PRINT January 1995


ALTHOUGH I HAD MET GEORGES DIDI-HUBERMAN some years before and was already an admirer of The Invention of Hysteria, his book on the Salpêtrière photographs, I didn’t really get to know him or to have a sense of his extraordinary range until the fall of 1983, when we both participated in a photography symposium at the University of Liège. Lodged with a faculty family just off campus, the two of us found ourselves over the course of several days breakfasting and talking and laughing together in the wan sunlight of this fanatically neat house, before trudging off to the lectures, most of which betrayed assumptions about photography and how to theorize it or write its history that neither of us shared.

It was in this context that Georges showed a film he had made called L’Optogramme, a strangely shadowy, almost invisible ten minutes during which the faint sounds of barking, of a door opening and closing, and of footsteps were heard. The film was based on the late-19th-century conviction that the body itself is a kind of light-sensitive impressionable matrix, so that, in the case of the retina for example, the last thing seen just before death would be fixed, as in a photograph, on the surface of the eye. In this connection, a murdered woman had in fact been autopsied, in an attempt to shave the image of her assailant off the retinal film of her eyes. Not only did L’Optogramme have a wonderful daring, it showed with great efficiency George’s conviction, which I shared, that a history of photography that conceived of the medium only as a process for producing images, ignoring its condition as an index, or trace, or imprint, would be an impoverished one indeed.

Georges’ work has consistently been organized around the implications—both for historical art, ranging from antiquity through the Middle Ages to the Renaissance, and for contemporary practice—of the index. Ignored by historians and theorists in favor of the concept of imitation, the index is structured around a notion of the visual object as literally having been caused by its model: by having come into physical contact with it. As the symptom bodies forth an internal condition, as the Incarnation is the becoming-flesh of the Word, so can resemblance be a function of the life cast. The implications Georges has drawn from this indexical form of visuality have been reflected in studies he has done that range in subject from Fra Angelico to Donatello, from the Shroud of Turin to James Turrell.

As the first to have published Georges’ work in translation (see October no. 29, Summer 1984), I am particularly pleased to be able to introduce him to a wider English-speaking audience.

Rosalind Krauss


The history of art is a humanistic discipline.“ There are certainly many ways to interpret this famous proposition of Erwin Panofsky’s, but they all come together, it seems to me, in the assumption—whether optimistic (in keeping with the ”humanist“ side of Panofsky’s proposition) or authoritarian (on the ”disciplinary“ side)—of a flawless harmony between two words each themselves unfathomable: ”history“ and ”art.“1 In order for the discipline known as ”art history" to exist without too much difficulty, it was necessary that the stream Art should flow majestically—as if in the very nature of things—into the river History.

Art historians today, more or less conscious heirs to entire strata of “humanisms”—those of Vasari and of Kant, for example, which Panofsky’s essay sets in a continuum (though a rather strange one if you think about it)—for the most part no longer question this postulated harmony, which seems to advance by its own momentum and has at least two strengths, the first being that of self-evidence. (“A work of art belongs to a history”—who would argue with that?) The second, more troublesome, guarantees each “domain” of the discipline of history its own autonomy, perhaps even its autarchy—in any case its specific area of competence (“To each art its own history” ). By this recessional logic of identity—scarcely “humanist” in the old sense, though certainly as old as the positivist thought of the 19th century—it becomes justifiable, even desirable, to know nothing but one’s “own” period: to study Fra Angelico without looking at Jackson Pollock, Donatello without knowing Robert Morris’ name. (This does happen.) After all, isn’t it obvious that Fra Angelico is best explored through the ideas formulated in 1480 by Cristofaro Landino, the “best art critic” of the Quattrocento?2 Isn’t the humanistic sculpture of Donatello best understood through Alberti’s De statua, or Pomponius Gauricus’ De sculptura?

Actually, though, the real trick is to know up to what point such evidence remains valid. The “euchronic” model—the desire to understand the realities of the past through the concepts of the past, and to reject encroaching ideas from any other temporality right from the start—is just an ideal. After a certain point, the only thing it reveals is its function as suture and psychic defense—that is, its value as illusion (and I don’t just mean the illusion of modesty by which historians recognize, if only rapidly to forget, that they are of course “conditioned” by their own time). Even the thirty years between Fra Angelico’s San Marco frescoes in Florence and Landino’s esthetic judgments created an anachronism: Landino barely understood Fra Angelico’s practice, perhaps because the categories he negotiated as a vernacular “art critic” were practically the opposite of those of the artist, who was concerned more with memory than with “history,” more with the image in the theological sense than with “art.”3 As for the humanistic prose of Alberti’s De statua, which struggles bravely to be simultaneously contemporary and even narrowly linked to the work of Donatello, it in no way takes account of that work’s particular characteristics—of form, iconography, feeling, or facture.

Beyond humanist history and its “euchronic” dream, structuralist history, as we know, put in place a broader schema in which two dimensions intersected and articulated each other: diachrony and synchrony, the evolutionary and the structural points of view. Today we perhaps should note that it is not only within a diagram of pure, simple complementarity that diachrony and synchrony can intersect or articulate each other. And we should note the kind of rupture that is produced at their intersection: where diachrony and synchrony touch, there opens a rift, a critical moment, a symptom—an effect of anachronism. Perhaps we must think of this effect as a fundamental paradigm of historicity with regard to art.

Anachronism! Isn’t this the most plague-stricken word, the insult the historian fires off at that final moment when he tells the philosopher or the anthropologist that he has nothing to tell him or to learn from him, then slams the library door in his face?4 Isn’t this word the epitome of the interpretation gone astray, the word evoking Pollock before Fra Angelico’s Sacra conversazione, or Morris before Donatello’s Judith and Holofernes?

But let’s take this last example literally, absurd as it may seem: what do the Judith and a Morris work have in common? Obviously—apparently—nothing: these sculptures have alien histories, styles, and esthetics. Everything separates them. For the traditional historian (who puts Donatello at the heart of history, and still effectively refuses Morris entry into the paradise of the “humanist discipline”) as for the Modernist critic (who postulates Morris’ absolute break from the “humanist” parameters of ancient sculpture), there is nothing to see here (rien à voir, as the French so aptly say). In Morris and Donatello, two historical worlds obviously stand opposed. To put them in contact would be to create a firestorm—the firestorm of anachronism.

Yet this contact exists—it exists historically, and consequently requires the attention of an art historian. It exists in the work of Morris himself, who has looked long at Donatello’s sculpture, and has taken from it an interpretative direction whose pertinence, I would briefly suggest, is a hundred times greater than most of the academic writing about the creator of the Judith. In a remarkable essay in Artforum in 1970, Morris argued for the interest, even the necessity, of inscribing Donatello’s work into his own critical and esthetic project.5 This shocking inscription—this anachronism—arrives within an argument of which I must summarize here at least two or three significant elements. First, the takeoff point of Morris’ reflection is an acute critique of “the history of art as a humanistic discipline”—that is, before everything else (even before Panofsky), of Vasarian art history, which he defines as “a systemless collection of technical, anecdotal, or biographical facts that were fairly incidental to the real ‘work.’”6 The modern, academic offspring of this form of art history, still lively, privileges the “function” over the “making.”7

A second moment in this argument extends directly from the first, in that it restores a fundamental value to the notion of process. A Minimalist sculptor writing in 1970, Morris insists that the simplicity of a form (a cube, for example) does not exclude—indeed demands—what he aptly terms a “phenomenology of making.” In the classical esthetic, process (most often reduced to the notion of “technique”) is only a means toward the end that we call “form.” Morris, though, lucidly asserts that processes are themselves forms, forms in motion, forms not static and “formed” but “forming” (that is to say, “de-forming”).8 And a third point consists in using a general theoretical reference to name this dialectic between form and process. Looking to Saussurian linguistics, Morris breaks with the sempiternal questions of priority between language and image to try to characterize—following the model of the arbitrariness of the sign, and of its limitation—what is structured and what is erratic in formal processes.9

And then comes Donatello’s Judith. The sculpture appears, anachronistically but pertinently, in a passage in which Morris discusses certain artists—Marcel Duchamp, John Cage, Pollock—who have a need to put into play this dialectic of the organized and the arbitrary in their “making process”:

Duchamp, Cage, Pollock, Johns, Stella have all been involved, in different ways, in acknowledging process. Quite a few younger artists are continuing to manifest the making process in the end image. But the tendencies to give high priority to the behavior end of making can be found in much earlier artists. Rather than modeling parts of the costume in Judith and Holofernes, Donatello dipped cloth in hot wax and draped it over the Judith figure. This meant that in casting, the molten bronze had to burn out the cloth as well as the wax. In the process some of the cloth separated from the wax and the bronze replaced parr of the cloth, revealing its texture. This was a highly finished work, and corrections could have been made in the chasing phase had the artist wanted to cover it up. It has also been claimed that the legs of the Holofernes figure were simply cast from a model rather than worked up in the usual way. Evidence of process in this work is not very apparent and could have been noticed only by the initiated. . . . What is particular to Donatello and shared by many twentieth-century artists is that some part of the systematic making process has been automated. The employment of gravity and a kind of “controlled chance” has been shared by many since Donatello in the materials/process interaction.10

The questions that Morris raises here by placing the Judith in this strange—anachronistic—constellation (alongside such works as Duchamp’s Female Fig Leaf, 1951) seem to me considerable, and in the end considerably complex. Let’s begin with the simplest premise: Morris mentions that in two places—Judith’s veil and Holofernes’ naked legs—in the facture of the Judith Donatello used a process of casting from “life”—from cloth, or from a living person. Morris’ analysis, which is supported by direct observation as well as by the statue’s restorer,11 raises a question of esthetic censure that the restorer in question lacked the conceptual means to grasp, and that art historians today carefully avoid confronting. They are generally content to evoke what for them is a “technical detail,” which they trivially relegate to Donatellian “naturalism.”12

Yet they know quite well the shock value, the anachronism, that such casting might introduce into the axioms of “art history as a humanistic discipline.” It’s not just that the casting of Judith’s veil from real fabric would by definition ruin the idea of a style in which the successful carving of draped cloth has been the benchmark of achievement since the time of the Greeks; or that the casting of Holofernes’ legs would throw the ideal of mimetic skill in rendering the human body out the window. (Rodin, of course, was to have some problems for the same reasons.) More: the casting element in the “making” of the Judith would exclude it from the humanistic definition of sculpture. For Vasari, sculpture was an art that “takes away matter,” subjugating and reducing it to an idea already “drawn” in the artist’s mind. The cast, on the other hand, far from reducing matter, proceeds from matter to matter, and manages quite well without any preconceived “idea of form” whatsoever. And Gauricus on principal excluded anyone who worked with plaster, and particularly those base artisans who practiced casting, from the entire history of sculpture.13

Why this censure, this exclusion? Because the “phenomenology of casting” escapes humanistic and particularly Vasarian thought. Because it constitutes a basic obstacle to art history, if we understand by “art history” a discipline of total euchrony. In fact it is an obstacle to the traditional notion of art. Vasari and Gauricus had to exclude it, and Alberti to ignore it, because, strictly speaking, it doesn’t require any of art’s rhetorical criteria: neither invenzione, nor idea, nor even imitazione. In casting, it is enough to duplicate the referent by contact, without making the optical effort to “imitate.” Above all, casting is indifferent to the traditional notion of style: two legs, even if cast in the Quattrocento, have nothing in themselves that is “Renaissance.”

Since the temporality of casting’s indexical process (analogous to photography’s ça a été [this has been], according to Roland Barthes) engenders a dubious history only (how does one date a cast?), casting is also an obstacle to the traditional notion of history. It is a process productive of anachronism. This is why it is most often thrown back to that “before the time of art”—that ill-defined sphere of the artisan—of which Cennino Cennini spoke willingly and of which Alberti never spoke, 14 or else to an “after the time of art,” that sphere, again scorned, of so-called “nonworks” or Duchampian readymades. (In fixing a veil onto a mannequin with wax, and melting it in the bronze, what was Donatello making if not a readymade “in fusion”?)

Now, perhaps, we can see that Morris’ gaze at Donatello, with its declared anachronism—the Quattrocento artist’s name beside Duchamp’s and Pollock’s—isn’t concerned just with a “modern” interpretation of a historical work, or with a personal one, but with a certain state of knowledge in art history, a state that it actually alters. In fact Morris brings out an essential aspect—not a “technical detail”—of Donatello’s sculpture: the aspect of anachronism, camouflaged or disbelieved by art historians from Vasari to John Pope—Hennessy, who could not, didn’t want to, take it into account. This certainly wasn’t a matter of saying that Donatello and Duchamp are “doing the same thing,” under the archetypal umbrella of some nebulous transhistorical generality. Morris wouldn’t dream of that for a minute. It is a matter only of recognizing the power of a richly anachronistic moment (Walter Benjamin might have called it a “dialectical image”15) to reopen one’s eyes to a discipline—art history—that was historically and academically constituted, in the 16th century, on the refusal to see certain characteristics that were nevertheless essential to the work of certain artists, even “humanist” artists like Donatello. From something the art historian Pope-Hennessy still can’t take into account, through his allegiance to an esthetic axiom dating from the time of Vasari (dating, that is, from a hundred years after Donatello’s death; an axiom itself anachronistic), the artist Morris can make a new historical and theoretical object, though an authentically Donatellian one.

Maybe art history, this beautiful discipline, should occasionally pass through these moments, these anachronistic tests—dangerous moments, true, because they destabilize and violently “de-identify” their object, making it suddenly “strangely uncanny.” I imagine, for example, that Aby Warburg’s famous visit to the Pueblo Indians would have had for him this heuristic function of crisis, of fertile disorientation: the Pueblo serpent ritual may have allowed the great art historian to rethink the notion of Pathosformeln, and in some way to rebegin the history of the figurative rituals of the Florentine Renaissance.16 In the same way, couldn’t a voyage through Morris’ problematic—through the fascinating serpent of his exhibition at the Guggenheim Museum17— help scholars rebegin the history of earlier sculpture?

I say this not only because Morris, like Donatello, uses the casting process so well. There is something striking in Morris’ art—striking to the point where it troubles a number of viewers—that allows us better to see how Donatello worked: I would call this “thing” a heuristics, a term borrowed from epistemology. Whereas axiomatic knowledge is deduced entirely from propositions that are unproven yet are considered absolute, heuristic knowledge (the etymological reference is to the act of meeting, of finding by chance, of discovering) advances from a working hypothesis without worrying about whether that hypothesis is true or false, “correct” or “incorrect.” The obvious violence that Morris’ work does to the traditional notion of style is in no way negative: it is fundamentally heuristic, a way of being at once exuberant and severe, erratic and structured.18 And in looking at it carefully, one can start to approach the stylistic paradoxes of Donatello.

These too have a basically heuristic character. They also seem insensible to the axioms of humanism. Or, rather, they make humanism into a practice of excess, enlarged, open, and itself a producer of paradoxes—of Renaissance and medieval forms, iconic and indexical forms, optically conceived forms and forms created blind, “ideal” forms that aim for perfection and forms left to chance, closely mimetic forms and forms that are twisted and impossible (Morris particularly likes the empty glove in Donatello’s Saint Louis of Toulouse, which blesses the viewer without an arm to fill it), harmoniously complete forms and forms that are free, experimental, chosen at the last moment, left to their own devices—forms devoted to formlessness.

This is why, leaving the Morris exhibition at the Guggenheim, and recalling that article of 1970 in Artforum, I suddenly had the urge to leave for Italy and write a book about Donatello. The history of art is probably a heuristic discipline.

Translated from the French by Diana C. Stoll.



1. Erwin Panofsky, “The History of Art as a Humanistic Discipline,” 1940, Meaning in the Visual Arts: Papers in and on Art History, Garden City, N.Y.: Doubleday, 1955, pp. 1–25.

2. Michael Baxandall, Painting and Experience in Fifteenth Century Italy: A Primer in the Social History of Pictorial Style, Oxford and New York: Oxford University Press, 1974, pp. 147–51.

3. See Georges Didi-Huberman, Fra Angelico—Dissemblance et figuration, Paris: Flammarion, 1990, pp. 14–47, and Devant l’image: Question posée aux fins d’une histoire de l’art, Paris: Éditions de Minuit, 1990, pp. 48–54.

4. The French situation here concerns me directly: the only two art-history libraries in Paris—Doucet and the Louvre—refuse by sratute to share their materials with professors or researchers from the École des Hautes Etudes en Sciences Sociales, who might, for example, wish to study Donatello. This ideological censure of the too few tools of knowledge is just one sign among many of an extreme disjuncture, generating ignorance rather than debate, between “history” and “theory,” the study of ancient art and the study of contemporary art, etc.

5. Robert Morris, “Some Notes on the Phenomenology of Making: The Search for the Motivated,” Artforum VIII no. 8, April 1970, pp. 62–66. Republished in Continuous Project Altered Daily: The Writings of Robert Morris, Cambridge, Mass., and London: The MIT Press, 1993, pp. 71–93.

6. Ibid., p. 73.

7. Ibid., pp. 71–73.

8. Ibid., p. 73.

9. Ibid., pp. 78–83.

10. Ibid., pp. 86–87.

11. See B. Bearzi, “La tecnica fusoria di Donatello,” in Donatello e il suo tempo: Atti del VIII Convegno Internazionale de Studi sui Rinascimento, 1966, Florence and Padua: Istituto nazionale di Studi sui Rinascimento, 1968, pp. 97–105. Morris cites a less complete text by the same author.

12. See especially Artur Rosenauer, Donatello, Milan: Electa, 1993, p. 249, and John Wyndham Pope-Hennessy, Donatello, Sculptor, New York: Abbeville Press, 1993, p. 281.

13. Giorgio Vasari, in Le vite de’ piu eccellenti pittori, scultori e architettori, 1550/1568, ed. G. Milanesi, Florence: Sansony, 1878, I:148, writes, “La scultura è un’arte che, levando il superfluo dalla materia suggetta, la riduce a quella forma di corpo che nella idea dello artefice è disegnata” (Sculpture is an art that, by removing the superfluous from the initial material, reduces it to the bodily form that has been drawn in the artist’s mind). And Pomponius Gauricus, in De sculptura, 1504, ed. and trans. André Chastel and Robert Klein, Geneva: Droz, 1969, p. 250, writes, “As for sculpture, in plaster, which is barely an art, there is no one to mention [as an artist]” (Gypso autent ob tenue artificium, nullus).

14. On the humanistic “repression” of indexical techniques—in particular in the critical fortunes of Donatello, and in other examples besides the Judith—see my “Ressemblance mythifiée et ressemblance oubliée chez Vasari: la légende du portrait ‘sur le vif’,” in Mélanges de l’École française de Rome/Moyen Age–Temps modernes CVI no. 2, 1995.

15. See Walter Benjamin, Paris, capitale du XIXe siècle: Le livre des passages, ed. Rolf Teidemann, trans. Jean Lacoste, Paris: Cerf, 1989, pp. 474–94.

16. See Aby Warburg, “A Lecture on Serpent Ritual,” 1923, in Journal of the Warburg Institute II, 1938–39, pp. 277–92.

17. “Robert Morris: The Mind/Body Problem,” at the Solomon R. Guggenheim Museum, New York, 1994.

18. The epigraph Morris chose for “Some Notes on the Phenomenology of Making,” borrowed from Ferdinand de Saussure, could in a certain way be considered a heuristic injunction par excellence: “Between the two extremes—a minimum of organization and a minimum of arbitrariness—we find all possible varieties.”