All PostsConte, Joseph. “American Oulipo: Proceduralism in the novels of Gilbert Sorrentino, Harry Mathews, and John Barth.” Chapter 4 of Design and Debris: a chaotics of postmodern American fiction. Tuscaloosa: U of Alabama, 2002. 75-111.
Conte discusses constraint in the context of his work on design and disorder, chaos and order, much of which does not concern me here, but he does offer some discussion of constraint that is worth looking at a little closer, particularly in the relevance to more “formal” constraints which are not as often discussed.
His term is “proceduralism” which he defines as “the adoption of a rigorous and efficient design in advance of the composition of the work of literature” (76) or “a book constructed according to a set of arbitrary and precise rules, arrived at in advance by their author with narrative consequences that are both generated and controlled by those rules” (77), and he contrasts it to “formalism” which “establishes the characteristics of a text inherited within a literary tradition” (76). His formalism is basically the workings of convention, though that definition is not necessarily read in such a narrow way. He explicitly links his “proceduralism” to “the classical era… that gave us the palindrome and acrostic” (76) as well as the Oulipo, which also links it to a literary tradition of a sort. The difference here is one that is a tradition of heterogeneity rather than homogeneity.
An important part of Conte’s proceduralism is the generative aspect of the design. The constraints not only control the text, they work as a machine to create the text, what Conte calls the “generative constraint”. This aspect of many constraints (just picking one’s I have already discussed: Roussel’s method and Calvino’s tarot cards in The Castle of Crossed Destiny) is the almost paradoxical work of limiting and generating. The generative constraint has a possibility for creating an output which exceeds the work put into it, a text that almost “writes itself”, though no guarantee can be made of such. It is just as likely that a constraint will be nothing more than what is put into it. In this sense, the generative constraint is like a machine for writing, a function in the language system.
With regards to the Oulipo he makes an important point regarding most of their early and well-known constraints: that the “rule-based constraints usually have their greatest effect on what I call compositional method” and are “syntactic or strategic” (77). The syntactic constraints effect how the text is composed of words — “the material structure of the language… is most noticably altered” (78) — while the formal constraint, the proceduralism, works at a larger level — “emphasiz[ing] preconceptions of global structure in the novel” (78) — the micro and macro aspects of the text, respectively. He observes that while the “American proceduralists” use extraliterary forms, the Oulipo more often use “syntactic or algorithmic devices”. This division is a bit specious as Mathews himself falls under both categories and indeed has done work with all those types of constraints.
Too easily falling into these categories I realize that I, an American non-Oulipian, am more interested, for my own writing, in the formal constraint, the macro level, rather than the syntactic constraint, the micro level.
He continues on with further generalizations about formal constraints: that they must be new, never done before; that they be arbitrary “unmotivated by any resemblance to theme, character, or the author’s biography” (83); that they are for “one time only” use. All these generalizations amount to a preference more than anything else. Why couldn’t any of these rules be contradicted? Certainly, there is room for reusing constraints, very few would be exhausted in one use, and I see no reason why a constraint couldn’t work in some way with theme, character, or the author’s biography, though in the sense that the constraint comes before the writing of the work, the connection with character seems least likely before the fact.
Conte gets into some chaos theory and Mandelbrot sets that I will leave to those better at comprehending such. He then goes on to discuss Sorrentino’s Pack of Lies trilogy, Mathews’ The Journalist, and Barth’s The Last Voyage of Somebody the Sailor, none of which I have read recently, so I will leave that for another time (all are books I want to reread, though I think Barth’s LETTERS is a better example of his writing under a formal constraint).