Detective fiction, particularly of the classical formula, seems to be unique among narrative genres in that it thematizes narrativity itself as a problem, a procedure, and an achievement. In fact, its very constitution as a genre is based on the complicated employment of certain narrative strategies: the point of a classical detective novel typically consists in reconstructing a hidden or lost story (that is, the crime); and the process of reconstruction (that is, the detection), in its turn, is also usually hidden in essential respects from the reader. Furthermore, as I shall attempt to show in detail, narrativity is inscribed in detective fiction in yet two other forms. The stories that are narrated in detective novels can profitably be described as stories of writing and reading insofar as they are concerned with authoring and deciphering "plots." And ultimately, the detective novel hinges on the social effects that the concealment and the "publication" of a particular story, namely that of the crime, will have. Thus detective fiction affords clear examples of the possible social dimensions of narrativity, and the history of the detective genre is marked, on one level, by growing doubts about the possibility of telling the story. [End Page 451]
The plot of the classical detective novel comprises two basically separate stories—the story of the crime (which consists of action) and the story of the investigation (which is concerned with knowledge).1 In their narrative presentation, however, the two stories are intertwined. The first story (the crime) happened in the past and is—insofar as it is hidden—absent from the present; the second story (the investigation) happens in the present and consists of uncovering the first story. On the concrete level of plot and setting, the link between the two stories is established in the following stages. Most classical detective novels start out with a community in a state of stable order. Soon a crime (usually a murder) occurs, which the police are unable to clear up. The insoluble crime acts as a destabilizing event, because the system of norms and rules regulating life in the community has proved powerless in one crucial instance and is therefore discredited. In other words, the narrative incapability on the part of society's official agents, their inability to discover and tell the story of the crime, thus threatens the validity of the established order. At this point the detective takes over the case, embarks on a course of thorough investigations, and finally identifies the criminal, explaining his solution at length. Thus, through the development of the second story, the absent first story is at last reconstructed in detail and made known. By reintegrating the aberrant event, the narrative reconstruction restores the disrupted social order and reaffirms the validity of the system of norms.
Conventionally, the coupling of these two stories is presented to the reader in a specifically involved manner (invented by Poe and standardized by Doyle). Employing Gérard Genette's and Seymour Chatman's distinction between story and discourse, one can define the narrative organization of a classical detective novel as follows.2 The usual constellation of story and discourse (the abstractable preexistent sequence of events and acts versus its mediation in a narrative) occurs twice over: the story of the crime is mediated in the discourse of the detective's investigation; and the story of the detective's investigation, in its turn, is mediated in the narrator's discourse (for instance in Dr. Watson's uninformed written account of Holmes's detection). In both cases the story is hidden for the most part so that the reader is doubly puzzled—trying to make out the mysterious crime story by way of the almost equally mystifying detection story. [End Page 452]
In spite of writers' preoccupation with surprising variation, the arrangement of narrative elements is fundamentally the same in the majority of classical detective novels.3 In fact, the genre has been remarkable for the rigidity of its conventional structures—especially during the 1920s and '30s, when the conventions were frequently canonized as explicit "rules" (for instance, by S. S. Van Dine or...