Chapter 15

The French Poststructuralist Model

The necessity of a separate essay on this subject is a result of the historical development of art history and theory since the 1970s. As a discipline, art history hasn't absorbed the interests and interpretive strategies associated with Barthes, Foucault, Derrida, Deleuze, Lyotard, and others in such a way that they have become indispensable. Individual writers and ideas have been widely discussed, and many contemporary practices of art history would be fundamentally different without the developments in French criticism and theory from the mid-1960s to the 1990s, but the discipline of art history has not made them central or normative.

This complex story–which luckily isn't my subject here—has resulted in a fairly clear distinction in art writing. The historians, critics, and philosophers I will be considering are, generally speaking, not considered as art historians. They are read as parts of the history of the reception of postmodern art, and as sources for art theory, but their texts would not generally be taken as examples of art history. Foucault's discussion of Las Meninas is widely read in graduate seminars, and still analyzed in art historical accounts of Velázquez's painting (for example in Benjamin Binstock's Vermeer’s Family Secrets), but it is taken as a philosopher's theory, not a scholar's contribution. Derrida's imagined conversation on one of Van Gogh's paintings of shoes is also still read in art history seminars, but as a contribution to art theory. Barthes's Camera Lucida is assigned in photography seminars, but as theory; I think only Geoffrey Batchen has suggested it might work as art history. (In Photography Degree Zero; his identification is discussed in my What Photography Is.)

There are also ambiguous and marginal cases, where the work may also considered as art history. Hubert Damisch's books such as A Theory of /Cloud/ and The Origin of Perspective are examples; so are Daniel Arasse's books such as The Detail or Vermeer: Faith in Painting. I might also name Louis Marin or Jean-Claude Lebensztejn in these terms; readers suggested Julia Kristeva, or Gilles Deleuze on Bacon, Jacques Dupin on Giacometti, Jean-Luc Nancy on On Kawara, Jean Genet on Rembrandt, Merleau-Ponty on Cézanne, Antonin Artaud’s response to a Van Gogh show at the Musée de l'Orangerie (see the comment to this Post, below), and Jean Baudrillard on trompe-l'oeil. (Thanks to Lyle Massey, Joacim Sprung, Barry Schwabsky, John Yau, J. Karl Bogartte, and John Kaly for these.)

When I posted a draft of this essay, Kristine Stiles, Aruna D'Souza, and others questioned the premise that these writers aren't art historians. They pointed out that art historians have routinely studied French poststructural texts for decades, and so Deleuze, Foucault, and others are in effect integral to the discipline. Certainly that is true for me: the North American experience of art history from the 1970s to the present includes all the authors I've mentioned so far. From this perspective it doesn't matter that texts by Derrida, Barthes, and others are introduced in seminars as theoretical models rather than art historians, because art history has always been a mix of theoreticians and historians; English-language writers like E.H. Gombrich and George Kubler are also sometimes introduced more as theorists than historians.

But luckily—again—my theme is not disciplinary definitions, but writing; I'm making these observations not to police what counts as art history, or to decide what counts as theory and what as history, but to open a way to talk about writing. To say that Derrida, for example, works as an art historian or counts as art history in his essays on visual art is to include curricular material along with the texts art historians produce—or to put it differently, it is to include source materials along with texts of art history. In terms of writing, there are differences between French poststructuralism and what is practiced under the name art history. An imaginary dialogue of unnumbered, unnamed female speakers—Derrida's dramatis personae in his essay on Heidegger and Van Gogh—is still not viable as an option for art historical writing. (The field is open enough to accommodate such an experiment, but I can only think of one or two examples. But certainly a young art historian wouldn't be well advised to write in the form Derrida adopted in that essay.)

What I am proposing in this chapter, then, is a closer look at the kinds of writing on visual art that were done in France, beginning in the 1970s and 1980s, largely following Roland Barthes. Because the subject here is experimental writing, I am most interested in texts where the form, narrative, voice, structure, and rhetorical devices are radical or unusual in relation to some practice.

The question of writing, in these authors, is not only a matter of the expanded sense of écriture or littérature that is developed in Barthes, Derrida, and others. For several of these writers, the corpus of writing against which they measure their own work, or to which they wish their writing to respond, is philosophy, but sometimes it is also art history. Damisch's Origin of Perspective was written partly against art historical writing on perspective that Damisch had read; on the other hand, Jean-Louis Schefer's The Deluge shows his awareness of that perspective literature, but also his lack of interest—his attention is focused more on art theory.

Arguably Barthes is the central author in this thematic, but he is absent from this chapter because I have written a book on what writing meant to him at the time he composed Camera Lucida. Small portions of that are discussed in the Third Introduction. Here I'll begin instead with an outlying text, Salvador Dalí's The Tragic Myth of Millet's Angelus, for reasons I will give later. After that prologue, the discussion of French texts begins with Jean Louis Schefer.

Revised 2024

Chapter 16

Salvador Dalí, The Tragic Myth of Millet’s Angelus

I'm more than a little aware that including this book risks losing readers: those who are serious about the discipline of art history, those who are care about Dalí's politics under Franco, and those who care about experimental writing. But as a reaction against early 20th century art history, Dalí’s book is unparalleled, and it was written in 1933, making it contemporaneous with writers such as Breton, Bataille, and Lacan. It is therefore an appropriate prelude to the poststructural experiments I will be considering. (Later in this project I will look at Breton, because his fiction has more to do with writing and images than writing on art.)

I wrote a chapter on Dalí's book in 1999, in Why Are Our Pictures Puzzles?—this is a revision and abbreviation of that chapter. That entire book is available, illegally, on academia.edu and on Scribd.

Salvador Dalí’s book The Tragic Myth of Millet’s Angelus, translated into English and published by that outlet of diluted surrealism, the Dalí Museum, is a brilliant text: it’s a sustained series of utterly improbable fantasies about Millet’s painting, soberly presented as a serious scholarly inquiry. For Dalí, “serious” means psychoanalytic, and so a reader need not be too shocked when, opening the cover of the original French edition, she encounters a double–page spread of an obscene postcard. (The English edition tries to be a bit more chaste, offering a reproduction smaller than a postcard.) 

Dalí then launches into an absurd psychoanalytic theory of the image:

Let us look at it. The mother, who could well be a variation of the phallic mother with a vulture’s head of the ancient Egyptians, uses her husband, whimsically “depersonalized” into a wheelbarrow, to bury her son while at the same time causing her own impregnation, being herself the foster–mother–earth par excellence. 

The double–image” of the phallus–cactus seems to us an unequivocal allusion to the desire to castrate the spouse, who, thus deprived of his virility and reduced to the state of a simple vehicle of social productivity, can no longer form a screen, or a hindrance in the direct relations of mother–son, or the rising sun of an absolute matriarchy. In the matriarchy, the mother wishes to substitute herself for the husband by replacing him in all “situations”; in the present case, that of a wheelbarrow. Thus she would like to play, be coaxed, a wheelbarrow rhythmically balanced by her son, himself at the zenith of his “heroic” athletic university student strength during which, in a matriarchy, he goes through a very short period of maternal idolatry just before undergoing in turn the fate of his father the moment he becomes a husband.

The passage contains most of Dalí’s theory of Millet’s painting: he wants it to depict a mother who has just killed her son and is anticipating being sodomized by her husband before she cannibalizes him.

There are many reasons to dismiss Dalí’s book—too many. He was famously undependable, excessive, theatrical, egomaniacal, and frivolous: as Freud said, he was a true fanatic. The book obviously revels in what it does to the Angelus, which was once a ubiquitous icon of middle–class values and pallid romantic yearnings. The Tragic Myth of Millet’s Angelus is an extravagant self–promotion, made even worse in the American edition by the addition of The Myth of William Tell (The Whole Truth About my Expulsion from the Surrealist Group). In terms of art history, Dalí is marginal even to those historians who have sought to privilege surrealism.

And yet the book is exemplary art history. I mean that in a very particular way: Dalí’s book exhibits many of the strategies of art history, and it is irreproachably well–informed about conventions that only became widespread long after the book was written.

Art historical methods, put to perverse purposes

While most art historians learn their psychoanalysis from texts, Dalí “learned” his psychoanalysis from the source itself—from Lacan. "Learned” is in quotation marks since it is far from clear that Lacan did the teaching. Their relationship in the years 1933–36 was formative for both of them, and the give–and–take is still insufficiently studied. Dalí had read Lacan’s thesis, which was finished in 1932, but his own essay “Paranoiac–Critical Interpretations of the Obsessive Image of Millet’s ‘Angelus’” appeared in Minotaure in the pages just before Lacan’s seminal paper “Problem of Style and the Psychiatric Conception of the Paranoiac Forms of Experience.” The relation between the two is not easy to disentangle, though as Dawn Ades has pointed out, Dalí derived some support from Lacan’s notion of the “objective and ‘communicable’” nature of paranoid phenomena. Certainly not all Lacanians are interested in pursuing his thought in Dalí’s direction, and Dalí did not help matters by telling a ridiculous and embarrassing anecdote about a meeting with Lacan in My Secret Life. (Dalí wonders why Lacan is staring at him so strangely, and then when Lacan leaves, Dalí discovers he has a scrap of paper glued to his nose.) 

In the absence of a good account of Dalí’s relation to Lacan, it is enough to point to the confluence of ideas, and the fact that Dalí’s book may have been written virtually simultaneously with Lacan’s dissertation. What art historian could claim anything similar? Art historians who use Lacanian theories learn at first from translations of partly unreliable and mysteriously edited transcripts of seminars that took place decades ago. To some degree Dalí and his theorist du jour were collaborators, or at least equals.

Psychoanalysis is not the only theory mobilized in the book, though it is the only one that is named as such. The entire book is set as an exploration of the painting’s power, and it is confessional about the effect the painting had on his life (given, of course, that Dalí’s confessions are rarely what they seem). Despite the many rhetorical reasons not to believe Dalí's stories about Millet's painting, it seems the painting did actually trouble him: even if the prose is unreliable on that score, the paintings testify to an enormous amount of time spent with the image. The encounter had effects that lasted over thirty years, from the first appearance of a Millet-motif in 1929 to Gala Looking at Dalí in a State of Anti–Gravitation in 1965; it was a central preoccupation, a guiding obsession or "obsession" for most of his working life. 

When I was teaching this text in a graduate seminar, two Spanish-speaking students (one from Mexico, another from Colombia) told me that in the original Dalí's tone is wild, rough, and informal, much more than in the English or French versions of the book. That undisciplined, unacademic voice is compatible with his insouciance about history and scholarship. The autobiographical mode Dalí deploys so carelessly has only come into art historical writing as measured deconstructions of the distanced authorial voice. (As in Krauss's descriptions of Clement Greenberg's face in The Optical Unconscious.) What could be more openly autobiographical, and less guarded, than the crazy photograph of “The beginning of an erection on Dalí while being photographed disguised as The Angelus”?

Dalí is also ahead of the curve in the anti-Kantian project of describing the "power of images." He recounts an instance in which a man knifed the Angelus and was put in an insane asylum, and uses that anecdote to make a point about the painting’s unpredictable power: exactly what David Freedberg and Leo Steinberg were to chronicle over forty years later. And Dalí does them one better by providing an appendix describing the outcome of Lacan’s interview with the man, in which the patient rants about Watteau’s Embarkation to Cythera, which in turn leads Dalí to promise an entire book on the “marvelous trilogy” of the Angelus, the Embarkation to Cythera, and the Mona Lisa. (Although that book never appeared, he did write an essay  explaining “Why They Attack the Mona Lisa”; it divides assailants into “ultra–intellectuals” such as Duchamp, and “more-or-less Bolivians” who throw “pebbles,” or just steal the image.) 

For Dalí as for the surrealists images are disruptive, disorienting, or sexually coercive, and though he treats the issue carelessly he is more at ease with the confessional mode, and with the thought that images can have deep and unexpected effects, than some the most adventurous postwar art historians. 

The same could be said of Dalí’s uses of several other art historical methodologies. Peircean accounts of the mobility of meaning, and texts about performative criticism, are not often as kaleidoscopic as the interpretations Dalí offers beginning with his very first illustration. Experiments in mingling “literary” and historical writing are overshadowed, in a sense, by Dalí’s literally incredible reminiscences, his purple prose, and his giddy mixture of criticism, philosophy, history, and confession. It is a sign of the text’s overwhelming inventiveness that it even has an echo of Derrida’s discourse on Van Gogh’s shoes: speaking of the man who had slashed the Angelus, Dalí says 

one cannot fail to compare this man’s case with that of Van Gogh, whose obsession with Millet at the crucial point in his madness, forced him to copy in his own style several paintings by Millet from postcards. On the other hand, there is every reason for stressing the automatic repetition of drawings of wooden shoes in the pictures of Millet.

It’s a funny passage, not quite on Derrida’s subject, but uncannily near it in the way it plays with Van Gogh, shoes, obsession, interpretation, and violence. 

I don’t want to imply Dalí’s book is in some way a model for what recent art history has been doing. His autobiographical writing, his confessional approach to artworks, his semiotics, and his combinations of criticism and history are quite different from the forms those strategies have taken in art history. But his writing is so free, and it embraces its methods so fully, that it can be understood as a more radical version of the relatively suppressed writing in and around art history.

Dalí’s formal analyses, for example, are sometimes astonishing—they have the same hypertrophied precision that went into his paintings—and his writing style is fluent beyond what most historians could accomplish. There are a number of points in the book where his rhetorical power is used to create a force of argument that is lacking in much art history:

How then explain and reconcile this obsessive unanimity, this undeniable violence brought to bear on the imagination, this power, this absorbing and exclusivist efficacy in the reign of images? How reconcile, as I said before, this force, this fury even in the reproductions with the miserable, tranquil, insipid, imbecilic, insignificant, stereotyped, and conventional to the most mournful degree, of the Angelus of Millet? How can such antagonism not seem upsetting? No explanation can henceforward appear valid to us if it continues to count on the belief that such a picture has nothing to say or “almost nothing to say.” We are convinced that causes of a certain importance cannot fail to correspond to such effects, and that, in reality, under the grandiose hypocrisy of a content manifestly the sweetest and the most worthless, something is happening.  

Sources
Like many art historians from Riegl and Gombrich to Jonathan Crary and Alexander Nemerov, Dalí collects an unexpectedly wide range of sources. The pornographic postcard is only his first popular image; as Gombrich was to do thirty years later, Dalí collected Angelus cartoons and postcards from all around the world. He also reports on storefront displays, sets of china, and a book he had as a child. And have any art historians cited sketches they first published in Playboy? 

When it comes to questions of technique, most art historians make use of existing conservation reports; Dalí convinced the technical laboratory at the Louvre to make an X-ray of the Angelus, to see if they could uncover hidden images. (The X-ray, which he reproduces, reveals a small curvilinear form on the lower margin which he claims is coffin–shaped, confirming his conviction that the painting is about the burial of a son.) 

His research is often surprising. Having found the vaguely “trapezoidal” outline of the “coffin,” he produces a photograph of a gravestone in Quinéville which is not only black and trapezoidal, but has a relief of the Angelus sculpted in it! (Needless to say, the discovery proves exactly nothing, but it is the kind of astonishing visual material that art historians often want to find.)

Dalí is also more adventurous than most art historians in his use of scientific images, though of course he puts them to irrational uses. Aside from the images of the praying mantis (they were only later to become clichés of surrealism), there is also a reproduction of a Surinam toad, “which even today causes me to shiver,” and he discusses genetics, aviation, and other scientific and technical subjects.

The book as object

Here, too, Dalí's book is exemplary of a certain kind of experimentation. The original French edition is bound in imitation of a schoolbook, with a sharp little metal clasp. 

The American editor has added information he thought Dalí would have liked, and interpolated additions throughout the book. On one page there is a reproduction of the cover of Science, with a photograph of a praying mantis; the issue announces the discovery that mantises have a single “cyclopean” ear in the middle of the faces. “Were Dalí not so ill,” the editor remarks, “there is no way of telling what interpretation his paranoiac–critical method would make of this recent discovery.” The most important additions are two reproductions of pornographic drawings by Millet, one showing two peasants making love “missionary style,” exactly as Dalí had imagined (and painted) in reference to the Angelus. Dalí’s interest in postcards and what is now called popular culture also receives comical confirmation in the form of snapshots taken by the translator and editor on a visit to Port de la Selva and Cape Creus, where Dalí saw rock formations that reminded him of the Angelus. It’s especially apt and unexpected to see the famous surrealist rocks oddly juxtaposed with tourist photos of the editor and his wife and friends waving at the camera. 

History as hallucination

There is one crucial difference between Dalí's use of all these methods and materials and their appearances in art history. Dalí only pretends to uncover truths about the Angelus, because actually the entire book is about his own hallucinations. Instead of setting out from a hunch about a picture and shoring it with evidence until it becomes a reliable explanation, Dalí begins with a hunch suggested by a hallucination, and then sets out to explain his hallucination, using the painting as evidence. He uses the painting to explain his own delusions. He is fully aware of the distance between the actual image and his fantasies: “it was precisely a question of the contrast,” he remarks at one point, “between the delirious image and the unalterable aspect of the known image… [i]n the case of the Angelus, the delirious productivity is not of a visual order but is very simply psychic.”

The book’s first three chapters chronicle his successive hallucinations: an “initial delirious phenomenon” in June 1932 when the painting suddenly appeared to him, leading to the “first secondary delirious phenomenon” and the “second secondary delirious phenomenon,” itself comprised of six further “phenomena.” The rest of the book purports to explain those phenomena, but it does so by circling back to them, re-imagining them, and introducing new hallucinations. The structure is labyrinthine. An excursus describes a series of subsidiary hallucinations, including a kind of Ur–image, an “instantaneous image associated with the Angelus,” which gave rise to a series of further images including the line of Lenin heads on a piano, a set of fried eggs, a line of inkwells on a loaf of bread, and several sculptures, including the one that Dalí had himself photographed while wearing. 

Even now, he says, the image “still continues to cause me anguish”: it seems that if any image is primordial in his imagination, this one is. But the excursus closes by saying that the analogy to the Angelus was clarified after yet another delirious phenomenon, the unexpected apparition of a set of china decorated in Angelus motifs—and that apparition is listed as number five out of six aspects of the “second secondary delirious phenomenon”: in other words, the structure closes on itself, bites its own tail.

Logically, if not chronologically or narratively, the opening moment of the book is Dalí’s sense of the “inexplicable uneasiness of the two solitary figures” in the Angelus. But after the initial delirious phenomenon, “the admiration and sudden attractiveness” he feels for the painting “was to occur again, and in what followed was corroborated by the ‘fervent’ skepticism with which my friends opposed my abrupt admiration for the Angelus.” The feeling must be true because everyone around me doesn’t believe it: a sign of paranoid schizophrenia, as Dalí knew perfectly well. 

His description is designed to be unreliable: it is a hallucination, first of all, and it is reported in a bombastic style with deliberately suspicious details. At the same time, he observes his hallucinations very carefully (that is, before he records them) and he plays off their “delirious” quality. In the first secondary delirious phenomenon, Dalí is lying on a beach, playing a game in which he arranges tiny pebbles and imagines them as colossal stones (menhirs and mountains). It’s also a sexual game, and he tries out “various locations, pairings and ‘situations’,” including “the evocative poses of lovemaking.” Suddenly two stones remind him of the Angelus and provoke “the strongest emotion.” Later, reflecting on the moment, he recalls how the stones formed not only a normal Angelus, but an exaggerated one, and “the sentiment of this exaggeration nevertheless helped make me conscious of the clearly delirious character of the association of ideas which were in question.” Hallucinations, in this book, are the impossible preconditions of anything that can be presented as history, research, or truth. It's a hyperbolic position, but what in the recent literature on history, the authorial voice, and subjectivity, from Derrida, Barthes, Foucault and Hayden White onward, prohibits such hyperbole?

What kind of text is this?

Of the many reasons not to read Dalí's book in the context of this project, the most compelling is that he inverts the purpose of art history so completely that he avoids the many issues of writing and history that preoccupy most of the other authors I am considering here. And yet The Tragic Myth of Millet’s Angelus is in full possession of the apparatus of early 20th century art history (the scientific “evidence,” the theories), and the issues and methods he so happily inverts are the same ones that continue to help determine what is legible as history, and what as writing. Reading this book only as a document of surrealism would overestimate the stability of art history and even the innovations of postwar experimental art writing.

Chapter 17

 

Jean Louis Schefer, The Deluge

I begin with Schefer not because The Deluge, The Plague is the most influential text on art of its period but because it is still one of the strongest, which is to say strangest, products of the moment in the early 1970s when French experimental writing on art was moving forward under the influence of philosophers like Derrida, Foucault, Barthes, Lacan, and Lyotard. (The next best choice would perhaps have been Hubert Damisch's Theory of /Cloud/, but in its English translation that book has been domesticated in art history in a way Schefer's has not.)

The book needs an historiographic introduction, because it combines at least five cultural moments: 1969, the date of Schefer's first book, which was reviewed by Roland Barthes; 1976, the date of The Deluge; 1995, the date of the English translation; and 2014-2024, the years I drafted and revised this. Each has its own theoretical and historical concerns, and a full account should probably be an historiographic review.

Here I would like to concentrate on the writing, as it appears—as it is legible—in the current moment. Let me propose five ways of reading what happens in The Deluge.

Against semiotics
In a brief, evocative Foreword to the English translation of Schefer’s double book on Uccello, Stephen Bann notes that Roland Barthes understood Schefer’s first book as a "decisive new tendency in the development of semiology," away from the idea of pictures as codes, and toward the notion that a picture is "a work of coding, not the depository of a system but a way of generating systems." Barthes's take comes from his own experiments with semiotics, including his early systematizing efforts, and his awareness of their limitations. He was open, even then, to a reverse movement of semiotics, one that developed almost twenty years later into books like Camera Lucida.

The idea of a unique text
Second, Bann also follows Barthes in proposing Schefer's work is a hapax or hapax legomenon, an expression used by classicists to name a word whose meaning can never be known because the language or linguistic corpus—Sumerian cuneiform, for example—records only a single instance of the word, which can therefore never be compared with examples in other contexts in order to determine meaning. In Saussure's terms, there are no oppositional grammatical structures that can lend it meaning, so it remains unique and opaque. The Deluge is a "a text without redundancy," a unique instance. Schefer supports this reading in various ways, for example by saying bodies in the painting work as anacoluthons (sentence fragments or grammatically defective sentences, as in some modernist fiction), or by proposing that iconography can't be appropriate because figuration in the painting "does not add up" (pp. 27, 30). At the very end of his book he provided Barthes with his theme by saying a painting is a hapax, "a text without redundancy" (p. 54).

Charles Sanders Peirce
Bann has also compared Schefer's project with one of Peirce's semeiotic categories. He mentioned this to me sometime around 2007, so I may be misremembering, but I think the category he proposed was the sinsign, which stands in Peirce's system for "an actual singular thing, an actual occurrence or fact." The more precise denomination would be "rhematic iconic sinsign," but the lexicon doesn't matter as much as the idea, which is that Schefer's text would be a unique object, responding, in an untranslatable fashion, to another unique object.

The Body
The context of the English translation is a book series published by the University of Michigan Press in the mid 1990s, called The Body, In Theory. The editorial board included Bann, Norman Bryson, Alphonso Lingis, Lyotard, Elaine Scarry, Susan Stewart, and Schefer himself. In that context, The Deluge belongs with the turn to theorization of the body in North American art history and theory, a turn that includes Scarry's book The Body in Pain and the Zone Press Fragments for a History of the Human Body (1989), books by Rosalind Krauss and Hal Foster, and, a bit late, my own Pictures of the Body (1999). Schefer himself speaks of the body in the opening sections of his book: paintings are like bodies, he says, because figuration "posits the enigma of an inhumation of the intelligible body; of something that must, but cannot, be understood as such." (p. 26) The body, in this book, is not only the "disaggregated" bodies of the painting, but the sense of figuration itself: an idea that surfaces also in Lyotard and Lacan, both of whom are alluded to in the opening two pages of the book.

Writing
From my point of view, it matters most that Schefer himself frames his entire enterprise as a project of writing. Paul Smith, the English translator, interviewed Schefer; the text is in Artforum (1997) and also online. Schefer says Barthes's work is "marked by a way of perceiving problems which is ultimately just a style of writing." Under Barthes's influence, Schefer says, he "rather quickly opted for a quite personal way of working… I soon understood (or decided) that my 'literature' would consist in roaming among historical or cultural objects—that I'd be a writer without a novel." In The Deluge, at the end of his introductory section, he says "to write about painting is therefore, first of all, to write" (p. 32). Part of what this means is exactly what I imagine Barthes thinking of Camera Lucida: "writing," Schefer concludes, "is never either entirely protected from nor assured of not disappearing" into the "void of meaning" that the painting proposes (p. 33). The same distance from "a novel" can be read in Barthes's Preparation of the Novel. This expanded sense of writing is the best fit, and I think the broadest context, for this project, so perhaps the time has come to revisit the idea that seemed most foundational to Barthes back in 1969.

It would be possible to multiply this list: a sixth key to reading could be the religious themes Tom Conley explores in his Introduction to the English edition. Lacan is another possibility: he is alluded to very obscurely in the second paragraph of the book. But these five are enough to articulate the issues of writing I want to explore.

Experimental
In what ways is this book experimental, given its starting points in philosophy? By this I mean that it can seem to be enough to say that Schefer thinks about Lacan, Lyotard, Barthes, Bataille, and a dozen others at a certain point in the 1970s in France—but if that context is taken as given, then in what senses is Schefer's text moving into places it does not recognize?

One, I think, is simply the movement away from art history. At a distance of nearly forty years, that move appears at once uninteresting (because it has been done many different ways since then, including in books by Schefer himself) and very problematic (because the general corpus of French poststructuralist writing on art remains awkwardly detached from much of disciplinary art history, theory, and criticism). When Schefer writes that "it might be wise to note the importance of Uccello's place in the history of painting," and then immediately adds that "history has to be written in respect to the text that we are reading and only at the point where we are reading it," he makes a disavowal of historical sense more stringent than Keith Moxey's doubts in Visual Time (2013). But when he goes on to say that history "is disengaged, torn away, adrift," he is back in the more comforting domain of interpretation: the painting itself shows this fragmented sense of history, and so it can speak of it to a writer and reader who still imagine themselves to be in history (p. 39).

Several of Schefer's recurrent themes, including color, texture, figure, and perspective, have more and less experimental moments, as he tries to compel his writing not to repeat familiar forms of interpretation, from formal analysis and semiotics to iconography. "Figures of writing," for example, may be "floating, drowning, veneering," or "sinking," but they are also classifiable. Even clouds—the subject of Hubert Damisch's 1972 book, which Schefer would have known—can be classified ("clouds of color, heimatikè néphélè [Aristotle], a fog of color," p. 46).

This movement from radical, nearly illegible rewriting of meaning, and back to observations made by a particular viewer at a specifiable time, repeats itself in many ways, and I think its oscillation provides the book's most interesting structure. The pendulum swings from obdurate hapax to plausible interpretation all repeat the fundamental swing from art history to writing and back, and often the writer's voice is at its most concerted when the pendulum is about to swing back.

Publics

The mazzocchio, the strange checkerboard torus that two figures wear (and other emerges holds onto, in its guise as a barrel) is a hapax in the history of perspective. I wrote on it in my own first book, on linear perspective, because it is known that Uccello made a special study of it. It was a frame for a hat, but from a painter's point of view it was a supremely difficult exercise in perspective. Studies of the mazzocchio itself were a stock in trade of art historians from the late 19th century to the first decades of the 20th. Schefer doesn't bother to explain that, which means his book was aimed in part at art historians—although it is difficult to imagine how he thought it might be received by scholars for whom the mazzocchio was exclusively a vehicle to demonstrate the artist's or the scholar's perspectival skill. 

Schefer works especially hard at re-imagining the mazzocchio: it is part of the repertoire of the demonstration of the failure of realism; it is a sign of the ruin of the system of figural painting; at the same time "it is the only thing that might be apprised in the painting" (pp. 48-9, 92). It is, therefore, close to a key to Schefer's principal theme, the uniqueness of the encounter of writing and painting and the ways they ruin one another. But, as with Krauss's version of Lyotard's "pulse," Schefer does not want to let the mazzocchio work in such a way as to produce enlightenment: he doesn't want to let it produce itself as a theme, and so he reworks his account, multiplying, erasing, effacing what he has to say. In my reading, the pages on the mazzocchio are the most determinedly experimental.

Is there a place between art history and the hapax?
Is there, then, a nameable position between art history (by which I mean, in this case, any writing that can be read as a reference to history, to objects made in and for some past) and experimental writing (which denotes, in this case, writing that wants to insist on the imminent mutual destruction of meaning in the encounter of writing and painting)? I don't think so.

In a crucial section toward the end of The Deluge, Schefer says that "what makes up some of the painting" is the fact that "only its complexity can be understood… and not its meaning." Meaning, which seems to play in the representation, is "not just an instant summation of the discourse behind the figures, but a summoning of language placed in a scene of impossibly specific writing" (p. 55).

From this position, it is possible to name the objects of attention: figure, color, fresco, mazzocchio, language, writing. But it is not possible to keep them in sight for very long, because the flood of opaque water, the color of urine and feces, keeps covering them (p. 36). This is where experimental writing becomes poetry whose only remaining theme is its own necessary senselessness.

Chapter 18

 

Hélène Cixous, “Bathsheba or the Interior Bible” from Stigmata

What is the best way to describe the form, the structure, of “Bathsheba or the Interior Bible”? In asking this I am aiming at another question, which is difficult to frame well without qualifying terms like structure, form, the literary, and the historical. The question I am aiming at is something like this: What is it about writing, in this essay, that is not noticed? In what sense does Cixous not notice writing?

So considering structure first: there are, initially, two sources of structure in “Bathsheba or the Interior Bible.”

Structure
First there are the 24 numbered sections, which Cixous describes as “steps in the direction of Bathsheba.” These steps have an irregular structure, in that they deviate several times from their principal subject. Steps 4, 5, and 6 are on other painters, by way of saying what the Bathsheba is not.  Steps 17, 18, and 19 take up the same theme, offering more comparisons. After Step 20, the essay is about Rembrandt's Slaughtered Ox, as a way of thinking about Bathsheba. So, following the essay's own image, this "staircase" that leads to “meditation” (section 20) is not straight. Not that it needs to be: but a reader who notes these deviations will also be thinking of themes of approach, direction, and avoidance. The idea of walking toward a painting, or walking down a staircase in order to think about a painting, is compatible with semiotic critiques by Barthes, Kristeva, and others in the 1970s who sometimes imagined visual art as a presence somehow "beyond" language, at the end of some capacity of language—something you can approach, but perhaps never actually reach.

Second, there is an ongoing ekphrastic narrative, in which Cixous takes note of things she notices. It would be possible to extract those passages from the essay and read them on their own; I think they would make a reasonably continuous narrative. She notices nudity (section 1), the dark background (section 2), Bathsheba's face (section 4), the direction of the light (section 7), the paint (section 9), the servant (section 12), the letter (section 13). There is a rhythm to these observations, which runs at a different speed than the rhythm of the numbered sections. It's as if observations are things that happen to meditations, rather than the other way around: her imagination is not driven by what she's noticed, but things she notices appear to her in writing when they coincide with her imagination. That relation between observed detail and themes is less empirical, in its way, than the essay's first source of structure (its numbered sections), because at least those go "in the direction" of the painting.

The anti-representational
In terms of logical claims, what matters in this essay is an argument against the sufficiency of naturalism. Rembrandt is praised for setting aside naturalism in favor of forms of interiority. This is the historical, logical content of the essay, and it has an interesting relation to what might be called the essay's literary interests, which are articulated around the metaphors of stigmata, pain, and death. Let me expand a little on the essay's logical claims in order to draw out the parallel with the literary concerns.

On the one hand, the claim that Rembrandt paints interiority is a standard part of his reception, one that Cixous will have known, even if she didn't get it from reading what she calls "the big catalogues" (section 24). This traditional value is clear for example when she says

That which wells up in Bathsheba, that which the letter has poured into her body, into her organs, into her brain, and which is working on her body, her face, her brow, from the inside. (section 19)

This is an idiosyncratic but fairly standard way of putting the claim about what is called psychological portraiture, which is a tradition in Western painting, a standard element of academic instruction in the Baroque (physiognomics), and a part of the Rembrandt reception since his lifetime. The soul paints emotions and thoughts on the face, and Rembrandt's skill was to have shown them there. A number of books have explored this thematic, in this case perhaps the most salient being Svetlana Alpers's Rembrandt's Enterprise: The Studio and the Market. Mieke Bal's enormous book Reading "Rembrandt" is partly about deconstructing this notion. When she expresses her disappointment that Vermeer painted "cupboards" and Holbein cared about Erasmus's appearance, she is repeating conventional terms of the appreciation of Rembrandt's psychological portraiture.

On the other hand, Cixous is also developing an anti-representational critique, which involves the claim that what is most interesting about painting is what it does not show. "Rembrandt paints the secret," she writes, "the trace of what escapes us," things "beyond the painting, beyond thought" (section 18). Some of those things are commonly noted in relation to post-Renaissance painting, such as painting "what has just happened" and "what is going to happen," but others, such as the idea of painting something "beyond thought," belong more to the late 20th century poststructuralist critique of representation. In the next essay in Stigmata, which is also the only other one about visual objects, she asks “what do we want to draw?” and answers, “it's not a question of drawing the contours, but of what escapes the contour, the secret movement, the breaking, the torment, the unexpected. The drawing wants to draw what is invisible to the naked eye” (pp. 19-20). That essay is a confessional, performative, phenomenological account of drawing, and this essay on Rembrandt brings the same interests to bear.

(It should be asked what in Cixous’s accounts is specific to painting or drawing, and what is a metaphorical or rhetorical projection of her interests in writing. For me, the few passages in the second essay that refer directly to drawing depend on their palimpsestic passages. But most drawings do not “combat themselves” in that way, and Cixous’s description of those passages is neither specific to the drawings, nor something that couldn’t be said about writing. But all that that is not my subject here.)

These are two elements of what I would call the essay's logical claims, its interest in the working of naturalism. There is an interesting parallel between those claims and the essay's literary interests in metaphors of stigmata, pain, and death. The preface to the book draws a distinction between scars (which are a residue) and stigmata (which are emptinesses), and readers will be thinking of that when they follow Cixous "in the direction of Bathsheba," that is, into the painting. That motion toward the inside is the rhetorical analogue, the figural companion, of the essay's arguments against naturalism. The question, from the perspective of writing, is how those two are connected—or if Cixous feels they need to be.

Dramatic scenes
There are some passages in which the movement of Cixous's thoughts toward interiority (Bathsheba's, the painting's, Rembrandt's, her own) produces dramatic scenes. Here is one, which opens, I note, with another staircase image:

Taking the red staircase, down to the bottom of ourselves, under the earth's crust,
This world is full of night and of golden stuff. The stuff of night is a clay. A mud. It is still moving, imperceptibly.
No landscape and no 'furniture' either. Instead of furniture, 'shelving,' 'shelves' of color. Bands, brush strokes. What do you see, there?
A man said to me, here's a cupboard. With linens piled up.
Another man, this one a painter: an architectural background, a pilaster.
A commentator said to me: the curtain has been drawn aside.
We have such a need of cupboards, of curtains, such a need to furnish.
The interior world is full of night and of golden stuff, of the stuff of night. Spools of night. (section 8)

On the surface, she is reporting what friends said when she asked them where Bathsheba and her servant are meant to be. Her friends tell her that all sorts of ordinary things are there, including the machinery of Baroque painting, pilasters and curtains. She wants to say that they refuse to notice the "clay," the "stuff," part of which turns out, in the next section, to be the "mud" with which Rembrandt painted. The background, for Cixous, is "night," interiority, David, skin, and blood: "The entire room is flesh. Sex." (section 8)

In terms of writing: this scene is a drama within the text, like a play within a play, with its own dramatis personae ("a man," "another man," "a commentator"). The kind of interiority it conjures has echoes of Beckett: it's threadbare, perhaps with a cupboard or a curtain. On the other hand the passage also has the deep golden and red tones of Gustave Moreau's interiors, and in that sense its literary precedent is also Huysmans or Lautréamont. It's a drama of blood, darkness, and sex: a very fin-de-siècle tableau.

What do these dramas do for the narrative? Are they optional flourishes or moments of pleasure in relation to the surrounding prose, or are they crucial enactments of ideas that cannot be simply listed? I don't see any places in the text that open that question, and I don't perceive it as a question Cixous is asking herself. For that reason it appears to me as a lost opportunity to make the text, and the response to the painting, more compelling. The undecided relation between the text's literary interests and its arguments is a similar case. In both, I wish that Cixous's sense of her project would have included meditation, exploration, figuration, or—in a metaphor better suited to her own book—suturing of historical argument and literary tropes, or descriptive narrative and drama.

This is the sense in which I would say Cixous does not notice writing in this essay. It is as if she is supernaturally aware of herself as a writer, playing with languages, etymologies, and narratives, but at the same time she takes her project, as a writer (and Derrida's encomium is prominently displayed: she is, for him, the best French writer of the 20th century), to be the enactment of the moment and possibilities of writing, rather than the analysis of the the kind of text that her writing has gone ahead and produced. For me, that is ultimately a reason I let myself stop reading: the text is not as reflective as it might be, despite—or, exactly, because—of the kinds of vigilant self-awareness she directs at writing in general. The text is doesn't perceive barriers or boundaries: that gives it a freedom, but it also issues a license, to readers, to stop reading carefully, to become less fully engaged in what works and what doesn't, in what's said and what isn't, and ultimately, in what is written.

Images
The images are set with absolute indifference into the text, at least in the English edition: I haven't yet had time to compare the French. The same is true of the second essay in Stigmata, whose subject is drawing: it reproduces two images, and there is minimal interaction between images and text.

Chapter 19

 


Gilles Deleuze, Logic of Sensation

Let me propose there are three levels of reading that make this book pertinent, and even exemplary, for the French poststructural response to writing about visual art.

1. The original French edition

The first of the two volumes of the original 1981 edition of Deleuze's book (published by Editions de la Difference) is text, and the second is illustrations (note the Roman numeral I on the cover, above; vol. 2 is titled "II - Peintures").

Tom Conley's Afterword to the English translation is exemplary in its attention to this fact, but even Conley, who is arguably the scholar most likely to take formatting and illustrations seriously, doesn't draw many conclusions from the layout. He notes that call-outs (references to the images in the second volume) are placed in the margins, "somewhat like title-summaries in manchettes in early-modern printed books, in which the text itself can be seen at once as a 'legend' underwriting the images or even as a component unit of a greater 'fable' built upon the composite character of words and pictures" (p. 131).

Here is a page from an early modern Bible, for example, with cross-references in the margins. (There are many more recent examples, including the original French edition of Barthes's Camera Lucida.)

Part of Conley's gloss on the "manchettes" (marginal call-outs*) is plausible: the use of marginal numbers does create the effect that the text is a "legend" (caption), but for me the practice means more that the text, considered as a whole, and the images, in their separate book, are equally important, that both are continuous, and that one is not interrupted by the other. The call-outs also remind a reader that the text does not often need to specify exact images, and in fact Deleuze doesn't always say exactly what image he means: a figure number is anchored well enough if it is in the vicinity of its sentence. In the English edition, the call-outs only refer to the small-print list of plates at the end of the book, and not to illustrations; but they are in the text, in square brackets, which places them precisely in relation to the grammar of the text. In that way the logic of a given sentence, and its singular referent, are closely bound. In Deleuze's usage, the vagueness about that relation is striking. Why, a reader may ask, does it not matter exactly when images are being referred to, or exactly how many images might be meant, or when a reader might choose to look at the images?

Conley notes that once an illustration has been called out, its number may not be given again in a subsequent passage, implying either that the reader has gotten to know the image in question, or that readers aren't expected to turn back and forth as they would in a conventional art history text. Conley suggests the two volumes be read "in juxtaposition," perhaps in the same disordered way that the chapters of the text can be read. (Deleuze says his chapters are arranged in order of ascending difficulty, but that statement immediately, and permanently, places in question the value and meaning of "complexity," inviting readers to read in other orders.)

Later in the introduction, Conley guesses that Deleuze might have owned eleven paintings that were added, without explanation, to the third edition of the French text; at least Deleuze probably had reproductions on his walls or floor when he was working on the book, because "the unlinked and paratactic quality" of his observations suggest he is "telling the reader to break frequently with the line of his reasoning by looking in detail at an ample quantity of pictures" (p. 132). Conley also takes note of the fact that the back covers of both volumes of the original edition have photos Bacon took of himself, which makes the covers look like contact sheets, and brings the artist's body and life back into the reading—but Conley's reading stops with those observations, and he moves on to other topics (p. 142).

The fundamental physical fact of the two volumes means that Deleuze's text exists alongside the paintings as a proximate but potentially detachable narrative. That property is made literal in the single-volume English translation, which is entirely unillustrated and has only a list of the paintings, in a remarkably tiny font (especially minuscule in the paperback--as if the editors felt a fiduciary responsibility to list the paintings they weren't reproducing.).

It is also pertinent to the phenomenology of reading that in the original French edition, some triptychs are foldouts, producing a suddenly more immersive experience. (My scanner couldn't accommodate the full width, which is approximately 24 inches.)

My experience reading in the original is that I seldom have image and text side by side, because it's too awkward. I turn from one to the other, looking or reading sequentially in either volume, then returning to the other for "ample" reading or viewing. To see images and text in strict parallel, as in a more conventional book, it is necessary to evaporate the physical books into digital images. (I reproduce this effect below: it is close enough to the experience I imagine many people have while they study the English translation, with a screen nearby to check references. Needless to say that sort of reading won't be what Deleuze imagined.)

I am not aware of any documents or further information about Deleuze's involvement in the design of the book, but as it was printed, the Editions de la Difference text is a material exemplification of a theme that Deleuze develops throughout the book: the possibility of writing in such a way that the images are not reduced to illustrations, decorations, examples, or mnemonics as art history typically does. (These three terms are explored as part of a critique of art history and visual studies in the Introduction to Theorizing Visual Studies, and also online in a project called Writing with Images.)


2. The logic of writing in Logic of Sensation

Two things become clear, I think, early on in a reading of either the French or English versions of Logic of Sensation: first, the text has an unusual form (chapters which reproduce an ascending complexity, but which may be read in any order); and second, that the unusual form is somehow motivated by the subject matter. Or to put it more exactly: the unusual form of the book may be a response to the paintings, to the "logic of sensation" that the text is exploring, or to both. It also becomes evident that the author will remain silent on these points, and that he is possibly working on understanding them as he writes.

The ordering of the chapters announces its open-endedness, its randomness, at the same time as its author asserts the chapters' logical order from simple to complex. This seems consistent with Deleuze's lack of interest in Bacon's development, except where it serves his themes. His text is a conceptual analysis rather than a chronology or history—and in that regard it does not require the images to be arranged in any particular order. Conley notes that this open-ended and yet structured presentation is consonant with Deleuze's interest in open-ended structures of argument, totalities such as "a thousand plateaux" that "cannot be accorded a finite measure" (p. 134).

The book all but proposes that its structure, its form, is analogous to the "logic" of its subject. As Conley puts it, "concepts move through and across his oeuvre analogously to the way painterly forms migrate to and from many places in Bacon's paintings" (p. 142). Some chapters are "thumbnail summaries of a theory of aesthetics," and others are fragments, or portions of larger arguments (p. 148). Conley thinks The Logic of Sensation gave Deleuze a logic of composition that he took with him to his later projects:

Before the Logic of Sensation Deleuze philosophizes and conceptualizes; after the work on Bacon a greater and more supple sense of flow, flexion, transformation, and bodily force becomes evident. The style becomes the very image of what Deleuze draws from the life he lived with the paintings. [p. 149]

These are all structural parallels between the "logic" Deleuze finds in the paintings and the text he produces. There is a strong parallel between Deleuze's central theory of sensation, which exists in levels and strengths, and comes at us with immediacy and without systematic mediation, and the chapters in his text, whose fundamentally disordered order and varying strength and concision mirror their content: the question is how to read his decision to represent the "logic" of sensation both in the structure of the text (its chapters, its "supple... flexion") as well as in the text's propositional logic (its argument).

The idea of writing in a series of differentially disconnected chapters has to appear as a parallel to the book's subject matter, which is a lifetime of differentially disconnected canvases. The idea of writing about the logic of sensation in a series of differentially disconnected chapters also seems appropriate, even if its logic is harder to deduce. In general, why write using the form of the object that is being explored? Since Deleuze isn't writing under the pressure of radical claims about the relationship between written form and content, such as the ideas in Adorno's "Essay as Form," it is not clear why his writing persistently explores the possibilities of presenting itself in levels, intensities, and encounters, even as it describes those very terms.


3. Writing against figuration and abstraction

The Logic of Sensation can be read as a model of how not to write philosophy at images, or imply images are philosophy, or that they're adequately imagined as philosophy, history, or criticism: Deleuze's text refuses to be a commentary, just as it refuses physical control of the images of the sort that is implied by conventional art historical or theoretical texts that incorporate reproductions into the flow of the printed page. The writing exists alongside the paintings, both because it is physically adjacent to the companion volume, and because it thinks by enacting parallel structures of force and meaning. Sensation is immediate, it is "translated directly" (Deleuze is paraphrasing Valéry; la sensation, c'est qui se transmet directement): unlike abstraction and figuration, it does not "pass through the brain" (p. 32, p. 28 in the original).

There is a problem, I think, in taking this literally. If the text was actually embodying or exemplifying sensation, it would cease to argue altogether. Yet I am continuously tempted to make a parallel between the theme of avoiding both figuration and abstraction, and Deleuze's own writing as an attempt to avoid both history and philosophy. A useful vehicle for this parallel is Deleuze's notion of the Figure.

The liberation of the Figure from figuration enacts the liberation of writing from description, history, theory, and criticism. The "very general thread" (le fil très général, an odd metaphor) that links Bacon to Cézanne, Deleuze says, is "paint the sensation," in italics in the original (p. 32; p. 28 in the French). It would not be misplaced, I think, to read this phrase as write the sensation. As Deleuze says of Bacon's supposedly grisly figures, "the Figures seem to be monsters only from the viewpoint of lingering figuration" (p. 123): a statement that could be made just as well about his own book.

There is a brief passage on Proust (about whom Deleuze had written a book), in which Deleuze agrees with John Russell's observation that Proust's theory of involuntary memory is similar to Bacon's practice. Deleuze comments:

This is perhaps because Bacon, when he refuses the double way of a figurative painting and an abstract painting, is put in a position analogous to Proust in literature. Proust did not want an abstract literature that was too voluntary ([that would be] philosophy), any more than he wanted a figurative, illustrative, or narrative literature that merely told a story. What he was striving for, what he wanted to bring to light, was a kind of Figure, torn away from figuration and stripped of every figurative function [arrachée à la figuration, dépouillée de toute fonction figurative]: a Figure-in-itself, for example the Figure-in-itself of Combray. (chapter 9, p. 56)

In Deleuze's critique, "there are two ways of going beyond figuration (that is, beyond both the illustrative and the figurative): either toward abstract form or toward the Figure" (p. 31). The Figure is a direct record of sensation, the object of Deleuze's study, and its "logic" involves such things as color, the frame, the contour, and other elements that are the subjects of The Logic of Sensation.

It wouldn't be inappropriate to read this passage, and Deleuze's subject in general, as an allegory of his own sense of what it is to write philosophy to one side of painting or literature, rather than for or as painting or literature. As Conley says, Deleuze's style "becomes the very image" of his experience looking at Bacon's paintings: it is meant to stand along with his experience of the art, just as the first (unillustrated) volume of the French original edition stands alongside the second volume of plates.

This sense of the Figure in the text, both in Proust and in Bacon, can be understood as a story Deleuze told himself about the kind of writing he meant to accomplish. He was experimenting with writing the Figure in the text of literature: neither "too voluntary" nor merely "illustrative," but a form of escape from both that was indebted to and dependent on their continuing presence. As such it is an especially strong model for how to write about images: such writing would be a deep challenge to academic modes of addressing images, up to and including poststructural theories by Tom Mitchell, Jean-Luc Nancy, and others, and it would have the interesting virtue of being not entirely easy to justify, maintain, or even understand. 

— Revised April 2024

_________

* From the CNRTL website, s.v. Manchette: "IMPR. Addition imprimée dans la marge d'un texte. Synon. marginale (vx).Ouvrage à manchettes. Il n'y a pas de manchettes dans ce livre (Ac.1935). Les notes marginales, également appelées additions marginales ou manchettes sont autant de sous-titres secondaires placés en marge et vis-à-vis du texte dont elles énoncent l'esprit (E. Leclerc, Nouv. manuel typogr.,1932, p. 229)."

Chapter 23

[Note to readers: if you find this page without its context, please have a look at What is Interesting Writing in Art History?, which will make sense of the context. Much of this particular essay is about Don DeLillo's Point Omega. I have written a brief essay on the philosophy in Point Omega here.]

Artworks in Fiction

—Art history: “the most evil business there is” 

(Thomas Bernhard, Old Masters)

At this point I come across a boundary, and a kind of uninhabited zone between known territories. Behind me is the literature I have been calling experimental writing on art, and even farther behind, there is art history. Ahead, the writing includes images but is not necessarily about art.

What does the land ahead look like? In the summer 2013 issue of Frieze, the novelist Ben Lerner speculated on the relation between novels and visual art. He mentions Thomas Bernhard's Old Masters, Javier Marías's A Heart so White, Proust's passage on the yellow wall in Vermeer's View of Delft, John Ashbery's Self-Portrait in a Convex Mirror, and Tim Clark's Sight of Death, which he says "has as much in common with novels as it does with most conventional art criticism." (That is an unsatisfying thing to say without qualification, because it could mean Clark's book is like a novel, or that it's unlike either novels or criticism.) All of Lerner's examples, except Clark, are fiction or poetry that address actual works of visual art. His own book 10:04, which appeared in 2014, briefly mentions several actual artworks. I'll discuss several of the books Lerner mentions, including his own, in the next part of this project.

Lerner then concludes that even though "Clark is rigorously involved with two real canvases,"

the novel is a space wherein such an experiment in art writing can take place before the existence of the art itself, where an encounter can be staged between individuals and/or art works that are not or cannot be made actual. Fiction that describes encounters with artificial objects that don't yet or can't yet exist is usually called "science fiction"… but we could also organize a genre of "speculative fiction" around virtual arts; Keatsian music; a painting that never dries. (p. 155)

This conflates two possibilities that I think it's important to separate.

On the one hand, fiction can address works of art that don't exist, and on the other, fiction can address works that do exist, or have existed, but it can propose "staged" encounters and narratives around those works. Let me number these in order to keep track of some ramifying possibilities.

(1) There are relatively few works in the first category; one is Balzac's "Unknown Masterpiece," about historical artists and an invented artwork. (Thanks to Joachim Sprung for reminding me of this.)

(2) The second category is what happens in every one of Lerner's examples (again excluding Clark, because that book doesn't present itself as fiction). It is also what happens, Lerner says, in his own forthcoming book 10:04, which mentions Judd sculptures and Jules Bastien-Lepage's Joan of Arc. (It's a strange article, because in the Frieze essay Lerner is advocating a possibility he hasn't practiced.)

Within the second category, fiction that stages encounters about actual artworks, there is an apparently small but important distinction to be made between (2a) writing that can be said to address a work of art, like Ashbery's poem, and (2b) fiction that happens to include art among many other interests.

In category (2a), a central part of the writer's purpose is to say something about specific artworks, and that brings it, distantly but distinctly, into the domain of this project. This category includes any number of ekphrases from ancient examples to the present. Writing poems about artworks has been a tradition in European writing since the Renaissance. (See Rensellaer Lee's Ut pictura poesis.) Ashbery, Auden, and many poets have written novels, stories, and poems that address artworks. The critic Raphael Rubenstein curated poems on contemporary artworks in the summer 2013 Art in America, and he has written some himself. Several important works of modernist fiction treat actual artworks extensively, such as Peter Weiss's Aesthetics of Resistance, and there are also contemporary and popular fictions that turn on real artworks, from Donna Tartt's The Goldfinch to Dan Brown's books.

On the other hand, Proust's novel is certainly in category (2b), because it's only in the world of art history that his reference to the "yellow wall" in Vermeer's View of Delft looms so large. When a long novel like Proust's stops briefly to evoke a painting, then the novel isn't about the painting: the painting is a vanishingly small point, a moment of the Real, in a larger fiction. (A lovely book that inadvertently demonstrates this is Eric Karpeles's Paintings in Proust.)

At the risk of being overly analytic, I think it's a good idea to distinguish those two possibilities, (2a) and (2b), from another: (2c), books that are said to have been inspired by images—if not artworks—even though those images aren't mentioned. Don DeLillo has said that Falling Man was inspired by the well-known photo taken on 9/11, and it has been said that inspiration at that remove might be more promising than attempting to write ekphrases of individual images that are not reproduced. (See for example this New Yorker blog.)

But (2a) remains an interesting possibility. There are works of fiction in which some measurable part of the author's purpose is the evocation of artworks that readers know, or could go and see. It is tempting, in the context of this project, to see what happens if such texts are read as if they are "experimental writing on art": as if fiction was partly, or even principally, a strategy to say more about the art than would be possible in the constraints of nonfiction.

I'll consider two examples: the first, a novel by Don DeLillo, goes to show how hard it can be to say exactly what's gained by descriptions of artworks in fiction. The second, Proust's In Search of Lost Time, indicates the complexity of trying to unweave descriptions of artworks from their fictional meshes.

Artworks in fiction: Don DeLillo's Point Omega

DeLillo's novel begins and ends with descriptions of what it's like to see Douglas Gordon's 24 Hour Psycho, a version of Hitchcock's movie slowed down so it takes a full day. At that rate, individual frames (usually invisible at 24 frames per second) appear as softly blurred individual images. Several excerpts on Youtube demonstrate that. It would be perverse to read Point Omega as an attempt to say things about 24 Hour Psycho that cannot be accommodated by art history, theory, or criticism, but if I was to read it that way, I might be able to distinguish descriptions that might occur in nonfiction from other passages that could only take place in fiction.

(Incidentally: there are several screening configurations of Gordon's piece. In the original show, the one DeLillo saw, there was one screen:

Later Gordon showed a version with two screens, one running backwards, so that they met in the middle with a single synchronized image (not shown here):

Apparently Gagosian also showed 24-Hour Psycho along with other Gordon videos in the same space. None of this has to do with DeLillo's book, but I want to mention it in case some readers have seen other versions. In light of the book, each different installation could lead to a different fiction.)

So, to distinguish descriptions that could be done in nonfiction from descirptions that require fiction: for example, the book opens with a straightforward account of the screening at MoMA in 2006, told by a third-person narrator. (Most of the book uses a first-person narrator, but this character is different; he is effectively first-person, but is given in focalized third-person to distinguish him from the book's principal narrator.) DeLillo tells us about the screen, the guard, and the ways people move around the room. That sort of introductory material can be found in art criticism. After a few pages, the narrator's descriptions become more exact, and some of his observations wouldn't be out of place in a film theory essay. He notes, for example, that "Anthony Perkins turns his head in five incremental movements," and the shower rings vibrate up and down in the famous scene where Janet Leigh's character is murdered (pp. 5, 9).

But then the third-person narrator notices things he doesn't analyze, and he even notices that he's not understanding:

He stood and looked. In the time it took for Anthony Perkins to turn his head, there seemed to flow an array of ideas involving science and philosophy and nameless other things, or maybe he was seeing too much. [p. 5]

The narrator misidentifies two visitors to MoMA as film theorists, and he watches them,

the academics, adepts of film, of film theory, film syntax, film and myth, the dialectics of film, the metaphysics of film, as Janet Leigh began to undress for the blood-soaked shower to come. [p. 9]

These passages wouldn't happen in academic film theory, criticism, or history, if only because they report on theories that are not themselves explained: they're about an awareness of theories, but they don't serve to articulate those theories. (In other words, they aren't philosophy or argument.) This is a simple and clear example of a way to talk about 24 Hour Psycho that would not be available in nonfiction.

When the two visitors whom the narrator mistakes as as academics leave, he thinks:

They went past the guard and were gone. They had to think in words. This was their problem. The action moved too slowly to accommodate their vocabulary of film. He [the narrator] didn't know if this [idea] made the slightest sense. They could not feel the heartbeat of images projected at this speed. Their vocabulary of film, he thought, could not be adapted to curtain rods and curtain rings and eyelets. [p. 10]

This is even more obviously not plausibly a part of academic writing, because it is a critique of such writing. The ideas DeLillo's narrator gives voice to could be analyzed, and then they might well appear in an academic context (a little like this one!), but the passage itself is a criticism, in a voice that's incrementally close to the author's.

The opening and closing chapters of Point Omega are the only ones devoted to describing the experience of viewing 24 Hour Psycho. But the body of the novel is full of parallels to the film, and each can be understood, in the perverse reading I am entertaining here, as a way of saying something about Gordon's film that could not be said either in nonfiction or in the straightforward descriptions in the opening and closing chapters.

First, there are several films mentioned in Point Omega: some real, like Sokurov's Russian Ark, and others invented, like the first-person narrator's own film about Jerry Lewis (pp. 22, 25-7). There are also echoes of specific scenes in Hitchcock's film, for example a memory one character has of a janitor who walked down stairs backwards, which echoes the detective's fall down the stairs in Hitchcock's film (pp. 32, 37, and compare the escalator described on p. 41):

Or I drove into box canyons, over hard dry cracked earth, car swimming in heat, and I thought of my apartment, two small rooms, the rent, the bills, the unanswered calls, the wife no longer there, the separated wife, the crackhead janitor, theelderly woman who walked down the stairs backwards, slowly, eternally, four flights, backwards, and I never asked her why.

(Here's a still from that scene in Psycho.) 

Those parallels exhibit a freedom not available to writers who write only nonfiction, but it isn't obvious exactly what they contribute—and part of the reason is that I am constrained by my reading to search for passages that enrich our understanding of Gordon's film, rather than passages that contribute to the novel's themes of eternity and the "omega point." If you're interested in Point Omega, you might say that the passage about the elderly woman is made more sinister by the echo of Arbogast's slow-motion murder, that we're all going backwards and down through life. If you're interested in experimental writing on art, however—so that the passage contributes to an understanding of 24-Hour Psycho—then you might say DeLillo's sentence enriches the artwork by introducing a different kind of slowness, the careful steps of the elderly woman, and by suggesting that Detective Arbogast's tumble, when slowed down, is like a person carefully walking backward. Im not sure where that would get you, but Point Omega permits it, passively. You would also, I think, be omitting most of the sentence, and you'd be unable to account for what its other images might contribute.

More interesting are the extended parallels between 24 Hour Psycho and the dilated time the narrator, his host, and his host's daughter spend in a small house in the desert, the setting for the main part of the book. The narrator's host, Elster, has ideas about eternity and infinity that work as hyperbolic extensions of the twenty-four hours of Gordon's film. He contrasts time in the desert, and especially the suspended time that he experiences after his daughter vanishes, against the time of cities, "dimwit time, inferior time, people checking watches and other devices" (p. 45). Elster visited the installation of Gordon's film in MoMA, and it reminded him of even vaster time scales. "We're the last billionth of a second in the evolution of matter," he says. For him, the film was "like watching the universe die over a period of about seven billion years" (pp. 50, 47). Point Omega is about "deep time, epochal time," and Teilhard de Chardin's concept of the "omega point," the convergence of all creation onto unity, which Elster interprets in a secular sense as the time when time will slow to a stop and people will become material: stone, desert, sky (pp. 52, 72, 98).

Things get complex from there, because even though DeLillo's theme is clear (he has Elster say it several different times, and the book's penultimate chapter, when Elster's daughter has gone, plays out the timelessness of "point omega"), the novel's many partial parallels and slant rhymes with 24 Hour Psycho make it difficult to say how the meaning of that film might be enriched by the novel, except in the trivial sense that the film elicited or crystallized the author's thoughts about deep time. I'll just give one example of that fairly unencompassable complexity.

Elster's daughter tells the narrator how she and a friend spent a day walking around the Chelsea gallery district: first they visited all the galleries, but then they went back again, street by street, and deliberately stayed out of the galleries. "It was like the idea of their lifetimes," and it "deepened the experience" (pp. 67-8). DeLillo describes the women going up and down the streets in a way that is reminiscent of his description of the people wandering in the screening room in MoMA, setting up a comparative analogy, something like this:

seeing Chelsea galleries : skipping them : : Hitchcock's film : Gordon's film

It's a complex analogy, and I can't quite put into words how, or even if, it contributes to my appreciation of 24 Hour Psycho. So at this point it might be best to give up my speculative reading, but the question remains: fiction definitely does sometimes add to our understanding of artworks, but as far as I know the mechanisms by which that happens, and the special resources fiction may provide, remain untheorized.

Proust

Proust makes many sorts of references to visual experience, visual memories, and artworks real and invented. Because there are no images in the novel, it is outside the current project: but it remains a central instance of the representation of the visual in modernist fiction. The principal question here is what counts as the visual in Proust, both in relation to his theories of memory, and more literally in relation to his interest in a actual paintings and architecture. I think on the whole Proust is far less accessible than he has been taken to be by critics who imagine he describes artworks.

I will consider the place of the visual in the novel from two perspectives: first, the question of what sort of visual references tie the novel to actual artworks; and second, the curious problem regarding what, in Proust's mind, works as evocation of the visual.

The novel's sense of visual artworks

The intensely visual description of the Combray church has traditionally been regarded as a subjective recollection of the church in the actual town of Illiers (I. 80). But as Germaine Brée argued in 1967, it seems likely Proust was looking at a reproduction of Vermeer's View of Delft. His striking line about "des ecailles et des egouttements gommeux de soleil," "flakes and gum-like driblets of sun," seems especially painterly. ("Proust's Combray Church: Illiers or Vermeer?," Proceedings of the American Philosophical Society 112, 1968, pp. 5-7.) This possibility makes the memory that much more complex, especially if it is read alongside Benjamin Binstock's plausible suggestion that Proust was not looking carefully at the painting (or rather, a reproduction of it) when he came to actually describe it later in the novel. 

If something of both these observations are true, it becomes interesting to contemplate what counts as ekphrasis, or as a relation to facts in the world. Within the novel's logic, there are "inner" and "outer" kinds of experience while reading (I.115-20), partial and more complete memories (many, like the madeleine memory, not completely explained for hundreds of pages), voluntary and involuntary memories, and the continuous reminder, from the narrator to himself, that the real issues are in his own mind, not in the world. (This is repeated nearly as a mantra, perhaps five times in the first hundred pages.) The fictional town Combray may be fused with the reproduction of a painting, and the reproduction fused with other unspecified memories, so that the novel never reaches out of itself, and cannot be properly said to attach itself to visual objects, like Vermeer's painting, which exist in the world for others to see. There are many counterexamples--first among them Giotto's Arena Chapel, where realism is exactly what is at stake, even though it is pointedly not the realism of a naturalistic painter--but the ones that matter are entangled in the text, embedded (a Proustian metaphor) in the novel's inner life.

The place of the visual in narrative

Another aspect of the visual is what counts, in the novel, as a visual experience or description. It is crucial that the madeleine episode is prefaced by an inventory of the house in Combray (I. 58-9). The narrator's recall is incomplete ("as if all Combray had consisted of but two floors joined by a slender staircase") but precise; and as soon as he completes his inventory he unveils the theory of "voluntary memory, the memory of the intellect" and goes on to describe the madeleine. In that fashion intensive, purely visual memories are left behind, at least so we're told, in favor of involuntary memories sparked, in this case, by taste. He isn't consistent in this, however, and by I.66 he is recounting in great voluntary detail the smells of the house. It seems that voluntary memory is linked to visual memory, and even voluntary recall of smells, tastes, and sounds can function as if they were involuntary. A characteristic use of the visual is in I.252, where the narrator says he would stand "motionless, looking, breathing," trying to "penetrate with my mind beyond the thing seen or smelt." He sees three steeples, two in Martinville and in Vieuxvicq, and tries to perceive "what lay hidden beneath them." The visual world is therefore an incomplete prompt for more inward and genuine experiences. (This is from a painting by Corot of Martinville; Corot is mentioned in the book.)

Often, however, the hidden meaning isn't a childhood experience or other instance of "lost time," but the expression of the search itself in writing. The entire interest of the visual in the novel can even be imagined, without much distortion of the narrator's intent, as a problem in writing. In I.256 for example, Proust's narrator resolves the deeper meaning of the three steeples not by discovering something new, but by writing about them; and the passage he presents us, which he says was written on the spot, hardly offers anything different from the initial description that precedes it in the novel.

It's possible to put this even more strongly, or literally. Visual memories are of interest often, and perhaps ultimately, because they can be enchained, linked in clauses and sentences to thoughts that Proust's narrator calls "philosophic," and which often involve other senses. An example among many is I.238, where a sharply visual description of a lily, pushed back and forth by the current of the Vivonne river, reminds the narrator of "certain victims of neurasthenia" who exhibit what we would call repetition compulsion, and then of "those wretches" in the Inferno who repeat their torments indefinitely. The point on which this sentence rests—and the point of the sentence—is the unexpected turn, at the end, where the thought of Virgil hurrying Dante past the damned reminds the narrator that his parents want him to catch up to them on their walk. The purpose of visual description, here and in general in the book, is to facilitate, introduce, or initially ground and focus enchained ruminations that move away from the visual and on into other senses, metaphors, allusions, and finally into unexpected insights. The visual is reimagined as writing.

(I owe more than the usual passing thanks to Kayla Risko, who pointed me to most of the passages I mention here. There is more on Proust and ekphrasis in Visual Worlds, co-authored with Erna Fiorentini.)


In the absence of conclusions

I hope these two inadequate analyses suggest how much more there is to be said. I don't disbelieve in the perverse attempt to read a novel, like DeLillo's, to see what it might be saying about an artwork: but I can't see how to proceed with the analysis. And I don't think Proust's special kind of imbricated complexities of memory and invention are a model for other people's fictions about actual artworks: but I am not sure how Proust's novel differs from others fictions in that regard. It's a fascinating topic, one that is not well served by the simple observation that novels sometimes contribute to our sense of the complexity of actual artworks.

Revised April 2024