Two experiments investigated factors influencing the integration of sentences. A priming technique using sentence recognition was employed. The experimental procedure involved presenting four to-be-remembered sentences to the subjects and then testing two sentences successively for recognition. In the experiment 1, positive priming effect was observed when two test sentences shared same subject noun (same subject condition), whereas negative priming effect was obtained in the same verb condition. Results of experiment 2 showed no significant priming effect in the same object condition. These results suggest that two sentences are integrated when they share same subject noun. A tentative model based on repetition rule and spreading activation explanation was briefly discussed.
The purpose of this study was to investigate hierarchical structure of concepts which was the basic organization of human knowledge. In order to compare two different types of hierarchy (concrete-abstract hierarchy and part-whole hierarchy), 40 subjects were asked to list kinds of 12 concepts and the other 40 subjects were asked to list parts of those. The concepts were divided into biological categories, object categories and scene categories. Results revealed that while more kinds were listed for the biological categories, more parts were listed for the scene categories. The conclusion was that concepts were organized into the two types of hierarchy, and that the two structures were different in salience among the categories. It was suggested that the difference was due to the relationship between superordinate- and subordinate-concepts and to the internal structures among subordinate-concepts.
The effects of inaccurate instructions on the differential-reinforcement-of-low-rate (DRL) performance of human subjects were investigated. The inaccurate instructions were divided into two types, i.e., the instruction which leads subjects to contact aversive consequence and the instruction which leads them to contact no clear aversive consequence. In this study, those two types of instructions were presented within a session on three-component multiple DRL schedules. Results indicated that the instruction resulting in aversive consequence caused subjects to respond according to schedule contingencies more quickly than the instruction resulting in no aversive consequence. The usefulness of the inaccurate instruction resulting in aversive consequence to eliminate the instruction-following behavior was emphasized.