Rules for Reading

In a century marked by literary narratives declaring and sometimes celebrating multiplicity and indeterminacy, several critical traditions have sought a single, stable structure to define narrative form. Structuralist narratologists developed systematic descriptions of narrative form, and computer-aided research in the humanities, with its empirically-based studies of the linguistic skeleton of narratives, extended this tradition. Researchers in both fields hypothesize that there exists a formal, rule-based structure that constitutes the minimum condition of narrative, a schema that can be extracted, studied and replicated. Although most narratologists no longer subscribe to this theory, much computer-aided research continues the search for a Platonic ideal. The purpose of this paper is twofold: to suggest new methods and goals for computer-aided studies of narrative and to demonstrate that visualizations of the process of interpretation might help us better understand narrative. Contemporary narratologists propose a division between classical and postclassical narratology1. Current computer-aided study of narrative is grounded in classical narratology, a tradition informed by structuralism and interested in the elaboration of a universal structure of narrative. In building a description of the archetypal narrative, classical narratologists ask two basic questions: 1) what is the most basic unit of narrative; 2) what set of rules constructs a unified structure from those units? In an answer to this question classical narratologists think they will uncover the skeleton that characterizes all instances of what readers identify as narrative. These structuralist notions still obtain in most applications of computing technologies to the study of narrative. Such research often separates form and meaning; much research proposes narrative units that can be identified and quantified, rendering a static—even statistical—accounting of narrative form2,3. Classical narratological models of narrative are attractive—discrete units and formal rules are easily computable—but perhaps dangerous for our discipline. Narratologists suspicious of structuralist notions of narrative form have proposed various new paradigms for the investigation of narrative1,4. Represented by a wide variety of approaches—feminist poetics to artificial intelligence—postclassical narratology in its most general sense involves an investigation of the methods by which readers invest strings of events with narrative structure. Drawing on cognitive models of reading, postclassical narratology posits narrative as a sequence of cues that promotes some structuring activity. The structuring activity may be rule-based, but the rules reflect preferences rather than unambiguous and universal forms5,6. In its shift from form to process, postclassical narratology questions common distinctions between reading and interpretation, arguing that the former is not prior to and free of the latter.
75
This questioning of the boundaries between reading and interpretation is crucial to postclassical narratology and to textual studies. Postclassical narratologists criticize structuralist models because the end is always already known—intermediate events are only interpreted in light of a known ending. The “retrospective” form that results ignores a crucial dynamic of reading—the projection an ending that makes sense of intermediate events, an ending that is constantly re-imagined as the intermediate events resist previous interpretations7. Thus reading might be understood as a time-based process of constant re-interpretation in light of an anticipated ending, and literary narrative as a textual collection of cues that trigger this process. Two decades ago Stanley Fish argued that we misuse the word “read,” that we pretend that there exists a text that can be perceived prior to interpretation and that reading involves some unstructured experience of that text8. In this sense a text is only what we, as readers, make of it. Jerome McGann extends this argument in his theories of reading as deformance9. In both cases, the text is as much in the reader as it is in the artifact, and what we think of as structure emerges from the interaction of the two. This interaction is evidenced by the emergent text. I begin my investigations of reading as interpretation with a study of clues in Arthur Conan Doyle’s Sherlock Holmes stories. Although ostensibly about something—a crime—Doyle’s fiction is literally about the (re)construction of something—the acts preceeding and constituting the crime. In its temporal inversion between the fabula and the sjuzhet, its incorporation of character-bound narrators and narratees, and its literalization of causal logic through the use of clues, Doyle’s stories seem to perform narrative. Detective fiction has been central to structuralist studies of narrative, and many theorists of detective fiction argue that clues are the critical component of the genre both on a formal and on a socio-historical level10,11. I begin my study of the emergent text by looking at how interpreters structure a narrative around clues. I give readers a copy of Doyle’s “The Red-Headed League” and ask them to mark points in the text at which they identify something as a clue. They use two colors to mark the text, one to indicate something that seems to be a clue when first encountered and another to indicate something that seems to be a clue only after reading further in the text. So readers mark the text the instant they identify some bit of information as a clue to the mystery, and they also have the option of flipping back and finding a detail that seems significant only after reaching some later point in the story. This layering asks crucial questions about the dynamics of reading detective fiction: How does anticipation of a particular type of ending contribute to the identification of textual cues and to the activation of preference rules that govern constructions of narrative? Schematic visual representations of narrative structures are common in articles and books on narratological theory, whether classical or postclassical, and these graphs are most often intended as a visible skeleton of the story itself. Although several postclassical narratologists construct and utilize visualizations of narrative, none attempts to graph narrative as a process, to capture the discovery of textual cues and the enactment of preference rules. In my attempts to do just that, I graph all markings made by readers, not privileging the most recent—most “correct”—structuring. Following this method, I hope that I better capture the dynamic of reading, the constant processing and reprocessing that results from an anticipation of closure, the emergent text.
REFERENCES Herman, David. 1999. “Introduction.” Pp. 1–30 in Narratologies: New Perspectives on Narrative Analysis, edited by David Herman. Columbus, OH: Ohio State University Press. Meister, Jan Christoph. 1995. “Consensus ex machina? consensus qua machina!” Literary and Linguistic Computing 10(4):263–70. Snelgrove, Teresa. 1990. “A Method for the analysis of the structure of narrative texts.” Literary and Linguistic Computing 5(3):221–5. Ronen, Ruth. 1990. “Paradigm Shift in Plot Models: An Outline of the History of Narratology.” Poetics Today 11(4):817–42. Herman, David. 2002. Story Logic: Problems and Possibilities of Narrative. Lincoln, NE: University of Nebraska Press. Jahn, Manfred. 1999. “‘Speak, friend, and enter’: Garden Paths, Artificial Intelligence, and Cognitive Narratology.” Pp. 167–94 in Narratologies: New Perspectives on Narrative Analysis, edited by David Herman. Columbus, OH: Ohio State University Press. Brooks, Peter. 1984. Reading for the Plot. Cambridge: Harvard University Press. Fish, Stanley. 1980. Is There a Text in this Class? Cambridge: Harvard University Press. McGann, Jerome. 2001. Radiant Textuality. New York: Palgrave. Moretti, Franco. 2000. “The Slaughterhouse of Literature.” Modern Language Quarterly 61(1):207–27. Ginzburg, Carlo. 1984. “Clues: Morelli, Freud, and Sherlock Holmes.” The Sign of Three. Edited by Umberto Eco and Thomas Sebeok. Bloomington: University of Indiana Press.