The Failed and the Inadvertent: Art History and the Concept of the Unconscious

The Failed and the Inadvertent: Art History and the Concept of the Unconscious

by James Elkins


The history of art historical responses to psychoanalysis has yet to be written. Art historians have imported a wide variety of psychoanalytic concepts, and psychoanalysis continues to be a major interpretive resource for the discipline of at history. But beyond the core of art historical texts that are directly and explicitly influenced by psychoanalysis is a much larger, and I think more important class of texts that do not cite psychoanalytic concepts, but would nevertheless not be possible without psychoanalysis and especially the fundamental concept of the unconscious.

This paper examines the ways that the idea (or notion) of the unconscious affects current thinking about the control artists have over their works; I argue that in this more general sense, psychoanalysis has tended to help art historians take away artists’ control and awareness of their own work, replacing it with the model of artists as workers largely unaware of what they do. Against this I argue that artists who are imagined to “preside over their work with their eyes open” can be more interesting subjects, both historically and psychologically.

The unconscious remains problematic. It has always been indispensable to psychoanalysis, and in some ways it is the essential precondition of psychoanalysis itself (Freud thought as much, and called it his most important discovery; see Freud [1900, 611]). Though it is still possible to speak of a single “unconscious”—there have been no theories that begin by claiming Freud was fundamentally mistaken in that one central tenet—the varieties and schismatic versions of the unconscious are sometimes so radical that only extended qualifications can demonstrate their allegiance to the originary concept. Even within Freud’s writing there are differences between the latent unconscious of the earliest writing, the formal “Ucs.” of the first topography, and the more generous “Unconscious” of the second. The concept of the unconscious is surprisingly insinuant, and in particular it is at once heretical (prone to radical and continuous critical revision), peripatetic (apt to travel to neighboring disciplines) and metamorphic (likely to be unrecognisable in new contexts). Ideally, a heresiology, itinerary and metamorphology of the unconscious would be required to demonstrate where it appears under other names (or under no name at all) (Laplanche, 1976, 1989, 5–16).

But the vagaries of the concept, central as they must remain for psychoanalysis, are less my concern here than their appearances in art historical texts. What chiefly interests me, and seems in need of investigation, are two traits of those art historical texts that I will loosely describe as psychoanalytic in tone, whether they are authored by historians or analysts: (a) they ascribe a large component of artistic production to the unconscious, and (b) they value the unconscious over the intentional. In short they propose—often “unconsciously”—that what is important about artistic creation is precisely what is unconscious.

This idea seems to me to be mistaken. It infects an wide range of art historical texts, from those whose indices list “Freud” and the “unconscious” to those that speak loosely about intention and remain silent on theoretical forebears. It appears most insidiously and commonly as an unargued assumption in texts whose authors would not describe their works as psychoanalytic or even psychological. What I have in mind here is not a comprehensive refutation of this idea, but three more local explorations that I take to be decisive for the possibility of sensible exchanges on this subject: (a) the case against the privileging of the unconscious in psychoanalytic art criticism, taking Dr. Liebert’s study of Michelangelo as an example; (b) the argument against the more common kind of text in which psychoanalysis is not mentioned but the privilege of the unconscious is nevertheless decisive, with an uncommon text, Michael Fried’s Courbet’s Realism, as the example; and (c) a reading of Cézanne intended to suggest how interpretations of unconscious imagery and intended forms can be brought together in a fruitful and nonrigid manner.

If there is a scholarly literature that deserves execration, it is uncreative psychoanalytic art history. Mondrian’s alleged exposure to the Urszene, Cézanne’s masturbation or the hidden labia in his landscapes, Leonardo’s traumatic “feeling of loss of penis,” and Vermeer’s ineffective “barriers” against his own concupiscence, are readings guilty of the most irresponsible cutting of context, regardless of their potential truth—and that is my first objection to the high value put on the unconscious (Niederland, 1976; Reff, 1962; Eissler, 1961; Kramer, 1970; Geist, 1988). The richness and nuance of historical research are excised in favor of violently reductive interpretations, and the result is artwork that is less interesting than it had been. To see repression in Vermeer, one cuts the sum of sensitive accumulated historical connections that have been patiently built around the name “Vermeer”; to see labia in a Cézanne landscape, one cuts the landscape with one’s eyes until nothing remains but the fetish and the excuse for presenting it. Leo Steinberg is best on this point. In a critique I will discuss below, he takes Robert Liebert to task for seeing a family drama in Michelangelo’s Last Judgment. It is “inappropriate,” Steinberg says, to collapse the “nearly four hundred” figures in the Last Judgment to a “three-way altercation” between Christ, the Madonna, and St. Bartholomew, and “to have St. Bartholomew out to knife the Madonna, with a rebarbative Christ dispatching his disciple to hell—and all this because the artist’s alleged oral deprivation in infancy continued to fester” (Steinberg, 1984a, 44). I do think that the virulence of this writing needs to be taken seriously, entirely apart from the historical and formal reasons Steinberg has for doubting part of what Liebert says. Steinberg’s energy is directed at the toxic effects of the strong, simpleminded stories that are forced into our contemplation by such readings.

Most criticism of psychoanalysis in art history has centered on the epistemological problems of explanation and evidence that are inevitable when the artist is no longer available to be quizzed about his youth (Ricoeur, 1970, 170 ff.). But questions of interest should not be overlooked: would we go to a museum to see a documentation of defense against multiple primal scenes? Or rather to see a painter whose meanings stretch from the asymmetries of Manet to the topography of the Dutch landscape? We value the Wolf Man, the Rat Man and others because of the lucidity of Freud’s detective work (Ginzburg, 1980); but we do not value Mondrian because there’s a chance he might have been documenting a defense against primal scenes.

Freud started these difficulties, of course, with his own aesthetic writings, in particular Leonardo da Vinci and a Memory of his Childhood. The book’s strangeness is not adequately explained by Freud’s ambivalent “self-description” as Leonardo, nor the fascinating manner in which Leonardo’s Renaissance science is both deprecated and explained by means of an uncertain fledgling modern “science,” nor by the troublesome epistemology and its well-worn narratives of vultures, Egyptian gods, and passive homosexuals, nor by the colloid of genres (biography, pathography, psychoanalytic reconstruction, art history, connoisseurship, synopsis, mistaken philology, aesthetics), and it may be that Freud’s text is best studied in comparison with other unaccountably “wild” texts. Nietzsche’s Birth of Tragedy is a possible parallel “case”: both have an eccentric position in an oeuvre of increasing coherence, both exult in a bricolage of subject matter, and both possess a rhapsodic, even lurid, theatricality.

The wildness is important here because it resulted in a conflicted, ambivalent, and partly unreadable account of the relation of Leonardo’s conscious and unconscious expression. Freud did not write a clear text, committed—let us say—to the sovereignty of the conscious mind. The turmoil of his Leonardo is salutary. It does not mean that the book is easily mined for method or for insights into Leonardo, but it does mean that later writers have available an example of the difficulties that should await them in coming to terms with a mind and with works that are certainly more interesting, and probably more insightful, than their own.

Shrinking Michelangelo

The two terms of my title, “inadvertent” and “failed,” and the title of this section, come from Leo Steinberg’s viperous review of Robert Liebert’s Michelangelo: A Psychoanalytic Study of his Life and Images (Steinberg, 1984a, 19484b). It is still without peer for its analysis of certain tendencies in psychoanalytic art criticsm. Steinberg takes Liebert to task for thinking that “meaning is invariably tied to an unconscious expression,” so that Michelangelo’s works are not thoughts or decisions but lapses, things that are failed and inadvertent. To Steinberg, “artists, if they are any good, preside over their work with their eyes open.”

A writer replying to Steinberg’s review objected that “artists do not stand in any privileged position vis-a-vis the interpretation of their works,” and that aspects of works that are “unconscious (read: unknown)” need not “represent lapses (read: failures).” In reply, Steinberg claims:

I have nowhere suggested that impulses from the unconscious informing the creative process must produce lapses or failures. What I said was that the specific unconscious motives Liebert assigns to Michelangelo and imputes to his works do not visibly inform the artist’s paintings and sculptures [Steinberg, 1944a, 52].

These positions are a little slippery in that they are easily misstated. Steinberg does not say that artists have absolute control or even “privileged position,” but he does say their eyes are open: they look, they see, they control what they can. And what they control is what is interesting. When it’s put this way, the intractability of the debate becomes clear: no matter how the argument might proceed, there will remain the difference between those who prefer to see control and those who do not. This is a false debate: though it seems to be about methodology, historical evidence, hermeneutics, and models of the psyche and creativity, it is actually a matter of preferences.

I take it that this issue is, in fact, unarguable, and I have put it in the context of this exchange because the quickness with which the two sides rally their arguments underscores its futility. Where there are sides, I side with Steinberg: the texture of thought and the kinds of challenge that are required in dealing with intimidating artists who are in control of their means provokes me more than do the ways of thinking required in psychoanalytic readings. In psychological terms, the two are distinct sources of pleasure—one, for me, slightly less engaging than the other.

Other issues raised by Liebert’s text can be argued more easily. The nudes in the back of the Doni Tondo are also a vexata quaestio within art history. They may be shepherds, angels, sinners, neophytes, Neoplatonic figures for love, or mankind ante legem—no explanation is fully satisfactory. Liebert proposes that they are youths “engaged in homoerotic interplay,” acting Michelangelo’s fantasy of internalized femininity. But that fantasy must have been unconscious, since the tondo was commissioned for a wedding, and identifiable homosexual imagery would have been scandalous. For the sake of argument, let us assume that Liebert’s reading is entirely persuasive, and also that he did not intend malice or irony toward the Doni family: what then was Michelangelo’s conscious intention? Perhaps he meant to portray shepherds (it is not inconceivable that he would have omitted the sheep), or pagan nudes, or figures for the soul. The strangeness of the figures—and their interest for generations of critics—would then be a failure. Unconscious forces would have impelled him to make some ineffective excuse as he worked in order to keep the truth of his presentation from his Ego: “Nudity is permissible in pagan figures,” or “Shepherds often sported with one another.” My point here is not that Liebert is mistaken, but that the Michelangelo he gives us is less interesting simply because the dynamic of conscious and unconscious is so crisp. Art historical meanings have become displaced vehicles of unconscious forces. Consider instead—again for rhetorical purposes—a Michelangelo who blended angel, shepherd, spiritello, neophyte, sinner, and homosexual. Such an act could not be described by a polarity of conscious and unconscious, but by a web of intended acts. It would be an accomplishment rather than a close call, and it would have the interpretive advantage of allowing us to ask how its half-dozen elements are related rather than imagining a rigid puppetry of drive, fantasy, and repression.

The Drunkenness of Noah in the Sistine has long puzzled historians because it presents all the figures nude, and because it does not show the Shem and Japheth walking backwards (as Genesis 9:23 describes). To Liebert this is evidence that “the vision of the fallen father… must… be masked in the artist’s mind, introducing confusion” (Liebert, 1983, 41). How can the psychoanalytic account allow for the possibility that the “confusion” is conscious, and that we are dealing here with a controlled ambiguation? I ask this question rhetorically. The Drunkenness of Noah is an typical case because the question of control cannot be easily adjudicated. There are a number of things Michelangelo could have meant. We can imagine, for example, that if Michelangelo had wanted to say that none are without sin, he might have chosen to represent the heads unturned. Or if he wanted to induce shame in the spectator commensurate with the shame of Noah, he might achieve that by showing us more naked bodies than the story required. Or if he wished to represent his own ambivalence toward naked men, unexpected nudity would have served the purpose. In making this list of hypotheses, I have mimicked the kind of art historical texts that do not mention psychoanalysis directly (which we will consider below), in that words such as “meant,’ “wanted,” and “wished” can be taken to mean conscious or unconscious intentions. How can the two be distinguished? It seems to me that our only way of deciding is to look for programmatic departures from the expected story. If all three sons were equally naked, if they stared equally at their fallen father, then we could say with confidence that Michelangelo meant to bend the story. But of course the scene is not uniform: there are different gestures and different displays of nudity. So here, as in many other cases, the choice is between a Michelangelo who says something complex—something, I think, that involves all the meanings I have named—by “mistake,” as a puppet reacts when its strings are yanked, and a Michelangelo who says those same things by pulling the strings and orchestrating the drama himself. Why assume the former? Or again, in psychological terms, What motive is there for preferring the former?

The unnamed unconscious in art historical writing

I turn now to the strange but common kind of art historical writing that is clearly psychoanalytic without openly declaring its allegiance. Such writing does not study the relation beween conscious and unconscious expression, as Freud did in his own aesthetic writings, and it generally refrains from mentioning specific terms of the Freudian metapsychology. Instead the texts simply concentrate on descriptions that the artists would have reacted to in the way that Freud says unprepared patients react to interpretations of their unconscious wishes: with amazement, suspicion, and anger.

An essay that goes partway in this direction is Julia Kristeva’s delicately written “Giotto’s Joy” (Kristeva, 1988, 34, 36). Though it introduces explicit Freudian and Lacanian theory (the latter source is not acknowledged), it does so primarily in order to be able to say that color, unlike form, requires a special hermeneutics. Color, Kristeva argues, is “a pressure… linked to the body proper,” “a physiologically supported drive,” and it therefore needs a special economic analysis that is out of the reach of art history. Having introduced a “triple register” of color theories, she turns away from theory and toward history and semiotics. Her essay has much more theory than the average American essay, but the tendency is similar: to make use of Freud while remaining independent.

In most recent art history, the traces of Freud are difficult to locate even where his influence is decisive. An exemplar of this kind of writing is Michael Fried’s book Courbet’s Realism, the fruit of a deeply layered encounter with psychoanalysis. Fried seldom invokes the Freudian lexicon (his footnotes are peppered with various feminist and literary-critical sources), but his entire enterprise in is avowedly an exploration of the visible traces of Courbet’s unconscious. Typically, the diffusion of the unnamed unconscious is in direct proportion to its declared importance, and the privilege of the unconscious in this book is virtually complete. Fried dismisses his subject’s consciousness literally in a parenthesis:

the philosophical, political and even moral connotations of realism made it all but inconceivable that a work of art… could be both realistic in effect and imaginative or metaphorical in relation to its materials. (I am convinced that Courbet himself was largely unaware of the aspects of his work I focus on in the pages that follow.) [Fried, 1990, 5]

In Fried’s view Courbet’s purposes are unconscious in the sense proposed by MacIntyre: what Courbet thought he made was radically at odds with what he actually produced. Fried’s Courbet is so desperately misguided about his enterprise that his conscious intentions have only a few point of contact with the project of Courbet’s Realism. What is important is unconscious, and the substance of the book is an exploration of the working-out, the pictorial manifestation, of that unconscious.

In accord with Freud’s doctrine of unconscious drives, the signs of Courbet’s unconscious are repetitive and conservative: they recur, seeking only fulfillment, periodic discharge and homeostasis (Laplanche, 1976). That means the sign of any given impulse in one of Courbet’s paintings is fundamentally the same as the corresponding sign in any other painting. I say “fundamentally” because the drives compete, and they interact with Courbet’s developing capacity and the particular occasions and purposes of his paintings, providing a wealth of variations: but through them all, Courbet’s “mastery” is necessarily, if temporarily, demoted to a form of automatism. Most of Fried’s time is taken in expositing signs of different drives rather than different signs of the same drives. It is not that there is no development in these motifs, and indeed their nuanced differences occupy much of Fried’s text: it is that the catalogue, the ticking off of the evidence, takes precedence over theories of development.

The principal unconscious signs make a short list. There is the left and right hand (the former with palette-sign, the latter with brush-sign), the body inclined toward the viewer, the softening and penetration of the lower margin of the painting, the “seeming physical proximity” of the painted image and the “surface of the painting,” the half-closed eyes, blurred twilight landscapes, and figures seen “as nearly as possible” back-to. Each of these is a psychological signifier—the closed eyes and tenebrous landscapes, for example, denote “consciousness on the verge of extinction” flowing outward, melting into its environment. They mark anxiety over, or denial of, the “vertiginous gulf between… painting and beholder,” coupled with “an intense absorption” in the artist’s own “live bodily being” (Fried, 1990, 65, 75, 57–59, 69).

A difficulty with this approach (not a reason to doubt its validity) is that it does not inspire a reader to go an work on other Courbet paintings. Imagine analyzing a work by Courbet that Fried does not mention, in a way sympathetic with Fried’s text. Such an analysis would involve the careful identification of the repertoire of unconscious strategies set out in the book; one would expect variations and perhaps a new solution, but nothing provocatively new. I do not mean that this is necessarily the case, only that Courbet’s Realism functions to make such a scenario seem likely, and therefore to stifle interest in further study.

Harold Bloom has praised Freud’s theory, saying it has provoked not only critics and disciplies but new theorists. By that standard, Fried’s theory is an exemplary end product, in the positive and negative senses of that phrase: it is a theory that provokes criticism and imitation more than augmentation.

In an importance sense a reader’s future understanding of Courbet is predicted for him. Courbet’s Realism, aside from its truth claims, is a strongly repressive text: it limits, at one and the same time:

1. The subject, since Fried’s Courbet is denied a priori the possibility of escaping from the drives of his own unconscious or of understanding and controlling them in terms of his own conscious program of Realism.

2. The writer, since Fried constrains himself to explicating the unchanging mechanics of the unconscious as he finds them in Courbet’s paintings. The unconscious, Freud said, does not know time. It does not sleep, it obeys only the primary process, nothing from the outside world disturbs its tidal rhythms. A text that takes the unconscious as its subject enslaves its writer to a similar regimen. The analogue in clinical practice might be a patient who does not go in for treatment: barring trauma, the expression of his drives will follow a certain routine. In the case of phobia, for instance, the symptoms may change as anti-cathexes immure the conscious awareness of the troubling idea—but the mechanics of the neurosis will remain static (Freud, 1911).

3. The reader, because we are compelled in turn to follow the same mechanics in our reading and in our reimagining. We are disallowed the creative possibilities of alternate readings, the pleasure of undiscovered interstices, the hope of further developments, and the dialectic interaction that are possible with fictional characters in novels (or accounts of artists written further outside the psychoanalytic paradigm, such as those in Fried’s other books). Freud was aware of the parallels between his work and the “mental processes” described by “imaginative writers”; one trait that sustains the freedom of our relation to fictional characters is the novelists’ mix of conscious and inadvertent explanations for their characters’ actions. Some things we know about Ivan Ilych he does not know about himself, but many other things are shared, and that provides a freedom for the reader that an unrolling of unconscious desires, no matter how pictorially compelling, must deny.

This triple slavery is, I think, a consequence of not developing three corresponding freedoms:

1. To the artist corresponds the freedom of conscious control. To the degree that an artist is imagined as presiding over a work with his eyes open, he or she is free to accept and deny, propose and dispose.

2. To the writer corresponds the freedom of analyzing that conscious control. Without the thematic of the conscious or of the relation between intention and inadvertency—and this latter is one of the most rigorously excluded topics in Fried’s book—the writer loses the possibility of representing whatever freedoms are ascribed to the subject. Those freedoms, fictional though they must be, can provoke mimicking freedoms in the narrative.

3. To the reader corresponds the freedom of reading a narrative about consciousness. Whatever narrative the writer has found will open similar possibilities for the reader. Imagine, for example, an analysis of the relations between Courbet’s Realist intentions and his unconscious drives: it would be possible to adjust the writer’s findings (as in my rhetorical adjustment of Liebert’s reading of the Doni tondo), rather than simply assenting, balking, or fleeing.

Mathematicians analyze equations according to their “degrees of freedom”: essentially the number of variables that are not fixed. When a system has more than one degree of freedom, it typically cannot be solved, and the business of analysis is then to find restrictions. But a system with a surplus of freedom is no more interesting than one that is wholly determined: the former is useless fantasy, the latter unchanging petrification. I find Fried’s book on Courbet compelling, largely convincing, and very forcefully argued: but I also feel the pressure of its repressions and exclusions.

Cézanne’s unconscious motifs

The readings by Sidney Geist and Theodore Reff, which I mentioned briefly in connection with Cézanne’s alleged masturbation and fantasies of landscaped labia, are the most recent examples of an escalating tendency toward the explicit and narrow. Earlier readings, beginning with Roger Fry and continuing through Meyer Schapiro, were more circumspect in their psychoanalytic overtures (Schapiro, 1952). (Fry in particular was open-minded about the issue of control. To him, Cézanne was “ignorant” of his deformations, while at the same time his “intellect… claimed its full rights.” [Fry, 1952, 48, 53; Spector, 1988, 49ff.]) But instead of reviewing that history, I want to briefly propose a field of unconscious motifs that has not been developed in the literature, in order to make a closing argument concerning the possibility of integrating the unconscious into art historical analyses.

Schapiro has described Cézanne’s “detached, contemplative relation to the world,” his themes of solitude, his love of things “beyond approach,” and above all his “abandoned catastrophic landscapes” that are “intraversible,” “inaccessible,” and “inapproachable” (Schapiro, 1952, 78). There is a structure to this detachment, consisting of a dozen or so repeated forms. We can look quickly at seven of them, as they appear in three landscapes done in the years 1869 - 73. The Picnic (Paris, Musée d’Orsay, c. 1869) has a form that may be called a splayed center—two trees fall away at either side, as if they have been pushed aside. Another form might be called an abyss, though the term is a little strong for this painting: the ground gives way toward the front, so that the figure of Cézanne is squatting on a declivity. The result is that the ground under our feet is surely much lower, perhaps on a slope or in a gulf. There is also a wave in the painting: a surge of rounded organic forms at the upper left, pinching the sky into a cramped space above them, rolling downward and toward the right, weakening as they go and deliquescing into a thin cloud.

If this reading seems excessive, consider L’Estaque, Melting Snow (1870-71): again there are the bent trees, the splayed center, this time voided, and again the wave that begins with thick puffed leaves, pinched into the top left corner of the canvas, and melting into a black, rolling horizon. L’Estaque, Melting Snow also contains an abyss—a real one this time. As Schapiro notes, “there is no foothold for the spectator”: if we were to step into the picture, we would tumble onto the house far below (Schapiro, 1952, 38). The abyss is one of a reperoire of devices that preclude our imaginative entry into the paintings. Sometimes the abyss is a barricade, other times a wall in the distance. Something of the wall also occurs here, since the field slopes steeply upward, as if in Oriental oblique projection or a child’s dream of inaccessibility. So, to be exact about it, we are not “excluded” from the picture but trapped in middle distance. After falling onto the house, we can neither clamber back up the slope nor climb all the way to the distant farmhouse. Cézanne’s landscapes can present versatile and formidable challenges to uninvited ingress and movement. And finally, the same painting contains a motif I will call the weak right margin. If you place a card over the painting, and slide it to the left until only a thin strip at the right becomes visible, that strip appears remarkably weak. It is homogeneous in aerial perspective, value and color, and its forms are dissolutive. Weak right margins are sometimes even more pronounced, and in sketches Cézanne sometimes abruptly leaves off when he comes to the inches just short of the right margin.

The Suicide’s House (1872-73) can serve as a final example—though the exposition could be extended to cover both earlier and later landscapes. Here again, it is difficult to enter into depth: we begin on what appears to be a level path, which slopes down so precipitously it is lost to sight. That abyss (which is mirrored by the steep slope coming down from the upper left) is then followed by an abrupt turn. The town, to which the path evidently connects, is pathless. It forms a barricade or wall in the distance, because it is impossible to imagine how it might be walked through. The berm is the only clear platform in the painting, and it does not look like a spot to relax. The distant town also forms a fragmented center, a place near the middle of the canvas that receives particularly intense attention and takes on the look of an imbricated armour or a geode of small crystals. Cézanne commonly paid special attention to such passages and, as we know from Earle Loran’s photographs, sometimes expanded their size so he could lavish them with detail. And The Suicide’s House has a splayed center, though not as strong as those we have already seen: the three windows of the central house tip to the left, and a moundlike house slips away to the right. And though it is especially attenuated, there is an unmistakable wave here as well: the swollen foliage at the upper left, crowding the top-left corner in trecentesque fashion, gives way to a lower and gentler horizon toward the right. And last, a severely compromised right margin, nothing more than a flattened sheet of sod, contrasts strongly with articulations of form and color along the other three margins.

In an exposition of this type I cannot hope to make a convincing case for these forms, and I invite the reader to compare paintings such as the Road at La Roche-Guyon (Northampton, Smith College, 1885), The Bay from L’Estaque (Chicago, Art Institute, c. 1886), Mont Sainte-Victoire (New York, Metropolitan Museum, 1885-87), and The Bibémus Quarry (Essen, Museum Folkwang, c. 1895). But what I have in mind here is something other than an exposition of an unconscious thematic; let us consider these observations instead as raw material for such an analysis. An initial question might therefore be: Are we in the presence of a comprehensible set of unconscious drives? Do these forms add up to a coherent whole?

The abyss, barricade, fragmented center, and wall in the middle distance certainly comprise a family, and each expresses the kind of detachment of which Schapiro spoke. The “intraversible” barricades and abysses, and the “inaccessible” middle distances and fragmented centers may signify an unconscious desire to keep the viewer at bay. The paintings deny us imagined wanderings and sometimes even the right to see where we might go. Yet there are clear differences between the members of this “family”; and other forms are not as easily related as these. The splayed centers also speak of psychological distance, but more ambivalently and less intelligibly. A splayed center is like a trap, set to spring closed, and it is also like an invitation. What is pushed aside is in tension, and that tension is readable as a figure for the painter—as if to say, I am pushing myself aside for you. The open V is a little theatrical (it is like Baroque swags, curtains, or repoussoir figures), and it conjures another painting, “prior” to the one we see, whose subject would have been a small copse. What we see in The Picnic are usurpers of that proper subject, and there is both psychological detachment and mistrust, since the landscape may reclaim its proper place—as indeed it does in later figureless motifs.

All this is self-contradictory enough. What could a weak right margin say about exclusion? And what could it mean to compress a landscape so that it pushes the sky against the top margin? Can a comprehensible meaning be assigned to the “Baroque” flourishes at the upper left of these canvases? Or to the sagging wave? And is an expanded, fragmented middle a sign of exclusion? The intensely seen little town in The Suicide’s House and similar paintings is in the upper center, the place which in a portrait is occupied by a face. Are we then to see figural allegories in the landscapes? Is the splayed V a sexual invitation or trap?

And does it make sense to claim that these forms are uniformly unconscious? Rilke thought that “the two processes, that of visual perception… and that of the… personal utilization of what is perceived, counteract eachother [in Cézanne] so that neither is too conscious.” Every painter knows that one can be dimly aware of forms like these, or not see them at all until someone points them out. Cézanne may have pondered his attraction to crowded centers, and he may never have given his right margins a thought. Here as everywhere else, we have no criteria for separating the recognised from the unseen, and therefore we have no justification for lumping them in one camp or the other.

The incommensurability of these forms speaks of “radical alterity”: it is exactly what one would expect of the unconscious, whose contents are unknowable and whose utterances must be disguised. In a Freudian dynamic, the abyss, the wave and their companions would require reference to “deeper” causes: perhaps the primal fantasies (castration, seduction, intra-uterine existence, primal scenes) or the aftermath of infantile sexuality and the Oedipus complex. These sources can still be implied in texts that remain at some distance from psychoanalysis. Fried names sexual and somatic impulses without invoking psychoanalysis, and Kristeva’s “Giotto’s Joy” stresses the somatic origins of color and lets more specific Freudian doctrines fade into the background. In the clinical setting, it might be possible to resolve them, and we could make similar attempts by recalling the salient facts about Cézanne’s family. But psychoanalytic art criticism, as Laplanche has said, is a discipline distinct from psychoanalysis, and one condition of its separateness is the lack of the clinical exchange. In addition, assigning drives to their sexual and familial roots produces the kind of impoverished, almost mechanical artist we have seen in Leonardo, Michelangelo, Mondrian, and others. Freud’s aesthetic investigations are replete with undecided issues, one clanking against the next. In the Leonardo there is a phallus, a kite, an Egyptian goddess, a parataxis, a history of homosexuality. In the essay on Michelangelo’s Moses, there is Oedipus, Morelli, Julius, and a finger on a strand of hair. Even the theoretical essays whip up odd concoctions. The second essay on the unconscious compresses blackheads, penises, schizophrenia, and philosophy. A strongly coherent theory in these arenas must be modelled on our fiction of transparent unified conscious intentionality—exactly the wrong theoretical model for ideas that we mean to say are unconscious.

Why, then, press the signs of an unruly, unknowable psyche into the domestic coherence of a conscious persona? Fried and Kristeva, among others, decline dogma in this fashion, but both hope to obtain something coherent from something puzzling. I mean to suggest that a further possibility exists. We can allow our artists to be complex, provocative, and genuinely intriguing, and at the same time use the possibilities that psychoanalysis has opened to us, if we refuse to be seduced by the etiologies and personal narratives of clinical practice. The result is not confusion or impressionism, but complexity, and its open-endedness acknowledges our love of questions whose answers would be crass.


I have tried to cast doubt on psychoanalytic art historians’ penchants for theory and closure. The most promising psychoanalytic art criticism is gentle but precise with its allusions to the Freudian corpus, circumspect in its archaeology of intention, and vigilant of the historical nuances and biases regarding the unconscious (for example, Leja [1990]).

Here are three kinds of questions that can be addressed to the idea that what is important (pleasureful, noteworthy, privileged for investigation, fertile for inquiry) about artistic creation is precisely (originally, preëminently) what is unconscious:

First, is this correct? How can we decide what is conscious and what unconscious in a work? According to what criteria, with what history of judgments, do we value unconscious over controlled elements?

Second, is this helpful? Do we understand our artists better? Can we now read the works with greater mastery, eloquence, or precision? Does psychoanalysis aid or enrich our “native” discourse?

And third, is this interesting? Are we seduced by new difficulties, new terms, new categories? Are we given fresh reasons to study paintings, to take pleasure in them, to learn to love them?

Since the subject is psychoanalysis, it is not inappropriate to end by suggesting that there are opportunities here for self-analysis. Why does art history’s fascination with psychoanalysis continue unabated? The simple fact that I have written this essay shows my own interest, even though I do not disagree with Derrida when he says that “we never dreamed of taking seriously… the metapsychological fable” (Derrida, 1972, 117). It can hardly be the case that we import Freud merely in order to be able to describe a wider range of phenomena: we also desire greater complexity, and with that desire we are once again enfolded in the dubious blanket of psychoanalysis.


BRION–GUERRY, L., 1978. The Elusive Goal. Cézanne, The Late Work, ed. W. Rubin. New York: Museum of Modern Art.

DERRIDA, J., 1972. Freud and the Scene of Writing. trans. J. Mehlman, Yale French Studies 48: 73–117.

EISSLER, K. R., 1961. Leonardo da Vinci: Psychoanalytic Notes on the Enigma. New York: International Universities Press.

FREUD, S., 1900. The Interpretation of Dreams. S.E. 5.

FREUD, S., 1911. The Unconscious. S.E. 14.

FRIED, M., 1990. Courbet’s Realism. Chicago: University of Chicago Press.

FRY, R., 1952. Cézanne, A Study of His Development. London: Hogarth Press.

GEIST, S., 1988. Sidney Geist, Interpreting Cézanne. Cambrdge, Mass.: Harvard University Press.

GINZBURG, C., 1980. Morelli, Freud and Sherlock Holmes: Clues and Scientific Method. History Workshop 9: 11 - 12.

KRAMER, M., 1970. A Study of the Paintings of Vermeer of Delft. Psychoanalytic Quarterly 39: 389 - 426.

KRISTEVA, J., 1987. Giotto’s Joy. Calligram, ed. N. Bryson Cambridge: Cambridge University Press.

LAPLANCHE, JEAN, 1976. Life and Death in Psychoanalysis, trans. J. Mehlman. Baltimore: Johns Hopkins University Press.

LEJA, MICHAEL, 1990. Jackson Pollock: Representing the Unconscious. Art History 13 no. 4: 542–565.

LIEBERT, R., 1983. Michelangelo: A Psychoanalytic Study of his Life and Images. New Haven: Yale University Press.

NIEDERLAND, W., 1976. Psychoanalytic Approaches to Artistic Creativity. Psychoanalytic Quarterly 45: 185 - 212.

REFF, T., 1962. Cézanne’s Bather with Outstretched Arms. Gazette des Beaux-Arts ser. 6, vol. LIX: 173 - 90.

RICOEUR, P., 1970. Freud and Philosophy. New Haven: Yale University Press.

SCHAPIRO, M., 1952. Paul Cézanne. New York: Harry N. Abrams.

SPECTOR, J., 1988. The State of Psychoanalytic Research in Art History. The Art Bulletin 70 no. 1: 49 – 76.

STEINBERG, L., 1984. Shrinking Michelangelo. New York Review of Books, 28 June: 41 - 45.



A version of this article was originally printed in the International Journal of Psychoanalysis (1994) and has been reprinted with the permission of the Editors in Chief, Professor Paul Williams and Dr. Glen Gabbard.


James Elkins is the author is Professor of Art History, Theory, and Criticism at the School of the Art Institute of Chicago, 112 S. Michigan Avenue, Chicago IL 60603.  Elkins’s research interests include theories of images, non–art images, connections between Renaissance and modernism, connections between science and art, and the historiography of art history. His recent books include The Domain of Images (1999), How to Use Your Eyes (2000), and Why Art Cannot be Taught: A Handbook for Art Students (2001).


All his publications and a resume are available at: Direct all responses to this essay to: