The analysis of style

Home / Education / ARChives / Discussions

The analysis of style


Published on before 2005


I've often been intrigued by the idea of recreating an artist's style in the cinematic medium. Back in October, I suggested, working from Ted Seth Jacob's ideas of symbolic and representational drawing, that an examination of what could be called "visual stereotypes" might deliver benefits in art authentication (rather, stylistic fingerprinting). A "visual stereotype" is the tendency of an artist to interpret the perceptual world in terms that can be translated onto canvas, paper, wax, marble, etc. To paraphrase, we never draw what we see, but what we think we see. This "thinking" about what we see is what led me to use the much-reviled phrase "visual philosophy"; that is to say, that a work of art represents a painter's total thoughts about the perceived world translated through his medium. What offended was the use of "thought" for a non-verbal conception of a primarily visual process; that is to say, some misunderstood me to mean that when I said a painting had a "visual philosophy" that I meant it could be translated into the language of (verbal) philosophy; that it was in some sense crudely literal.

I failed to faithfully summarize Jacob's actual conception: that is, that in the process of being faced with certain complex visual phenomena - such as the nude, for example - our aesthetic understanding of WHAT we are seeing is always mediated by preconceptions of what we THINK we see. The problem isn't that acquiring knowledge of the body, say through anatomy, is an evil, but on the contrary, that it is our lack of understanding that forces us to rely on certain presuppositions, stereotypes about form, or what Jacobs brilliantly diagnoses as a "symbolic" or "word-based" interpretation of reality. He instances the child's stick figure, and calls this an example of "symbolic" drawing. It is a drawing that is only tenuously related to the appearance of the model. Just as seagulls seem to accept abstract representations of their mothers' beaks in place of the genuine article, so the mind seems predisposed towards certain rigid or fixed "stereotypes" or "symbols" of living forms. These symbols are "word-like" in the sense that the draughtsman divides parts of the form into symbolic units, as is clear in primitive art. Citing Jacobs on the symbol-word connection out of context caused some confusion: it seems that people thought he was suggesting this "linguistic" interpretation of the visual world implied a semantic or grammatical structure; that it was some form of post-structural semiotics insinuating itself into visual art; a cold-blooded abstraction seeking to suck the life from painting, and set the Word in the throne of another realm.

Symbol itself might be a debatable choice, as your distinction between "allegory" and "symbolism" in Tolkien suggest. I proposed "visual stereotype" in the place of "symbol" as it seems to me that Jacobs was referring to the stock quality of visual symbols rather than their conduciveness to conduct the general through the particular; ie., to become particular manifestations and eternal symbols of the divine. Which is where, incidentally, I think Stefan's objections to Symbolism founder; they are a rationalist materialist insistence that everything can be isomorphically signified, when the very things that are most vital to contemplate do not necessarily lend themselves to naturalistic treatment. We may want to cut straight to the content, but the form itself - the myth, the symbol, the fairy-tale - rather than being an obfuscation or childish reluctance to tell it plainly is rather the only way that the general can find its particular manifestation. "Visual stereotype" implies the formulaic quality of the mind when faced with the near-infinite complexity of the visual world.

What makes the process of drawing, or representing, the world an almost spiritual undertaking is not merely the subject matter of the art-work; the process becomes the attempt to free the mind of its preconceptions of form, to sympathetically apprehend the object. Drawing, then, becomes a means of impersonation of the visual world through the total understanding of the artist; considered separately from the ideative or narrative intentions of the artist. The quixotic attempt to realize the visual world through a medium is the mind's struggle to enter into the uniqueness of the object itself; in other words, to capture its character, its essence, its gesture, its soul. This is why a humble still-life, a drawing of a jack rabbit, a patient representation of a sunbeam moving over a cheerless hillside, is laudable and self-justified. The very ACT of drawing becomes a means of sympathetic identification between subject and object.

Which brings me back to my original point: if we could compile a codex, of sorts, of an artist's stereotypes, by paying close attention to how he uses the resourceful instruments of line, shading, edge-definition, and colour to convey visual and tactile qualities of the real world ... we could then test our understanding of his style by programming our discoveries into an object-modeler. "Test" in the sense that the quest for artificial intelligence is in reality a computational model of our scientific understanding of the brain; once AI approaches were expert engines based on a rule-based interpretation of the mind; later experiments have attempt to digitally recreate the basic neurological unit, the neuron; all of these attempts are MODELS for the sophistication, or lack of, of our scientific understanding of mental processes. In a sense, modelling our codifications of an artist's visual stereotypes through graphics software would be one crude, but striking way to see how close or dismally far our approximations come.

To quote from my initial letter on visual philosophy:

Along this line, then, some paramount questions: (1) What are the qualities of line? I.e., how exactly do different Masters use line to describe such things as: space, volume, mass, texture, materiality (hard & soft), colour, atmospherics, edges, relief modelling, movement and action, underlying forms (bones, muscles), torsion, transparency and opaqueness, and rhythm? (2) What "facts of drawing" do certain Masters concern themselves with from the start of the drawing? I.e., is Ingres concerned with the sculptural form of his sitters, Rembrandt with the fluctuations of light, Mantegna with orchestrated spaces, etc.? (3) How are these "facts of drawing" assembled to create a total impression of artistic truth?

These aren't questions I'm interested in the abstract, but to which I want concrete and reproducible answers. I am not interested in squeezing the body of facts into the corset of preconceived theory: all propositions must be proved with strict reference to actual marks on paper or canvas. The only proofs that are valid under this investigation are visual proofs: artistic canons, aesthetic theories, etc., must take a second place to the vital question: What is the (visual) meaning of certain qualities and usages of lines, washes of colour, brush-strokes, etc., in the creation of artistic truths?

If our codification of an artist's stereotypes were sophisticated enough, we could attempt to apply them to a three-dimensional form scanned into a computer. This compilation of stereotypes would extend to the most basic questions of "what type of line is used to render the circumference of a sphere against variously toned backgrounds", to, "what conventions does the artist typically employ to represent an eye, ear, nose, curved lip", all the way to "what mannered system of proportions does he use to represent the body itself"? There would be stereotypes or structural forms for atmospheric perspective, for the shape of tree trunks, for the consistency of materials (liquid, bronze, flesh, hair, etc.)

Of course these are potentially endless; but we might restrict our scope to a small set of objects: apples, carnations, a young woman's hand, a portrait bust, etc. Then, working with a three-dimensional model in the computer, the programme would consult this codex of stereotypes or structured forms to alter the shape, proportions, of the object in its memory. Graphics rendering programmes can already simulate the effects of light passing through semi-opaque layers of skin, capillaries, oil paint, etc. The result then would not need be photo-realistic, but rather truly stylistic.

I'm not proposing anybody try this at home. It would probably be a colossal waste of time.