianna, Britt, Emily, Jeannie, Justin and Tammy.For more information on each artist, the full report on this study can be referenced [7].I collected visual and verbal material from each artist as they individually worked on their self-portraits.At the end of each art-making session, each artist recorded their answers to an interview protocol (taking approximately 10 minutes) that I provided them with ahead of time.The protocol asked them to recount what they did in that session, describing what tools they used, how they felt, etc.Once the self-portrait was finished, I met with each artist to conduct a semi-structured follow-up interview (approximately 45 minutes), in which I sought additional details and asked questions about the meaningfulness of the experience overall.The goal was to cast a wide net and discern what aspects of the experience were salient to each particular artist.
All of the material from each artist constituted one lived experience example.

Findings
As many others have written, technology has always been involved in artistic practice.Indeed, the Greek techne, a root of our technology, can be translated as art.It is unsurprising, then, that machines were involved in the art-making of all the artists in this study.For instance, artists shared their progress on social media, photographed themselves to create reference imagery, searched online for other references and inspiration, etc.
But more deeply, this study offered findings about the experience of art-making, and here I focus on those results in order to demonstrate what is at stake with the question of machine as artist.

Feeling
As artists go about their work, they experience feelings.In this study, a pair of feelings that all the artists seemed to experience were tension and relaxation.For the artists in my study, art-making was, for the most part, relaxing.Still, the experience was marked by moments of tension.Sometimes this was a bodily tension-for example, Brianna and Emily, both painters, mentioned their back pain.In other cases, this was a psychic tension-Justin, a stained glass artist, described the anxiety of watching his painted glass pieces firing in the kiln, hoping they wouldn't break.
A key point with regard to tension which most of the artists mentioned was "the hump," to use the metaphor of climbing a hill.In Emily's words, the "hump" is "a period of frustration before I can coast, when I don't like how it looks."But once an artist crests the hump, to quote Jeannie, "you have a bit of a view of how things are. . .Now my vision of what the whole piece will look like is much clearer."After this, most of the artists described working faster and more easily as the piece neared completion.
Another interesting feeling that came up in these artists' experiences was the question of finishedness.Several of the artists seemed to know confidently when their self-portrait was completed; Brianna and Jeannie, however, thought their piece was done but it might not be, and that they might do some more work on it.Emily, for her part, says her piece is still unfinished, even at the time of this writing.As Brian told me in our interview, the question of whether a work is finished is a timeless and tiresome one in art education.But it is a feeling, and artists must develop a sense for this feeling.

Intention
The next group of findings has to do with the artists' intentions in their work.These artists intended to create art, and their intentions guided their attention and actions as they did so.Intention manifested in how the artists used reference materials, took breaks and stepped back, handled mistakes, and asked questions.
The artists in my study used reference materials of various kinds, including photographs, sketches and the work of other artists.Interestingly, several of the artists described using the work itself as a reference.Emily, for instance, was creating a multimedia installation involving Polaroid photographs and a painting.One of the photographs came out all blue because of a mechanical failure.Emily then used that blue as a reference for the background color of her painting.Reference materials were not used, apparently, in any predetermined way; rather, they were sought and used when the need and usefulness arose.They were used in two main ways: first, as direct visual references, in which the artist tried to duplicate an image; and second, as indirect references, so that the artist could get a general sense of what something looked like.For instance, Jeannie referred to a book of machine imagery so that she could get a "mechanistic impression" as she worked on depicting the farming equipment in her painting.Another aspect of reference materials was that they were only used up to a point.While they were used heavily at the beginning stages of each work, they were abandoned somewhere around the "hump" (described above).Brianna said it well: "I don't worry about matching the photograph anymore-I just want to do what will look good in the painting."At this point, the work itself becomes the artist's key reference.
As the artists sought to realize their intentions in their work, they took breaks from time to time.That is, the artworks in this study were not created in one smooth session.Every artist took breaks, each lasting minutes to weeks.Even while they were working, the artists took micro-breaks, what I call stepping back, when they physically distanced themselves from the work in progress, getting a fresh perspective on the piece and taking a short pause from effecting the work.Emily described the value of these tactics: "When I take breaks and then come back, I have completely fresh eyes on the painting.When I work for long periods of time, I end up making more mistakes and redoing a lot of that work." Emily's mention of mistakes brings up another theme relating to intention.In my study, all of the artists except Brian described mistakes they made and how they handled them.Examples included getting proportions wrong, not aligning things correctly, and applying too much paint.Oftentimes, these mistakes were sources of the tension described above.Just as with the reference material, the artist's relationship to mistakes changed as the work developed.Early on, the artists worked hard to fix their mistakes.Brianna, for instance, redrew various body parts several times to make them the right size and proportion.Later on, though, mistakes are embraced.This is evident in Brianna's account: "And when I think I'm done, I realize [the hands] are way too big.It bothers me, but at the same time I like the distortion."Afterwards, she does not change them.Also along these lines, Emily was inspired by the mistakes of famous artists.In our interview, she described a painting by Cézanne which had an orange splotch that she guessed was accidental.If Cézanne could embrace his mistakes, she reasoned, then she could as well.As she told me, "Those accidents are the fun part, after all.It's completely unintentional, but it's intentional that I leave it." I have suggested that the use of reference material, taking breaks and handling mistakes are all done in relation to an artist's intention.This intention does not emerge from nothing.Rather, it is developed early on, in the ideation stages of art-making, and this is done through questioning.In this study, most of the artists engaged with the ti esti of the self-portrait-that is, the question of what it is for something to be a self-portrait.Several of the artists also asked what it meant to be themselves, and to be a person.Different artists answered these questions differently, and their answers guided their art-making.That is, the artists depicted themselves in their self-portraits according to how they conceptualized themselves as people.Brian, for instance, described his self-portrait as the story of his whole life "collapsed into this one thing," while Jeannie focused on how "my existence is on the border. . . the interface between inside and outside," and for Emily her self was about "connections and relationships."

Lifeworld
The last group of themes relates to the lifeworld, i.e., the way people live, and all the things that are part of their worlds as they do so.This group of themes shows that art-making was not a closed loop, isolated from the rest of the artist's life.Rather, it was part of their life, and they brought aspects of their life to their work, and vice versa.This can already be seen to some extent based on the themes described above.To add further clarity, in this section I describe how memories, other activities, and other people and art were integrated in these experiences of art-making.Memories played a major part in these artists' experiences.All the artists spent time consciously reflecting on their pasts, but in addition to this memories rose up spontaneously.Of course, such is the case for all of us.But with the artists, the memories influenced their self-portraits.Brian, for instance, at one point covered much of his painting in white house paint, signifying attempts he had made to cover up his past; and Jeannie, reflecting on childhood memories of motion sickness and imbalance, added a depiction of the inner ear (which she deemed related) to her self-portrait.
Just as art-making is not separate from an artist's history, it is neither separate from an artist's present.Over the course of making their self-portraits, all of the artists in my study did other things.In some cases, the artists had day jobs.Even the professional artists worked on other projects while doing their self-portraits.Several of the artists mentioned getting insights or ideas for their self-portraits while they were doing other things.Tammy, for instance, got a flash of inspiration while she was out grocery shopping.Britt and Emily found themselves thinking about their self-portraits while at work.
Similarly, most of the artists involved other people in their experiences.Justin, for instance, asked his son and wife for feedback when he was trying to capture his likeness in his painting."Anyone who comes to visit my studio, I ask for feedback."Sometimes others' involvement was more indirect.Jeannie, for instance, found guidance on her piece through conversations with her sister and a discussion group, even though they didn't talk about her self-portrait directly.
And finally, just as these artists were not ahistorical monads of people, their self-portraits did not stand alone.Rather, other artworks played a role in the making of the self-portraits in this study.In some cases, these were other works by the same artist; Brian and Jeannie described working on multiple pieces at once, and how the pieces entered in conversation with each other as they developed.In other cases, the artists drew on works by other artists, as mentioned above in the discussion of reference materials.For example, Tammy browsed other artists' works while ideating her self-portrait, and Britt and Emily found themselves inspired by artworks they saw in galleries.

Can Machines Experience?
I have argued in this paper that one reason we value any given art object is because of the experience that its maker had while making it-whether or not the viewer really has any idea of the particular nature of this experience.Following philosophers such as Nagel, we can say that this experience is not reducible to an externally observable process.With the admission that this experience is fundamentally ineffable, I have sketched some aspects of the nature of the art-making experience by presenting results from a qualitative, phenomenological study of art-making.I discussed these themes in three groups: feeling, intention and lifeworld.
On my analysis, the question of whether machines can make art has been reframed as the question of whether machines can experience.That, of course, is a much-discussed question, with staunch defenders of both answers.It is not necessary to enter into a deep analysis of their arguments here.For those who would answer that machines do not experience (for the record, I find myself in this camp, at least for machines at present), this leads to the conclusion that machine-made objects cannot truly be art.Or, perhaps, taking art to be a matter of degree, they are not art as much as human-made art is.
For those who suggest that machines at present can indeed experience, it must be proven that the themes above manifest in the work of artists.The difficulty here is that they may not be readily observable.Do we see a "hump" in the work of machines?Do they seem tense at some times and relaxed at others?Can they agonize over whether a work is done?Machines can certainly use reference materials, but does the machine's relationship to reference materials change as the process goes on?Do machines benefit from taking breaks and stepping back?Can machines make mistakes, recognize them, and then sometimes correct them and other times embrace them?Can machines ask questions?Machines can evidently work with other people and machines and learn from other artworks, but can machines bring their past experience and other activities to bear on their art-making?It seems to me that the answer to most of those questions is no, lest we warp the meanings of the words we use far beyond the usual.Perhaps that is a limitation of language, or evidence of the insufficiency of anthropocentrism.To that end, in discussions of AI the question has been raised of the whether machine intelligence ought to be like human intelligence.Airplane flight, after all, is nothing like bird flight, and that is not just acceptable but actually preferable.Could the question of machine artistry be analogous?Perhaps not, if what we value in the experience of art-making are the aspects that we identify with, i.e., the human aspects of the experience.And yet perhaps so, if we find it stimulating-intellectually, aesthetically or otherwise-to explore how a being so unlike ourselves experiences something.
There is also the issue of imagined experience.When a viewer encounters an art object, they may know nothing about the artist who made it and consequently very little about how it was made.The viewer, then, can only imagine what it was like to make the that there was such an experience in the first place.Supposing a viewer comes across a work made by a machine but thinks it was made by a person, the viewer will still imagine an experience.Effectively, then, it is a work of art.
The situation would be something like the scene at the end of Nabokov's The Real Life of Sebastian Knight, when the narrator visits his brother Sebastian in the hospital.He was running late and feared that Sebastian would die before he arrived.What a relief when the nurse told him Sebastian was still alive.She let him step into the room, even though Sebastian was asleep and shouldn't be bothered.The narrator listened for a few minutes, and was comforted-changed, even, as he says-by hearing the breathing of his still-living brother.He felt communion with his brother's soul.But after the narrator left the room, the nurse apologized and told him that the man he'd just seen was not his brother.There had been a mistake.His brother had actually died, after all.But even under these circumstances, the narrator realizes the genuineness of the effect that hearing the breathing had on him-even though it was counterfeit, so to speak.
To be sure, there would have been no such effect if the narrator knew he was listening to just any sleeping person; and likewise, if a person knew that an object was made by a machine, it would not quite be art.That is certainly not to say the object would have no value-only that there may seem to be something missing.

Questioning
To close, I would like to return to a point that came up in the findings: the importance of questioning to the experience of art-making.I might even say that questioning is the crux of art-making.
Such a perspective could be considered the completion of a shift that started during the Renaissance.Before then, art was seen predominantly as a matter of physical effectuation.To put it uncharitably, mindless handiwork.It was the work of those who we now call the Old Masters who emphasized that art is the work of the head and heart as much as of the hands [15].This set the stage for our recognition of the artist's genius, which in the 20th century Becker so diligently sought to dismantle.Still, it is beyond question that we now see the "mental" side of art as critical, if not defining.The past century has shown us conceptual art, art made through assistants, etc.In my study, we saw that ideation-engaging with the ti esti question-established the intention that guided the art-making.
Extrapolating a bit, perhaps we can think of questioning as the critical aspect of the experience of art-making.This suggestion is supported by the accounts of any number of artists.Paul Gauguin, for instance, said his art sought to satisfy his "terrible itch for the unknown" [6, p. 33] and one of his most iconic works is titled with three questions: Where Do We Come From?What Are We?Where Are We Going?Mary Oliver showed this as well in her poem "The Man Who Has Many Answers."She writes: ". . .While the man who has only questions, / to comfort himself, makes music" [12, p. 69].
And so another way of understanding the question of whether machines can create art is to ask whether machines can ask questions.Famously, Picasso is supposed to have said of computers: "But they are useless.They can only give you answers."For now at least, that still seems to be the case.So if machines cannot experience and cannot ask questions, then it seems doubtful whether machines can create art, in the fullest sense.And yet it is beyond question that they can create art-like objects that have many of the characteristics of art.At least for now, in my view, true art-making is the purview of humans alone.This should not be cause for despair.I would change the question somewhat: In this century, we should not be worrying about whether machines can be artists.They can make art-like objects and that is enough.Rather, I think, we should ask how machines can help more people have experiences of art-making-as part of a broader aim to awaken critical and creative thinking throughout society.

Preprints
(www.preprints.org)| NOT PEER-REVIEWED | Posted: 19 July 2018 doi:10.20944/preprints201807.0363.v1 I analyzed these examples independently at first and later compared the results between the different examples to discern what was shared and what was unique.I used an iterative scheme of open coding for phenomenological themes, as described by van Manen (2014).