Talking pictures (no vision needed)

Tom Wolfe's famous jibe that "without a theory to go with it, I can't see a painting" is quoted in this book, but the author (professor of English at the University of Chicago and editor of Critical Inquiry) evidently does not see the point. Or if he does, he prefers not to address it. That he has missed it altogether is the more likely hypothesis, since he translates Wolfe's claim into the peudo-Horatian maxim ut pictura theoria. Whereas Wolfe was saying the opposite: ut theoria pictura.

In spite of the title Picture Theory, Mitchell never intended, he tells us, "to produce a 'picture theory' (much less a theory of pictures)". So why the title? It turns out to be the sillier kind of word game one usually associates with Derrida, of whom Mitchell is evidently a keen admirer. We have to read picture as noun and verb simultaneously. Then we can understand that Mitchell is in the business of picturing theory. Theory about pictures, that is. Got it? (If you are too slow-witted, Mitchell actually explains the joke for you.) In spite of the disclaimers, Mitchell allows himself to be hailed on the dust jacket as "one of America's leading theorists of visual representation". It therefore should come as no surprise to the reader to discover that he has no theory of visual representation at all. He discusses -- very occasionally -- writers who do have a theory of visual representation, but he seems not even to have a theory about their theories. In short, he does not apparently draw any distinction between theorising and rabbiting endlessly on about the views of others.

When he does discuss the work of those who at least understand what theorising is all about, he shows an alarming tendency to be on the wrong wavelength. He says, for example, of Nelson Goodman that Goodman offers "a clarifying respite from metaphysical distinctions like space and time or nature and convention" and tries to emulate in the field of art Saussure's achievement in linguistics, ie, to exclude "linguistic change from the study of language". As the Duke of Wellington said to the man who mistook him for Mr Jones, if you can believe that, sir, you can believe anything.

Mitchell's book, nevertheless, has theoretical pretensions of its own. It claims to be a work of "applied iconology". But this iconology is neither the study of the paintings which feature as religious objects in eastern branches of Christianity, nor one half of Panofsky's well-known distinction between iconology and iconography. To understand what pure iconology is, we are referred to Mitchell's earlier book on that subject, Iconology. The difference between pure and applied iconology, it appears, is that the former is about images whereas the latter is about pictures. Pictures are "the concrete, representational object in which images appear". But by the time we reach chapter two, entitled "Metapictures", Mitchell has obviously already become confused by his own distinction. This chapter should have been called "Metaimages". In all of the examples discussed, from Vel zquez to the duck-rabbit illusion, it is the image that is self-referential, not the picture. Magritte's famous painting of the pipe that is not a pipe (the French title twice misspelled and again for good measure in the index) is said to achieve "self-reflexivity". This presumably means that the picture now hanging on the wall in an American museum has acquired a reflexive self. Magritte evidently did better than he knew.

Actually looking at pictures is the last of Mitchell's concerns. He himself said of Iconology that if it contained any insight, it was of the kind that "might come to a blind listener, overhearing the conversation of sighted speakers talking about images." Much the same could be said of his new book. When the "concrete representational objects" do eventually come up for discussion, they are treated in a curiously unconcrete manner. Although this book has many more illustrations than Iconology, it is interesting to note that only rarely are the relevant dimensions given. There is, for instance, a whole chapter on "Blake's Art of Writing" with 11 illustrations, some full-page, others half-page or less, but at no point is the reader told what the size of the original was. It is as if the "concrete" page were simply irrelevant to a discussion of Blake's calligraphy and compositional techniques.

Artists, in brief, have not succeeded in convincing Mitchell that there is anything of interest about their concrete products at all. They become of interest only when someone verbalises about them. At that point, Mitchell comes in as host to his own chat-programme, and the guests on the show are other chat-programme professionals. With disarming candour, he tells you what their professional qualifications are: they have to be able to write "that curious hybrid of mainly prose discourse compounded from aesthetics and other branches of philosophy, as well as from literary criticism, linguistics, the natural and social sciences, psychology, history, political thought and religion." The result is well exemplified in Mitchell's own flabby, jargon-ridden English, full of "sutures", "otherness" and "ideologemes".

Writing, as one might suppose, is a recurrent theme in Mitchell's essays. But he does not seem to have gone into the theory of writing any more deeply than into the theory of visual representation. He repeatedly refers to Homer's audience as "readers". And out comes the hoary old story about writing originating in "pictures". This is good enough for Mitchell's purposes, because it can be used to back up one of his more revolutionary sounding claims: that "all media are mixed media". The argument produced in support of this thesis is itself pathetic: since words conjure up images, and images conjure up words, the two are inseparable. One might as well argue that if restaurants serve up dishes that are described on the menu, that proves that eating is actually a verbal experience.

In short, Mitchell's main concern is not pictures, or even art, but the cocoon of verbiage that surrounds both. His own contributions to the verbiage will delight readers who collect items for Pseuds' Corner. They include such priceless gems as "Description might be thought of as the moment in narration when the technology of memory threatens to collapse into the materiality of its means" and "The image/text is neither a method nor a guarantee of historical discovery; it is more like an aperture or cleavage where history might slip through the cracks."

What the book lacks in theoretical acumen and clarity, however, it makes up for in unrelenting political correctness. In ringing tones, the champion of visual representation tells us: "Like the masses, the colonised, the powerless and voiceless everywhere, visual representation cannot represent itself; it must be represented by discourse."

The publication is in fact a motley assortment of articles by Mitchell already published elsewhere. They range from pieces on "slave narrative" to the television coverage of the Gulf War. What was the point of disinterring them? Their author modestly hopes that, collected together, they might serve as "an introduction to a discipline (the general study of representations) that does not exist and never will". Amen to that.

Roy Harris is emeritus professor of general linguistics, University of Oxford.