TY - JOUR
T1 - Prominent characters and events organize narrative understanding
AU - Morrow, Daniel G.
N1 - Funding Information:
Experiment I is part of the author’s dissertation completed at the University of California at Berkeley. where the author was supported by predoctoral Grant 5 T32 MH 15860-03 from the National Institute of Mental Health. I thank Eleanor Rosch for serving as the chairperson of my committee and for her advice on Experiment I. Experiments 2-4 were performed at Stanford University, where the author was supported by postdoctral Grant 2 T32 MH I5 157-06 from the National Institute of Mental Health. I thank Ellen P. Francik. Stephen Greenspan. Gisela Redeker, Heather Stark, and Deanna L. Wilkes-Gibbs for their comments on the research and/or on the manuscript. 1 am also very grateful to Herbert H. Clark for his advice on the research and several drafts of the paper, and to Ruth Kimchi for many comments on the manuscript.
PY - 1985/6
Y1 - 1985/6
N2 - Readers understand a narrative by constructing a representation from a sequence of sentences. They must identify which parts of the narrative are most prominent in order to assign the appropriate referents to referring expressions and construct a coherent representation. The present experiments demonstrate that properties of the characters and events expressed by narratives are often more important for guiding referent assignment than the order in which these parts of narratives are mentioned. In particular, in congruent narratives, where the protagonist (i.e., the most important character of the narrative) participates in a foreground event (i.e., part of the narrative plot) and the nonprotagonist participates in a background event, readers chose the protagonist as referent for a subsequent pronoun, regardless of the order of mention of the events the two characters participated in. However, in incongruent narratives, where the protagonist participates in the background event and the nonprotagonist in the foreground event, readers were less confident and relied on the order of mention and status of events, choosing the nonprotagonist in a last mentioned foreground event and favoring neither character when the foreground event with nonprotagonist was mentioned first. When narratives more sharply delineated foregrounded from backgrounded events, readers were less confused and chose characters from foregrounded events. When narratives contained pronouns that referred to places rather than characters, readers chose referents from the last mentioned place regardless of the kind of character or event it was associated with. The results suggest that readers combine information about the characters, events, and places that narratives express with information about their order of mention in order to assign referents and that this process is part of the process of constructing a model that represents the world described by the narrative.
AB - Readers understand a narrative by constructing a representation from a sequence of sentences. They must identify which parts of the narrative are most prominent in order to assign the appropriate referents to referring expressions and construct a coherent representation. The present experiments demonstrate that properties of the characters and events expressed by narratives are often more important for guiding referent assignment than the order in which these parts of narratives are mentioned. In particular, in congruent narratives, where the protagonist (i.e., the most important character of the narrative) participates in a foreground event (i.e., part of the narrative plot) and the nonprotagonist participates in a background event, readers chose the protagonist as referent for a subsequent pronoun, regardless of the order of mention of the events the two characters participated in. However, in incongruent narratives, where the protagonist participates in the background event and the nonprotagonist in the foreground event, readers were less confident and relied on the order of mention and status of events, choosing the nonprotagonist in a last mentioned foreground event and favoring neither character when the foreground event with nonprotagonist was mentioned first. When narratives more sharply delineated foregrounded from backgrounded events, readers were less confused and chose characters from foregrounded events. When narratives contained pronouns that referred to places rather than characters, readers chose referents from the last mentioned place regardless of the kind of character or event it was associated with. The results suggest that readers combine information about the characters, events, and places that narratives express with information about their order of mention in order to assign referents and that this process is part of the process of constructing a model that represents the world described by the narrative.
UR - http://www.scopus.com/inward/record.url?scp=0002089166&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=0002089166&partnerID=8YFLogxK
U2 - 10.1016/0749-596X(85)90030-0
DO - 10.1016/0749-596X(85)90030-0
M3 - Article
AN - SCOPUS:0002089166
SN - 0749-596X
VL - 24
SP - 304
EP - 319
JO - Journal of Memory and Language
JF - Journal of Memory and Language
IS - 3
ER -