A cognitive-psychology framework for the structural regularities of narrative text. The dominant formulation, Stein and Glenn (1979), models a story as a setting plus one or more episodes, each consisting of an initiating event, an internal response, an attempt, a consequence, and a reaction. Mandler and Johnson (1977) and Rumelhart (1975) proposed earlier variants; the Stein and Glenn rules became the most-cited reference set in research on children's narrative comprehension and production.
Within Stein and Glenn's framework a complete episode contains:
A story can be a single episode or a chain of episodes, with each episode's consequence sometimes feeding the next episode's initiating event. Mature readers parse narrative against this template implicitly, which is why narrative passages place a lower processing load than expository text on most readers.
Narrative familiarity is itself a schema. Readers recognise the story-grammar template from oral tradition and early literacy and use it to predict what comes next, anchor characters across long-distance anaphora, and integrate inferences. This is what Coh-Metrix's narrativity dimension captures and why narrative passages score easier on cohesion-aware readability tools than expository texts of equal lexical difficulty.
For reading assessment, story grammar has two implications. First, narrative passages cannot be assumed to discriminate at upper proficiency: the schema scaffolds candidates' comprehension and compresses the distribution of scores. Tests aiming to separate B2 from C1 readers should not rely on narrative alone. Second, for early-grade and lower-proficiency assessment, narrative is fairer than expository because the genre's familiarity reduces construct-irrelevant variance from rhetorical-structure unfamiliarity.