The types of agency in fictional literature are familiar in their general actions. The first is intradiegetic, concerning characters or other agents within the diegesis (the fictional world) whose actions bring about other events and outcomes. The details of intradiegetic agency—involving physical actions, character motivations, and the chain of cause and effect—are the stuff of much that passes for critical analysis in literature courses, practical criticism, and discussions and interviews with authors (how this event occurred, why this character acted in this way, and so on). However, as important as this type of agency can be in discussing a work, my main concern here is with the next two levels of agency.

The second level, also familiar enough, is intratextual, most often thought of as a narrator. This is an agency within the text but not necessarily within the diegesis that presents the narrative itself and its intradiegetic agents, as well as some material that is not part of the narrative itself but related to it (for instance, epigrams). This agent, the narrator, has arranged (or at least seems to have arranged) the details of the narration, placed the cues and initiated the process of the interaction of syuzhet/plot and style. Narrators, according to the classifications given by Gerard Genette, can most simply be typed as "heterodiegetic" (parallel with what is commonly referred to as "third-person" narration, "with the narrator absent from the story he tells") and "homodiegetic" (parallel with "first-person" narration, "with the narrator present as a character in the story he tells") (Narrative Discourse 244-245).

However, it is not clear that the agency of narration must always be a personified "teller." Even with the medium of print fiction, there is some disagreement among critics about whether, for instance, a story consisting only of dialogue still has a "narrator" in any meaningful sense of the word. What role a "narrator" might have in placing other, nondiegetic elements within the text (title pages, chapter headings, etc.) is also somewhat ambiguous. For example, Manuel Puig’s Kiss of the Spider Woman consists only of dialogue, fragments of documents, and some footnotes. Does it really have a "narrator" as such? When we turn to other media, such as film, which use elements other than the printed or spoken word to convey stories, the status of an overall "narrator" for the story is even less clear.

So far, we have been dealing with the issue of agency within the (print fiction) text, which presents enough problems of complexity as it is [3]. But what about the third level, extratextual agency, the agent that has created—or seems to have created--the text itself? If characters in stories cause actions to occur, and if those stories are told by the characters themselves or by a heterodiegetic narrator, then who—or what—has created the narrator and the overall narration of the text? The simple, common-sense answer (in print fiction at least) is the "author," the historical person who wrote the book or story. Again, though, within written fiction alone that common-sense answer is not as clear as it seems to be. What, for example, is an "author" when the historical person (say, the author of Beowulf) is unknown? Who gets the credit when more than one author is involved (for example, The Inheritors, by Joseph Conrad and Ford Madox Ford or the science fiction stories of "Lewis Padgett," actually the husband-and-wife team of Henry Kuttner and C.L. Moore)? And what role, if any, does that author play in a text once it has passed beyond his or her or their direct control? Is it the author or the narrator, for example, who should be understood as the creator of non-diegetic elements in the text?

Genette himself would deny the relevance of such questions to the analysis of narrative proper. To identify an author with a narrator, Genette writes, "is perhaps legitimate in the case of a historical narrative or a real autobiography, but not when we are dealing with a narrative of fiction." Furthermore, we must not confuse the narrating activity or the views of a narrator with those of the book’s author, however similar they might be: " . . . the narrating situation of a fictional account is never reduced to its situation of writing" [ND 213-214, Genette’s emphasis]. Narrowly applied, Genette’s admonitions make sense; they are similar to other warnings against the "intentional fallacy" that would boil a work’s essence down to pronouncements or facts about the work’s creator while saying little about the work itself. To ask about agency, though, is to have a rather different aim: not necessarily to seek "meaning" or understanding through appeal to an author’s biography or other works but to ask an even more basic question: Is there intradiegetic or intratextual evidence of extratextual agency? In other words, is there evidence within the diegesis or the text (aside from title pages, prefaces, acknowledgments and so on) of how this text came to be? How, if at all, does the text acknowledge (or pretend to acknowledge) that agency and why? In addition, I must wonder how my own knowledge or expectations of authorship affect my experiential awareness of the narrative and its diegesis. Finally, I can ask if the text acknowledges readers’ knowledge and expectations about authorship--their contextual awareness—and how such acknowledgement, in turn, prompts certain kinds of readings and reactions.

return to top