- Semantics, Pragmatics, and Narrativity:A Response to David Rudrum
In his article "From Narrative Representation to Narrative Use: Towards the Limits of Definition," David Rudrum argues that definitions of narrative based on what the text represents are fundamentally flawed: "As long as narratology remains tied to [a conception of narrative as representation], and tied to a philosophy of language that foregrounds signification above and before questions of use and practice, it seems that a satisfactory way of defining and classifying its subject matter will continue to elude it" (203). Here, in a nutshell, is the argument. Narrative has traditionally been defined as the representation of a sequence of events. But this definition fails to capture the distinction between a set of instructions for building model airplanes (Figure Two in the text) and what Rudrum regards as a genuine narrative, namely a Calvin and Hobbes comic strip (Figure One). To distinguish the narrative status of these two representations of sequences of events, we must take into consideration how the text is used.
Defining narrative, in this perspective, consists of spelling out the rules of a specific "language game," to invoke, as Rudrum does, the celebrated metaphor through which Ludwig Wittgenstein anticipates speech act theory and current interest in questions of language use. Rudrum is not trying to tell us that what a text represents—its content, its signified—is irrelevant to its narrative status: "Perhaps [a representation of a sequence of events] is a necessary condition of narrative, but it [End Page 188] does not appear to be a sufficient one. Something more is needed to make a text a narrative" (198). But he regards context and use as far more important constituents of a definition than what the text is all about: "What gets identified as narrative (or not), is first and foremost a function of social conventions, rather than exclusively formal or linguistic concepts" (200; my emphasis). In other words, semantics plays second fiddle to pragmatics.
In this response, I would like to defend the primacy of semantics over pragmatics for the definition of narrative by taking a critical look at two of the arguments on which Rudrum builds his case.
The "Reading As" Argument
To describe narrative as a language game (or more precisely as a semiotic game, since Rudrum's examples come from the graphic domain) means that a given textual object can be put to several different uses, one of them "narrative," and the others something else. A text, in this view, is like a deck of cards: depending on the rules, you can use it to play Bridge, Canasta, Poker, or Hearts, and the rules of these games do not entirely depend on what the cards represent: kings could be replaced with cows and queens with horses, and the four suits could be bells, flags, acorns and coats of arms, without impact for the rules of the game, as long as suits are arranged in a specific hierarchical manner. In speech act theory, similarly, you should be able to take a proposition, for instance "the cat has food," and performs different illocutionary acts with it: assert it, ask about it, order it, or make it the object of a promise or threat. It would of course have to be a very big cat, and you or one of your loved ones would have to be the meal, for "the cat has food" to form a credible threat (not to mention embedding in an if . . . then construction): here the concrete situation and the particular reference determines the usability of the propositional content. This is precisely what Rudrum wants to tell us about narrative: the narrativity of a text depends on the context.1 To make his point, he imagines communicative situations that make it possible to read Figure Two as a story, and Figure One as a set of instructions.
But what are the rules of the narrative game? Though Rudrum never spells them out, we can reconstruct them from two sources: (1) his unequivocal labeling of Figure One as a story; and (2) his attempt to read the airplane instructions as narrative. According to Rudrum, Figure One "'tells' (if that is...