Psychology ETDs

Publication Date



Changes in the location, duration, and frequency of pauses in spontaneous speech were studied as a function of two levels of cognitive complexity, two of situational stress, and three levels of a dimension of the subjects' cognitive style (scanning). Since pauses reflect the location in a sentence where cognitive activity takes place, their variation in numbers and length allow inferences about underlying speech production processes. Subjects were classified into three groups based on the extensiveness of perceptual scanning, a personality characteristic hypothesized to be related to speech production processes. Scanning style was evaluated by means of factor scores derived from a principal component analysis of fifteen variables derived from four perceptual-cognitive tasks (n = 43). The eighteen speech subjects performed the stress by cognitive task complexity within-subjects portion of the experiment. They talked about neutral and stressful cartoons (stressful content manipulation) on simple descriptive and interpretive levels (task complexity). The average duration of a subject's pauses and their frequency of occurrence were tabulated separately for pauses occurring at the beginning of a clause (early) or within clauses (late}. Early pauses were found to increase in duration with stress. They also increased with task complexity, but only for the two higher scanning groups. The complexity by style interaction approached statistical significance. The number of early pauses did not change with any independent variable. Late pauses decreased in both number and frequency with increased complexity. The duration decrease again depended on style and on stress condition as well. All style groups decreased pause length on the more complex task (interpretation) when content was neutral. Only low scanners showed this decline in the stressful condition resulting in a possibly reliable three-way interaction. The results allowed inferences about the functions of pauses and the processes which underly speech production. Evidence for planning, lexical choice, and review functions of pauses is presented. A model of speech production based on slips-of-the-tongue (Fromkin, 1971) is described and expanded in light of the present findings.

Degree Name


Level of Degree


Department Name


First Committee Member (Chair)

Carol Elizabeth Conrad

Second Committee Member

Samuel Roll

Third Committee Member

Richard Jerome Harris



Document Type


Included in

Psychology Commons