Three experimental tasks-spontaneous telling of a story, reading, and parsing the story-were used to determine whether empirical data reflect the narrative structure of stories and can be predicted by a plot unit analysis of the stories (Lehnert, 1981). It was found that spontaneous pause durations at sentence breaks were highly correlated with the importance of these breaks as predicted theoretically. Only low correlations were obtained, however, when reading pause durations were correlated with the model. As for parsing values, the value of the correlation coefficients depended on whether stories had sufficient superficial linguistic cues to help the subjects in parsing. It was concluded that spontaneous pausing not only reflects the narrative structure of stories, but can be used as a guide to constructing theories of narrative structure as well as for deciding between competing theories. © 1984.
Gee, J. P., & Grosjean, F. (1984). Empirical evidence for narrative structure. Cognitive Science, 8(1), 59–85. https://doi.org/10.1016/S0364-0213(84)80025-7