| Abstract: | Eighteen children, randomly assigned to one of six conditions, were trained in the use of two-word subject-verb (e.g. “mommy go”) utterances. The groups differed in (1) the number of different semantic relations theoretically underlying the subject-verb utterances on which the children were trained, and (2) the extent to which these semantic relations were associated with nonlinguistic events. The results indicated that the children acquired the subject-verb utterance forms more readily when they were also exposed to the events to which the utterances actually referred. Provided that the subject-verb utterances were related to ongoing events, the number of relations “underlying” these utterances did not appear critical. These results suggest that the child's acquisition of novel and grammatical utterances describing nonlinguistic stimuli highly similar to those in his natural environment are amenable to experimental control. |
