Examines cohesion in the introductions to some scientific articles and compares the patterns to those from popularizations. Discusses a computational model of cohesion. Argues that readers of scientific articles must have a knowledge of lexical relations to see the implicit cohesion, whereas readers of popularizations must see the cohesive…

Studies the construction of coherent narrations of events in a picture storybook by children and adults. Applies a causal network discourse analysis. Analyzes the structure and content of the network representation of the main character. Reveals differences in content and structure of the narratives. (HB)

Analyzes children's verbal disputes for the development of discourse markers "because,""so,""and,""but," and "well." Shows that the markers were primarily used to mark the exchange structure of discourse when they were first used by children two years seven months to three years six months and that the first markers to do so were "but" and…

Describes how interactive hand gestures made during conversation help maintain the conversation as a social system. Relates the methods and results of three experiments designed to test this theory. Finds that interactive gestures maintain involvement with the interlocutor without interrupting the flow of conversation. (HB)

Reports the findings of a study designed to investigate children's deictic changes, use of speech act verbs, and preference for reporting system in their retold narratives. Claims that a linear age function emerged and that children's mastery of direct and indirect speech extends over many years. (HB)

Analyzes the referential strategies used in narrative discourse by 10 adults and 60 Japanese children aged 3 to 7 years. Determines the factors underlying choice of nominal versus elliptical forms. Discusses results in terms of cognitive, social, and linguistic factors underlying referential choice. (HB)

Uses eight languages in five distinct linguistic areas to examine two hypotheses regarding text generation and analysis and to illustrate their reciprocity relative to narrative discourse. Demonstrates how these hypotheses yield salience schemes and constituent analysis which mutually corroborate and correct each other. (KEH)