Mitchell, W.J.T. Picture Theory. Chicago, IL: University of Chicago Press, 1994.
Mitchell moves from the 'metapicture' to the 'imagetext' (sometimes 'image/text') and, oddly, only just skims over the comic book. The imagetext is his recommended approach to handling metapictures, and indeed any art that mixes pictures and text.
Metapictures are images that, like any 'meta-' art, call attention to their material and formal nature. The famouse duck/rabbit and Magritte's "Treason of Images", for example, both force the viewer to contemplate how pictures work. In the first case, the image refuses to be either a duck or a rabbit, but it also refuses to be both simultaneously. Its nature requires that the viewer acknowledge that the picture is a two-dimensional representation, and therefore can be contrived in two dimensions to defy the laws that objects in three dimensions have to obey. Most optical illusions do that, in fact.
"Treason of Images" explicitly informs the viewer that representation and reality are not the same thing. That is not a pipe at all. It's a realistic drawing of a pipe. In fact, what you've looked at is a scanned image of a realistic drawing of a pipe. The painting forces us to recognise that we treat images as transparent signifiers. The image just is what it stands for. We mistake analogy (this thing is both like and unlike that thing, and the differences are just as important as the similarities) with allegory (this thing is that thing). This process, forcing a viewer to see representation for what it is, contrivance, is almost always what we mean when we talk about metafiction or metacinema or meta- anything. It looks you in the eye and says "I'm not real."
The imagetext is a tougher concept. Mitchell's concerned with how to analyse those moments in art when text and picture are invoked in the same context. He charges that the most common method up to then (1994) was interdisciplinary, largely consisting of people from English dabbling in visual theory (hi!), though something similar happened on the Art History side of things in which text is used as the ultimate explicatory medium for pictures.
The real problem, as Mitchell sees it, is that such an approach is "grounded in the sense that art history provides a 'visual analogue' to the study of literature" (84) and "that verbal and visual media are to be seen as distinct, separate, and parallel spheres that converge only at some higher level of abstraction" (85). For example, this approach leads us to the conclusion that Romantic painting and poetry were attempting to do essentially the same things, one in text and other in pigments, and that the 'real' relationship between them is conceptual. The result of that kind of study is that we place anything new into old categories, merely confirming what we think we already know. Like any categorical approach, it means that if we encounter something radically new, we have to either construct yet another category, or pretend that it doesn't exist, and our first instinct is usually to do the latter.
He breaks his objection down to three specific things. First, the interdisciplinary approach presumes "the unifying, homogeneous concept," that everything follows a set of master rules that we contrived before having seen all the examples. Second, it "ignores other forms of relationship" than comparison, failing to see not just difference and contradiction, but totally different things like metonymic relations or even the recognition that the relationship is entirely unmediated or non-negotiable. Finally, it enacts "ritualistic historicism," the tendency to merely confirm the pre-existing historical categories, slotting new things into old boxes, which I explained above, a move that also tends to towards committing the fallacy of teleology.
The imagetext, as he explains it, is his alternative to the above analytical practise, one that grounds itself in the "heterogeneity of represetantional structures within the field of the visible and the readable" (88), and to "engage with vernacular form of representation" (88-89). By the 'vernacular,' he means looking at art forms in their historical and conceptual context, deriving conclusions from them, rather than imposing theory on them (and anyone who reads this blog regularly knows that that's music to my ears). As he says, the "best preventative to comparative methods is an insistence on literalness and materiality" (90), meaning that we ought to locate our analysese in the primary material, how it actually functions, in its concrete time and place. In a sense, he's talking about studying those things that Benjamin calls the aura, the time/space position occupied by the art itself. (Of course, the aura is a concept, not a thing, so either that comparison isn't all that apt, or Mitchell's call to pay attention to the material isn't as concrete as he'd have us believe.)
The last bit of his argument is a little harder to swallow, and even he admits he hasn't exactly 'proved' that his premise is true in a 'literal and material' sense, but it's an interesting place to start. He argues, put simply, that all art is mixed-media, specifically that it's all enmeshed in the picture/text interpenetration. He never quite demonstrates that this is true of paintings. He points out that text is often illustrated in visual arts, that even paintings titled "Untitled" have titles, and that the institution of art history is inherently textual. However, that doesn't convince me that pictures always contain text, or require text for their mediation. Animals, creatures without speech as we know it, can relate to pictures in observable ways.
The flip side, though, that text is an inherently visual medium is absolutely correct, and a point that most people overlook on a regular basis. "Writing, in its physical, graphic form, is an inseperable suturing of the visual and the verbal, the 'imagetext' incarnate." It "deconstructs the possibility of a pure image or a pure text" (95), because text is a kind of image. Proving this claim is as easy as writing the same words, any words, in two radically different fonts. The stylistic (pictorial) aspect of the text on the page does make a difference to how we read it. Three years of marking hand-written exams has taught me that neat, clean handwriting can make an essay seem more intelligent than it is, and vice-versa. In the opening scene of Great Expectations, Pip reads his parents' names on their tombstones. Having never met them, the narration explains, he builds a perception of what the look like based on the shapes of the letters.
The imagetext approach is one that requires looking at the specifics of any given piece (not just a 'text' anymore) and how it deploys and employs, invokes and evokes the visual and the textual. It pushes us to look at not just similarity and difference, but different kinds of relationships, including antagonism, dissonance, division, tension, juxtaposition, contradiction, and the previously mentioned possibility that there is no relationship we can describe (which itself a kind relationship), that they are incommensurable.
As a final note, out of interest, Mitchell identifies a feature of comics that has been on my mind lately. In his words, most explicitly mixed-media arts follow "traditional formulas involving the clear subordination and suturing of one medium to the other, often with a straightforward division of labor. In the typical comic strip, word is to image as speech (or thought) is to action and bodies. Language appears in a speech-balloon emanating from the speaker's mouth, or a thought-cloud emerging from the thinker's head. […] Narrative diegesis [I.e., narration] (cp, Prince Valiant's "Our Story…") is generally located in the margins of the image, in a position understood to be 'outside' the present moment of depicted action, scenes, and bodies" (91-92).
What Mitchell claims here is generally true. For the most part, text and picture do not exist on the same plane in comics. In fact, in the industry-standard construction of American comics, the pencillor illustrates the panels without word balloons or text boxes, and the letterer adds those after the fact, covering the art. The text is literally a different image on the page, another layer on top of the picture. Somewhere, for every American comic book, there are art pages that don't have word balloons at all.
There are instances, many of them, of the mixing of text and picture, though. The separation isn't quite as stark as Mitchell implies. The title page of an issue, for example, is traditionally a splash page with the title of the story, the list of creators, and the hero's name and/or title of the series written in visually arresting letters. Think of the familiar three-dimensional 'Superman' text, for an example (that example is a cover page, not a title page). Those words simply hang in the air, though they're still not part of the image, they're not separated by thick lines and boxes.
For a truly radical approach, however, you need to look at the comics of David Mack. Though his website seems to consist of nothing but links that do nothing and potraits of his characters [damn you, David Mack's web master!], in his mini-arc "Echo" for Marvel's ongoing Daredevil (volume III), he created a series of stream-of-consciousness images that travel through the memories of a deaf girl who takes the name 'Echo.' The narrative content is fascinating in the context of the series, but the form of the story is stunning. It does precisely what Mitchell claims comics don't do. It places text and image on the same plane most of the time, relenting to a narrator box only occasionally. Echo's words and feelings are rendered in pigments and textures that are the same as those of the surrounding images and directly interact with them.
Mack also dispenses with traditional panel structure much of the time. His pages are layered, image on top of image, but those images are often not separated by 'artificial' devices like lines and boxes, or if they are, the colours and textures of the art is used to demark the panels. If an image is all drawn in crayon, then the edge of the panel is also in crayon, for example. He, in one sense, 'flattens' his pages so that all the imagery, textual or pictorial, is made of the same stuff, the same colours and textures. When I read that arc, I realised that it is the logical extension of comic books as an art form. It gets quite close to fully integrating the pictorial and the textual such that it's almost impossible to say where one ends and another begins. Mack's work is the imagetext approach from the artist's position.
Posted by orion at July 11, 2006 6:53 PM
