Formalism and Time

The most prominent generic features of the novel have received remarkably little systematic attention in our criticism. "A lengthy fictional prose narrative": that is the definition of the novel I learned in high school, and I am still mulling over each one of its terms. In Nobody's Story I tried explicating as well as historicizing fictional, and I have recently been pondering lengthy, the most thoroughly neglected word in the definition. Length has generally been treated by theorists of narrative under the headings of time, temporality, and duration, and a cursory survey of these analyses reveals numerous ways of rendering sequence simultaneous. For example, Mieke Bal, in considering the relation between the time of an element of a fabula and the time of its narration, argues that we should examine only the relative patterning: "The attention paid to the various elements gives us a picture of the vision on the fabula which is being communicated to the reader." ¹ Since the relation of parts to each other is the relevant question, the length of the novel ceases to count; the internal pattern of The Last Chronicle of Barsetshire may be set down as concisely as that of The Turn of the Screw. Nothing in this sort of temporal analysis would help us develop a concept of length.

Bal's procedure is typical of narratologists, whose fondness for graphs and charts is notorious. One, for example, graphs the rhythm of long stretches of narrative time; another pictures successive ideological choices as the corners of a single box; yet another compresses the infinite variety of agents and acts that might be encountered over time in any possible narrative into six basic actants. Just as charts are typical of formalist studies of narrative structure, attention to the grammatical features of single sentences characterizes analyses of narrative discourse. Charts and sentences attest to our need to make the narrative object as short as possible. As David Carroll pointed out over a decade ago, they appeal to our continuing desire to see, in a single, instantaneous act of perception, things that are, in fact, not visible, and he correctly asserted that such formalisms rest on an oculocentric bias. ² I would like to discuss another bias that seems to me implicit in them: a bias against the very thing under analysis, that is, extended temporal sequence--length--itself. Formalist analyses seem bent on showing that, although a novel represents temporal sequence by means of temporal sequence, it nevertheless has, or should have, a form that can be made apprehensible all at once, in a picture or a fractal.

I have already described two quite different kinds of narrative analysis as "formalist." Narratological graphs and charts represent forms in the sense of structures that organize, arrange, or order the parts of a narrative. This sense of form is perhaps the oldest use of the term in aesthetics, ³ and when Fredric Jameson makes a Greimasian diagram of the ideological elements to be combined in Balzac's Cousine Bette, or when Gerard Genette lists the relative speeds of events in Proust's A la recherche du temps perdu, they are engaged in the classical activity of displaying the overall shape, indeed the symmetry or shapeliness, of these novels. However, the Russian formalists, as well as the more recent analysts of narrative discourse, often mean something different by form: they mean the style of the work, the grammar, syntax, verb modes and tenses, and rhetoric. ⁴ Their analyses descend from late antiquity and usually involve claims for the specialness of literary, or fictional, style, specifying its significant departures from normal narrative grammar.

We might think of these two ideas of form as opposites, for a picture of the general relation of a narrative's elements results from a process of abstraction, whereas the features of its individual sentences are supposedly more sensually immediate and concrete. Form as arrangement or structure seems molar, an outline of the whole; form as style seems molecular...