Files in this item
|(no description provided)|
|Title:||Context Effects on the Representation of Meaning|
|Author(s):||Roth, Emilie Matarasso|
|Department / Program:||Psychology|
|Degree Granting Institution:||University of Illinois at Urbana-Champaign|
|Abstract:||Three experiments were conducted to examine the effect of context on the representation accessed for category terms. The first experiment employed an on-line reading time paradigm to establish that context affects the representation accessed for a category term at the time of comprehension.
Experiment 2 and 3 investigated the properties of the representation accessed for category terms in the presence of context. Experiment 2a was a rating study. Representativeness ratings were obtained to determine how well different exemplars fit people's idea of a category term presented in a context sentence. The results indicated: (1) that context-dependent category representations generate graded representativeness distributions of exemplars; and (2) that these representativeness distributions can not be derived by assuming that the exemplar most strongly suggested by the context sentence serves as the category representation. Experiment 2b used a reading time paradigm to show that the representativeness ratings obtained in Experiment 2a predict implicit category verification time.
Experiment 3 examined the process by which membership in a context constrained category is established. Subjects were timed as they decided whether an exemplar could be a possible referent of a category term presented in a context sentence. Two variables were manipulated: (1) whether the exemplar was a possible referent (true vs. false) and (2) whether it was judged (by an independent group of subjects) to be related to the representation of the category term in that context (related vs. unrelated). Decision time was significantly faster for related trues than for unrelated trues. For false items, decision time was significantly faster when the item was unrelated. The results indicate that membership in context-constrained categories is not determined through an exemplar search process. No effect of typicality (as determined in the absence of explicit context) was found.
The conclusions drawn from the results of the studies were: (1) Context-dependent category representations are similar in structure to the category representations studied in the absence of explicit context. (2) The entire representativeness distribution generated for category terms in the presence of context is determined by the constraints imposed by the context. (3) The typicality structure observed in the absence of explicit context does not influence categorization processes when context is introduced that suggests a subset of exemplars with different characteristic properties. Possible ways current models of semantic memory could deal with the results are discussed.
Thesis (Ph.D.)--University of Illinois at Urbana-Champaign, 1980.
|Date Available in IDEALS:||2014-12-13|