Cognitive Narratology and the Art of Character

One of fiction’s strangest achievements is its ability to persuade us that a consciousness exists where there is really only language. A character may consist of scattered gestures, fragments of dialogue, habits of attention, and a few remembered scenes. Yet the reader begins to imagine a whole person: someone with motives, blind spots, desires, fears, and private histories that the novel may never fully describe.

Cognitive narratology studies, among other things, how this process works. Drawing on narratology, psychology, linguistics, and cognitive science, it examines the mental activity involved in reading fiction. How do readers construct fictional worlds? How do we infer what characters know, believe, want, or misunderstand? Why can a character feel psychologically complete even when the writer tells us relatively little about them?

For writers, one of the most useful conclusions is that characterization is collaborative. The author supplies evidence. The reader builds the person.

Readers perform this work almost automatically. Imagine a character who receives a letter, turns it over several times without opening it, puts it beneath a stack of newspapers, and then spends the afternoon repairing a broken chair. The writer has not told us what the character feels. But readers immediately begin making inferences. Fear, avoidance, guilt, grief, embarrassment: possibilities accumulate around the action.

Fiction depends on our tendency to infer mental states from incomplete information. Cognitive theorists sometimes describe this process in relation to “mindreading,” the everyday ability to guess what another person might be thinking from what they say, do, notice, conceal, or fail to notice. Literary characters invite the same activity.

Natalia Ginzburg makes remarkable use of this principle in Family Lexicon. Members of the family emerge through recurring phrases, habits, irritations, and private jokes rather than through sustained psychological explanation. The father’s temperament, for example, becomes inseparable from the peculiar language he uses to judge the people around him. Character accumulates through repetition. We gradually know these people because we learn their patterns.

Penelope Fitzgerald often works with even less. In The Blue Flower, characters frequently reveal themselves through small practical choices, abrupt exchanges, and their responses to social situations. Fitzgerald rarely interrupts to tell us exactly what every gesture means. The reader has to connect fragments. A personality develops through inference.

This can be useful for writers who are tempted to explain too much. Suppose a novelist wants us to understand that a character feels inferior around his older brother. The manuscript could simply state this. Or it might show him correcting his grammar before the brother arrives, laughing too quickly at a joke, exaggerating the price of his apartment, and later remembering an apparently insignificant childhood competition. Each detail adds something to the reader’s mental model. No single detail needs to announce its meaning.

Jean Rhys takes this further in Good Morning, Midnight. Sasha’s consciousness is fractured by shame, memory, and social anxiety, but Rhys does not organize these elements into a tidy diagnosis. We experience Sasha through repetitions and evasions: the rooms she chooses, the clothes she worries about, the encounters she anticipates and dreads. Her mind becomes vivid partly because it remains difficult to summarize. That difficulty matters. A convincing character is rarely one whose psychology can be reduced to a single explanation.

Edward P. Jones’s stories offer another useful model. In collections such as Lost in the City, characters are frequently introduced through concrete circumstances rather than an explanatory backstory. A seemingly ordinary decision can reveal pride, resentment, tenderness, or self-deception. Jones trusts readers to construct the emotional history surrounding an action.

This idea has important consequences for revision. Writers usually know far more about their characters than readers ever will. They may know what happened when a character was seven, why she refuses to speak to her sister, what kind of student she was, and precisely what she fears will happen if her marriage ends. That knowledge can help the writer produce the character, but it does not all need to appear in the book.

The question is which evidence the reader actually needs. The author cannot fully reproduce the experience of encountering the manuscript without prior knowledge. A book coach can. The coach sees only what the pages provide and can therefore notice discrepancies between the character the writer intends and the character the reader is likely to construct.

Perhaps the author imagines a protagonist as cautious, while the manuscript makes her seem passive. A father is supposed to be loving but emotionally clumsy, yet his scenes contain only criticism. A later act of betrayal is intended to complicate our understanding of a character, but nothing earlier in the book prepares us to recognize that possibility.

The solution is not always more explanation. A coach might instead suggest changing what the character notices in a scene, introducing a contradictory gesture, adjusting a line of dialogue, or allowing an earlier action to acquire new significance later. Just as importantly, a coach can identify passages where psychological commentary is doing work the scene has already accomplished. Removing explanation can sometimes make a character feel larger.

Cognitive narratology helps clarify why. Readers do not need a complete inventory of a fictional mind. They need enough evidence to begin constructing one themselves. The strongest characters often seem to extend beyond the boundaries of the page precisely because some part of them remains unwritten.