Arlyne Moi

Towards a Justifiable Conception of ‘the Autonomous Artwork’ in Today’s Artworld

Thesis for a "Hovedfag" in Philosophy at the University of Bergen - Spring 2005.

Back to table of contents - Next chapter



Since the nineteenth-century, the autonomous artwork[1] has been one of the touchstones in the field of art. It is often baked into discussions about artistic creation, identity, interpretation, and value. When the expression is not specifically mentioned, it is implicitly present in other usage such as “the artist-genius”, “the work in itself”, “originality”, “the hermetic work”, “non-purposiveness”, “ineffectuality”, “art’s separation from epistemology and morality”, “the artworld’s play-logic”, ‘transgressive art’, and ‘Modern art’, etc. If interlocutors do not assume some version of the work’s autonomy, they are often reacting against it, challenging it.

‘The autonomous artwork’ is a controversial issue

Interpretations of Immanuel Kant’s Critique of Judgment[2] §1-60 established ‘autonomy’ as pre-eminently relevant for thinking about ‘a field of art’, and throughout the nineteenth and twentieth-centuries of Modernism, its relation to the field—artworld[3]—complexified, culminating in a swamp of disparate conceptions, all vying for legitimacy. Under High Modernism, with the exception of directions such as Social Realism (e.g., Diego Rivera), most Western artists were pretty much on the political periphery;[4] as long as their expressions were confined to the art institutional setting, they received little political interference, at least on the face of things.[5]

Yet today’s artworld experiences a transition: Now artists are leaving art-institutional settings and entering the everyday spaces where historical consequences, instrumentality, responsibility, personal ethics, relations of power and stated intentions are readily apparent. One pertinent Norwegian example of this transition is Kunst Passasjen located in the Oslo Metro, a typically commercial space [Illustration 1][6] Other examples are Bergen Kunstforening, Nasjonalmuseet in Oslo and the Astrup Fearnley Museum of Modern Art, Oslo. Bergen Kunstforening was, until a few years ago, the venue for artworks traditionally understood as autonomous; it has now been displaced with Kunsthallen, which mounts demonstratively politically enmeshed exhibitions.[7] Nasjonalmuseet in Oslo recently re-hung their “basis exhibition” according to themes, some of which are markedly morally enmeshed (‘the vulnerable human being’). As for Astrup Fearnely Museum, they regularly mount exhibitions highlighting artworks relating to our common everyday life-world; expressions urgently engage with content such as temporality, memory, commercialism, consumerism and social commentary.[8] These four examples are instances of late modern ideology[9] attempting to dismantle the distinctions famously put in place by Kant, between the domains of aesthetics, epistemology and morality. Still, in spite of the dissolution of clear institutional boundaries and demonstrative instrumentality, many of the dogmas and dreams about the autonomous artwork continue—some of the basic assumptions of Kantian[10] (i.e. Modern) aesthetic theory are kept, Kant’s “Copernican turn” in aesthetics[11]—but these assumptions are ever more critically crowded by questions concerning the social-political function of art, artist’s moral responsibilities, the psychology of aesthetic perception, the sociological relations of art, gender or race-related issues, and the role of art in interpreting the needs and desires of the polis.[12]

Furthermore, if one accepts, as I do, that ready-mades are art, one must accept that anything can be art.[13] Because of this, the late modern ideological trends that deny moral/epistemological independence of artworks seem to make good sense, because the Kantian aesthetic point of view—the view that swept moral and epistemological considerations aside—becomes increasingly difficult to maintain simultaneously as receivers experience it as incomprehensible and irrelevant. For example, formerly it was common to address AIDS in news media, law-courts, health-industry publications and from the pulpit, but now this theme is directly addressed in art. By changing the context in which AIDS is broached, a morass of moral concerns arise: Intuitively, it seems misguided to don a so-called disinterested attitude and restrict one’s reception to “art-internal” concerns such as the golden mean, or to focus on taste, or the work’s supposed indeterminate purposiveness. Can we assign AIDS-related artworks a separate sphere of value on account of some understanding of their autonomy? What makes them different from AIDS-related non-art? Andrew Bowie, for example, asserts that the significance of an artwork lies in its ability to reveal, by some special status, what nothing else can.[14] But hasn’t the loss of an exclusive institutional setting, political and instrumental artworks, and the ready-made, made a mockery of that claimed special status? ‘The autonomous artwork’, in its many guises, is under redoubled attack. At one end of the spectrum of today’s artworld, advocate of Modernism could say:

‘The autonomous artwork’ remains an important expression for the artworld; it denotes something beyond the threshold of determinate knowledge, if not moral constraints, and it is not directly instrumental. ‘The autonomous artwork’ is crucial for ensuring the artist’s freedom of expression and art’s contribution within advanced capitalist societies. Moreover, there are many justifiable ways of interpreting ‘autonomous artwork’, which share in traditional Kantian commitments.

At the other end of the spectrum, a dissenter could respond:

‘The autonomous artwork’, in most of its interpretations, is wishful thinking. It hides a multitude of deceit. If anything about the artwork were considered autonomous today, it would be to the detriment of cultural life generally. Artworks are directly instrumental and inter-relational with epistemological and moral concerns. Most aspects of the Kantian moments of aesthetic judgment are defunct.

One international example of these competing voices is Lucy Lippard and Hilton Kramer: When Lippard, in a catalogue text for the Art & Ideology exhibit (1990) at the New Museum of Contemporary Art in New York, stated: “All art is ideological and all art is used politically by the right or the left.” Kramer replied: “[…] this movement toward the politicization of art in this country is an attempt to turn back the cultural political clock to the Stalinist social consciousness of the 1930’s.[15] Yet these competing voices are also heard in Norway: In 2003, fifty-nine Norwegian artists and art-professionals were surveyed on the theme of autonomy.[16] To the question: If art is autonomous, then what is it dependent on?” one participant answered “The magic of the night”. Another replied “Itself”, while a third answered “A whole lot of people, who are willing to make use of its physical potential”. To the question “Why is art morally responsible?” answers varied: “Because art has to do with qualities, language, distinctions that involve communication between human beings.” Contrariwise, “Art in itself has no moral responsibility”. The participants were also asked about the work in relation to the artist, with the question: “Why is an artist not morally responsible?” Answers ranged from: “Because he is an artist” and “A morally responsible artist is no artist,” to “I don’t think it is possible to both call yourself an artist, or even a human being for that matter, and at the same time claim that this is not morally demanding.” Hence, both internationally and locally, ‘the autonomous artwork’ is a controversial issue, ambulating from being something wanted and claimed, to being something contested or denied.

‘The autonomous artwork’ is a confused issue

There are a myriad of construals of ‘the autonomous artwork’ jumbled in the cacophony of the artworld: the artist’s private expression; ontological independence; the separate status of the artwork in society, its separation from truth and moral judgments; the receiver’s disinterested reception; the work’s radical ambiguity; its self-sufficiency for either interpretation or justification; under-determination by language; transcendental essence, transgression of norms, non-purposiveness; an empty game; powerlessness or infectivity; the fragment; absolutely un-reducible unit; non-defineability; just to mention a few. One problem, as I see it, of what has been written on the theme of ‘the autonomous artwork’, is that writers tends towards “myopia”; there is a lack an appreciation of the enormity of different concurrent conceptions.[17] Indicative of this “myopia” is Atle Kittang’s very useful essay, “Til forsvar for autonomiestetikken—rett forstått”, where he only acknowledges one explication of ‘autonomous artwork’ as legitimate.[18] Meanwhile, many art-professional’s practice seems to indicate that they have not reflected much over what they want ‘autonomous artwork’ to mean, and how the different conceptions can conflict, and how ‘the autonomous work’ may be distinct from instrumental works. Self-contradictions arise. For example, when Dror Feiler’s Snow White and the Madness of Truth (2004) [Illustration 2] made international news, the art critic Yoram Kaniuk,[19] in his discussion of Feiler’s work, did not seem to be expressing an Adornoesque duplicity when he said: “Kunsten er ikke hellig. Kunsten forandrer ikke verden.” (Art is not holy. Art cannot change [anything in] the world.) According to a reasonable interpretation, Kaniuk’s assertion is problematic: The first sentence indicates that the artwork is not autonomous because it is not cut off from human moral and epistemological judgments. Therefore, we presume, it can be judged according to some standard of truth and morality. But the second sentence contradicts this: Art is ineffectual with regard to the daily activities of humans. Thus it is “holy” in the sense of blameless; we must assume the work is not able to encourage people to hate or kill, empathize or mourn; it is merely a goal unto itself, has no determinate significance for events in the world. Thus from the same mouth, the artwork is first criticized for wielding power (the instrumentalist position) and thereafter criticized for lacking power (an autonomist position). Another inconsistency about the Feiler incident was that Israel’s ambassador Zvi Mazel, a significant political person, was invited to an instrumental artworld but thereafter scolded for not conforming to an autonomous artworld’s rules. Yet it may be argued that Mazel himself defended his “demounting performance” by playing the role of autonomous “the artist-genius”, claiming to act from feelings without cognitive deliberation: “I acted out of how I felt, I could not do things any other way.”[20]

But are these just instances of inconsistency, or could the artwork’s autonomy and its instrumentality—the touted double character—coherently co-exist? Here is, as I see it, the contemporary scenario: The artist ardently desires that her works be instrumental (useful, significant and valuable in the political, morally laden every-day world); simultaneously she claims the work’s autonomy (it is free to address any theme via any media and be presented in any venue).[21] As soon as the consequences of the work’s instrumentality come to bear, the artist runs for cover under various and sundry of Modernism’s conceptions of autonomy. Then the whole language game of ‘the autonomous artwork’ is deployed.[22] This scenario poses a striking problem because most theories designed to account for the work’s autonomy rely for their paradigmatic examples upon Modern artworks, preferably from the first half of the twentieth-century—Formalist and Abstract Expressionist works, poetry and literature by the likes of Joyce and Mallarmé. These examples are invoked to justify the freedoms of late modern, demonstratively instrumental, political, and morally enmeshed works.[23] Insofar as contemporary artworks are defended from onslaughts by way of theories, arguments and examples from Modernism’s hay day, ‘the autonomous artwork’ may denote a closed epoch; it was an important concept, but perhaps it is mostly irrelevant for contemporary works? In any case, if it is going to be dismissed, or if it is still a viable notion for today, the tangle of conceptions needs to be thought through carefully once again, and, as well as reflecting over how the various understandings intertwine, their claims and arguments should be tested in light of today’s art.

Problemstilling and thesis statement

Can the disparate understandings of ‘the autonomous artwork’ be laid out, discussed, reflected over, and a revitalized synthesis be constructed, which could be relevant for contemporary artworks?[24] My thesis-statement is that in light of post-Duchampian and demonstratively political-instrumental artworks, the field of possibilities has narrowed for what ‘autonomous artwork’ should mean. With a narrowed meaning, the expression ‘autonomous artwork’ can regain relevance for today’s artworld. The narrowed meaning I suggest retains some sceptical insights as expressed by Kant himself, but also some insights from his many interpreters. Furthermore, the notion of ‘an undecidable character’ can help overcome some of the problems experienced with ‘the double character’ of artworks.

Overview of chapters

Chapter 2 begins by looking backwards, at the roots of the expression ‘autonomous artwork’. It also brings to light some of the assumptions inadvertently swallowed when the adjective ‘autonomous’ is joined with ‘artwork’. The backwards glance resumes in chapter 3, with what ‘autonomous artwork’ meant in light of Kant’s judgment of taste, which provided a veritable deluge of “building blocks” for establishing a whole field of autonomy for art. Chapter 4 follows up with an alternating account and discussion of the most prominent understandings of ‘autonomous artwork’ prevalent today. These understandings tend to selectively appropriate and build upon aspects of the Kantian synthesis of what ‘autonomous artwork’ entailed. Admittedly, the task of chapter 4 is overly ambitious; some of the discussions are sparse, yet I see no way around it if one is to overcome the myopic problem addressed on page 4. In order to comply with the limitations of 115 pages, I strategically choose to treat in a more cursory fashion the discussion of the artwork’s autonomy as being situated within the art institution; and choose instead to focus on the Kantian legacy, which entails that the work be viewed in light of the artist and the receiver, and the ontological approaches with both weak and strong commitments. The art institution will nevertheless be ubiquitous throughout this paper. The chapter concludes by constructing a provisional synthesis of the moments of autonomy for contemporary artworks. Meanwhile, some of these moments are still problematic. Chapter 5 examines and discusses some of the views of Heidegger, Blanchot and Derrida. The goal is not to exhaustively present and discuss the aesthetic philosophies of these three thinkers—115 pages do not provide the scope for that—but to examine how they understand the artwork as autonomous. This can shed more light on the problem-points of chapter 4: the assertion that material and formal features are primary and thus grounds for asserting the work’s autonomy; the assertion that the artwork’s autonomy is based on the fundamental incommensurability between the artwork’s symbolic form and the symbolic forms of speech and language; and the notion of the artwork’s double character. As the chapter wanes, a second tentative list of moments of autonomy is suggested, which is a synthesis of thoughts from these three thinkers. Where they disagree, Derrida’s undecidability thesis is given preference. Chapter 6 takes the two synthesis of moments from chapters 4 and 5, compares and discusses, and then hazards a third hybrid of moments of autonomy for the contemporary artwork. It is, I feel, a sober, careful and justifiable list that does not make large claims about the work’s autonomy, but what remains is highly significant. Chapter 6 concludes with the following question: Is the project—of constructing a conception of ‘autonomous artwork’ that would be justifiable for contemporary artworks—obsolete? I conclude in chapter 7 with a review of the main points.

Throughout the paper, I have tried to address the problem expressed by Friedrich von Schlegel, that in the so-called philosophy of art, one of two things is usually missing: either the philosophy or the art. Prompted by Schlegel, I have, where possible, integrated the matters under discussion with talk of specific artworks, many of which are in the Astrup Fearnley Collection, Oslo. My hope is that the works will not be reduced to “mere illustrations”, but that they can be occasions for thinking through the problems of ‘the autonomous artwork’.



[1] Throughout this paper I use Arthur Danto’s coinage “artwork”, with “work” as synonym.

[2] Kant’s Critique of Judgment will here to fore be referred to as CJ.

[3] By “artworld”, what I mean is a microcosm generally understood as the social, economic and political domain in which artists both work and find support. It is a domain made up of many people who cooperate by means of shared conventions that allow them to coordinate their activities. Through repeated cooperation by people who are similar enough in their function to be considered the same, we can speak of an artworld. (See, for example, Becker, 1998, p. 148.) According to some thinkers, it is qualitatively different than the rest of culture; according to others, it reflects the general culture and is overwhelmingly connected to it. Some thinkers would make a distinction between artists on the one hand, and the rest of the artworld on the other. In this paper, “the artworld”, “the art institution(s)” “the field of art” are treated as synonyms. Within the artworld there are many institutions (e.g., museums, art publishers, but there are also habitual practices which are institutionalized).

[4] By “political” I mean broadly “life in the polis”: political, religious and other everyday concerns.

[5] I allude to the CIA’s involvement in Modern Art during the Cold War, as addressed by Frances Stonor Saunders engaging book Who Paid the Piper? London: Granta Books, 1999. Saunders tells how the CIA heavily supported non-figurative visual art, abstract music, Modern Art theory and cultural activities that could fulfil Cold War propaganda purposes. Her chapter 16, “Yanqui Doodles” deals with the CIA’s involvement in MoMA.

[6] It starts near the passageway between the lobbies at Jernbanetorget Metro Station.

[7] Spring 2005, the Bergen Kunsthall mounted Time Suspended, which deals with the theme refugees and human rights: the challenge of longstanding conflicts. This exhibition is a collaborative effort on behalf of video-artists and human rights organizations such as the Raftos Foundation. As well as viewing videos dealing with the Israel/Palestine conflict, the Kunsthall has “curated” a seminar series by non-art professionals (politicians, journalists, judges, human rights activists and researchers) to present their views to gallery-goers.

[8] A good example is Everyday Aesthetics, Astrup Fearnley Museum of Modern Art, Oslo, 27.09.03-30.11.03. Among the artists represented: Louise Lawler, Sherrie Levine, Richard Prince, Jeff Koons and Charles Ray. For anyone familiar with contemporary art, these names will indicate Postmodern approaches.

[9] I use the term “late modern” rather than “Postmodern”. By this, I refer to ideologues who might be considered Postmodern, in the sense that they want to deconstruct the distinction between the domains of art, morality and epistemology: Jacques Derrida and Roland Barthes, but also American authors writing in the October Book series, MIT Press, such as Douglas Crimp, Hal Foster and Rosalind Krauss.

[10] By ‘Kantian’, what is referred to is the many interpretations of Kant’s Critique of Judgment, be they focused on the psychological aspects of aesthetic experience or the formal and expressive dimensions of aesthetic experience.

[11] What I mean by “Copernican turn” is the insight that it is not what the artwork holds for a individual to discover, but what the individual brings to the work, both a priori features and attitude that constitute and determine their aesthetic experience. (See chapter 2).

[12] A good example is the “ Cultural Studies” movement, where the traditional study of Art History is augmented or displaced with Visual Cultural Studies, more akin to social science. Contrast, for example, Jonathan Harris’ The New Art History: A Critical Introduction (London: Routledge, 2001), with Ernst Gombrich’s classic The Story of Art: Gombrich’s chapters follow chronological history, while Harris’ book is organized according to themes dealing with capitalist modernity, feminism, the subject’s identity, sexuality, structures and meaning in art and society.

[13] Here I am stating an important assumption. When Duchamp’s ready-mades attained the status of art, a revolution in the artworld occurred: Either the artworld entered a new dispensation (something that before could not have been art now could be art), or something was realized about art that never before had been apparent, namely, that calling something art is the primary aesthetic judgment. And ‘art’ is increasingly understood as being, in some respects, a proper name like “George” (this will be discussed in chapter 4). Meanwhile, in asserting that anything can be art, I am also already assuming an important sort of autonomy for the artwork, and will throughout this paper—that in being able to be anything, the work is independent from the necessity of being something, as was traditionally understood as a criteria for art, prior to the advent of non-figuration. Accepting ready-mades as artworks entails already having accepted a modicum of autonomy for the artwork.

[14] Bowie, 1995, p. 34.


[16] Frigstad, 2003, pp.37-62.

[17] An important exception is Peter Bürger’s excellent “Critique of Autonomy” in Kelly, vol. 1, 1994, pp. 175-178, where he mentions at least 8 different construals. Still, there are many he does not mention. This problem reveals Kant’s wisdom in describing the autonomous field as having many moments. (See chapter 3).

[18] Translation: “In Defence of the Aesthetics of Autonomy—Correctly Understood”. As far as I can see, what I call the myopic problem is perhaps a necessary evil, and perfectly understandable in relational to the Modern research-university scholars’ in-depth focus on one small segment of a field, which begets the highly specialized ‘fag idiot’ or ‘nerd’ as we would say in English. I am not trying to knock Kittang, as I feel that his New Critical approach is one of the most valuable understandings of the work’s autonomy; it is just too limited.

[19]; Nyberg, Jan, “Den uskjønne kunsten”, Bergens Tidende 23 January, 2004; Rosenbergs, Göran, ”Ambassadøren og antisemittismen”, Bergens Tidende, 24 January, 2004.

[20] Ibid.

[21] An excellent example was in 2003, when art-students, under the banner of “kunstens viktig funksjon: å skape offentlig debatt” (art’s important function: to create public debate), broke into a private house (in disrepair) and started altering it. They admitted breaking the law, but “art’s important function” was above the law (in this case, the function was to point out the inconsistency that, while it is unlawful to let one’s car rot on the roadside, it is lawful let your house fall into disrepair). (“Protesterer med kunst”, in Bergens Tidende, 22 October, 2003, p. 2.)

[22] ‘Language game’, in this context, means all the things expressed in phrases such as purposiveness without purpose, taste, the powerlessness or ineffectuality of art, the incommensurability of aesthetic phenomena to concepts, or the autonomy of reception, etc., the web of Modern notions used to defend and justify the artwork.

[23] Examples are legion, but three famous examples suffice: Robert Maplethorpe’s X Portfolio was defended against obscenity charges by reference to his use of the “golden mean” and formal compositional features, the autonomy of the institutional setting, “disinterested interest”. (See Crimp, 1993, pp. 6-13) To this Arthur Danto exclaimed, “[These museum directors] are arrogant Kantians who treated these extraordinary images as formal exercises…” Arthur C. Danto, "Censorship and Subsidy in the Arts," Bulletin of the American Academy of Arts and Sciences 47/1 (Oct. 1993), pp. 25–61.

André Serrano’s Piss Christ defended by Lucy Lippard according to the dictates of “art-internal” criteria. “The work is a large colour print, 60 x 40 inches. It is a Cibachrome print. This means it is glossy with deeply saturated colours, and its surface is very delicate, easily ruined by a fingerprint or a slight speck of dust. The image is not recognizable as a crucifix floating in the artist's own urine. The jar cannot be read from the print. Rather, a crucifix is presented in a golden, rosy medium within which constellations of tiny bubbles have been frozen in space. For all we know the crucifix could be suspended in amber or polyurethane. The formal qualities of the work are, in fact, quite mysterious and beautiful.” Lippard also used arguments based upon the title’s ambiguity, the artist’s private expression, the artist’s stated intentions and the context of the autonomous art institution. is a congressional transcript record.

When Sally Mann’s Immediate Family was accused of paedophilic content, they were defended with the claim that all the receiver is doing is looking in a mirror of her own values. (See ch. 4).

[24] By ‘contemporary artworks’, I choose to limit my focus to works such as are found in the Astrup Fearnley Museum of Modern Art, Oslo.