Picture Theory.

. . . confusing all the traditional divisions of labor.

- W. J. T. Mitchell, Picture Theory

Picture theory is about the invasion of image by text and of text by image in post-World War II American culture. It is also a book that conspicuously exemplifies what it talks about. W. J. T. Mitchell was originally trained at Johns Hopkins University, in what was at that time a conservative English department focusing on canonical works of English literature with a small admixture of American literature. Like many university people working in the humanities today, Mitchell is more or less self-taught in what he now most cares about doing. Though he is professor of both art and English literature at the University of Chicago, he is not doing quite what his Ph.D. authorized him to do.

Picture Theory is a prolonged investigation, by way of many diverse examples, of what Mitchell calls the "imagetext." Though he recognizes, of course, that words and images have always contaminated one another in the West, that the dream of a pure artifact that would be all word and no image or all image and no word has always been a chimera, nevertheless he argues that we are at a historical moment in the West when the visible and the readable are interacting in new ways. In these new varieties of media, picture may be even more important than word. Mitchell calls this a "pictorial turn." The long domination of the printed book is giving way to cinema, video, and other forms of multimedia as the determining factors in our culture. For Mitchell, the starting point for an investigation of this turn "is with language's entry into (or exit from) the pictorial field itself, a field understood as a complex medium that is always already mixed and heterogeneous, situated within institutions, histories, and discourses: the image understood, in short, as an imagetext." As this citation suggests, Mitchell is more a Foucauldian than anything else, though he draws on a wide variety of theoretical predecessors: Ludwig Wittgenstein, Nelson Goodman, Murray Krieger, and so on. His approach to the imagetext question is subtle, eclectic, commonsensical, and passionately committed to the belief that imagetexts are not only always historical but that they also matter because they have a role in making history.

Mitchell's book quietly resists the overall generalizations of theory: "The concepts of the meta-picture [as in Magritte's La Trahison des images (ceci n'est pas une pipe), discussed at length] and the imagetext . . . are simply crystallizations of what has become common sense in postmodern picture theory." The relations between word and image are for Mitchell always complex, and different in each case. This means that "the power of the imagetext is to reveal the inescapable heterogeneity of representation." The "'wrinkles' and differences in representation," he says, "its suturing together of politics, economics, semiotics, and esthetics, its ragged, improvised transitions between codes and conventions, between media and genres, between sensory channels and imagined experiences are constitutive of its totality. The theory of representation appropriate to such a model would have to be a pragmatic, localized, heterogeneous, and improvisatory totality." Mitchell's book is heterogeneous in the same way. He cares more for the fascination of exploring his examples in detail, with flexibility and pragmatic common sense, than for working out any new commanding theory that might transform the field. And I care more for that too.

Nevertheless, there is a central and original insight that unostentatiously governs Mitchell's discussion of his examples. This is most overtly stated in the brilliant chapter titled "Ekphrasis and the Other." For Mitchell, the relations between image and word are always "paradoxical." Word and image seem to get along for the most part harmoniously enough, at least until you begin thinking seriously about their relation. Names of pictures often seem adequate enough as labels, for example, and illustrations for novels really do illustrate the novels. Talking movies have long since been accepted. They did not destroy the power of an exclusively pictorial medium, as once was feared. The long tradition of ekphrastic poetry, which describes an artwork, imaginary or real, shows that the power of language is by no means struck dumb by pictures or other nonverbal artifacts. Nevertheless, for each medium, according to Mitchell, the other medium is to some degree opaque. Each medium is ultimately, at some farthest reach where their easy cooperation no longer works, an unknowable and unrepresentable other. "The ekphrastic image acts, in other words," says Mitchell, "like a sort of unapproachable and unpresentable 'black hole' in the verbal structure, entirely absent from it, but shaping and affecting it in fundamental ways." This black hole theory of the image/text relation is of great interest. Some readers (this one for example) may be sorry that Mitchell did not make more of it.

Provocative as the black hole theory is, however, perhaps even more interesting and significant is the range of examples Mitchell discusses. This is the most challenging aspect of what he calls the "de-disciplinary effort" in which his work participates. If the "pictorial turn" means a "new significance of visual culture," he says, "literary studies cannot simply 'add on' the study of film, television, and mass culture to its lists of courses without changing the whole map of the discipline, and it cannot stabilize its new relation to art history and visual culture with paradigms of structural comparison." Picture Theory is not at all like books published twenty or thirty years ago when the disciplines were more firmly in place. It is neither a work of art history, nor a work of literary criticism, nor a work of theory within a single discipline (like Frye's Anatomy of Criticism), nor a work of philosophy, nor a work of comparison between the sister arts of poetry and painting/sculpture (Mitchell is sharply critical of such comparisons), nor a work of cultural study, though it intermittently seems like all of these. Picture Theory belongs rather to a new transdisciplinary regime that heralds, like so much else in the humanities and social sciences today (for example, the rise of cultural studies), the rapid dismantling of the separate integrity of the old disciplines. Picture Theory discusses, among other works, artists, and writers, Saul Steinberg's drawing The Spiral, a New Yorker cartoon by Alain, Wittgenstein's Duck-Rabbit drawing, Velazquez's Las Meninas, Magritte's La Trahison des images, a Mad magazine cover, William Blake, Wallace Stevens, William Carlos Williams, Percy Bysshe Shelley, Doris Lessing, Homer, Wordsworth, Toni Morrison, Robert Morris, various photographic essays, Nelson Goodman, Spike Lee's film Do the Right Thing, CNN coverage of the Gulf War, and Oliver Stone's JFK - an incoherent list if measured by old disciplinary standards. Just what institutional forms will take the place of the old disciplines we do not yet quite know, nor do we know just what training will be appropriate to prepare to work in the new dedisciplinized humanities. It certainly will not be the kind of training Mitchell received in the old English department at Johns Hopkins. The development of protocols for such training is a major task in universities today, but Picture Theory foreshadows the shape of things to come or already here.

J. Hillis Miller is Distinguished Professor of English and Comparative Literature at the University of California, Irvine.