Character Studies

I readily agree with James Phelan that Chatman's model of narrative should be expanded. All models smooth the data for the sake of argument. Models must decontextualize, and while designed to show variation, they prefer "constants" to "variables," to borrow Phelan's terms.¹ This is as true of representations of persons as it is of models. Phelan's model of authors, resources, and audiences accommodates variables so much that it may not really be a model—until you return to an individual text that it models. The extensible list of resources, "the etc. at the end of the middle column" (8), allows for description or what in other contexts has been called surface reading focused on a particular textual production (Best and Marcus). Such contextual particularity is a sure sign that Phelan is indeed what he professes, a rhetorical narrative theorist—an approach almost naturally congenial to me. I particularly appreciate his hospitality in inviting "characters [to] … be part of Chatman's model" (5).² I will focus on that resource, turning a light on its implications for nonfiction.³ I propose that character typology be considered a resource comparable to genre. Characterization relies upon the interplay between individuals and types, variables, and constants. Readers fill in a narrative's gaps in the indications of personality according to different roles and prototypical functions, from the "round" protagonist to the "flat" member of the chorus. In biography as well as in realist fiction, agents are read as referring to both literary and social conventions. Thus, fictional characterization is one member of a large species of narrative representation of persons, which relies on typological modeling to communicate with its audiences.

In Phelan's model, "audiences have agency" in responding to a narrative's resources (10), as when they register the paratext and other indications of a specific genre. Readers in a sense choose which preloaded apps to use according to marks of authorial source and cues such as footnotes, specific day-month-year birth dates, or absence of fictitious place names.⁴ Similarly, typologies of persons preload from all sorts of social encoding, including genres beyond printed realist fiction.⁵

Character has had a revival as research topic, though of course no one speaks about any work of fiction for long without some mention of the agents of the events. Characters capture readers' attention, and teachers try to school novices to see beyond mimetic and thematic functions of character to look at other resources.⁶ In the realm of life narrative, it is absurd to deflect attention from the agents, the subject(s) being the very pretext of an autobiography or biography. Yet this is not to say we read nonfiction only mimetically. We should avoid solely referential readings of the subjects of life narrative as much as of fictional characters. Hayden White, we recall, agitated the world of historiography by calling attention to the narrative genres that shape even the most exhaustively documented histories. Critics, similarly, should do more to dispel the notion of clear formal borders between factual and fictional narration.⁷ A factual rendition of a real person is as mediated and almost as constructed as a fictional character is by authorial and audience agency and typological expectations. The biggest difference lies in documented events. Biographers may overlay certain plots such as tragedy or triumph over adversity, or may embellish childhood experiences, but they break their genre's contract if they contradict well-known evidence of events. Omit many events, by all means invent focalization or dialogue, tell it out of chronological order, but do not attribute the deeds of Caroline Herschel to Florence Nightingale, who never discovered a comet; do not say that Nightingale died in 1856, a martyr to her service to British soldiers in the Crimean War.⁸

I pursue the difference of nonfiction from fiction somewhat further, not because I need to argue that they share common techniques; this is widely accepted. Rather, I would invite further discussion of what I call strategic typologies, in the spirit of Phelan's contributions to narrative ethics. Typologies like gender can be deployed strategically, not...