To investigate the hypothesis that cognitive rules govern writing behavior, Carter's signaled stopping technique was used to study language and cognitive effects in public relations messages. Principles from Grunig, et al (1985) Axiomatic Theory of Cognition and Writing, which proposed premises, axioms and definitions about writing, were applied to one group of 30 college students and another group of 29 students to test for important differences in signaled stops for two versions of a story about alcohol abuse. Control groups read a story without figures of speech, and experimental groups read a story worded exactly the same, but with analogies and examples. Stops for the following reasons were coded: (1) confusion about the text, (2) need to reread, (3) need to ask a question, (4) reflection about the text, (5) sudden insight or understanding, (6) to agree, and (7) to disagree. Results revealed that signaled stops for thinking and for understanding were significantly greater for the experimental groups, whose text contained figures of speech. The figures of speech caused few confusion stops, however, and there was no clear pattern for questioning stops. The findings suggested that, while the signaled stopping technique requires refining, it is a useful new way for testing messages for understanding and meaning. (SKC)