Abstract
Six college undergraduates received programmed concept training on three kinds of intraverbal relations. These relations involved definition, exemplification, and example identification questions. The experimenter presented the questions, the subject answered them in writing, and the experimenter provided specific corrective consequences. After completing the training on a concept, the subject immediately received a test on the concept. The test included novel questions similar to the kind used in training (extension tasks) and question types that were not used in training but which were also considered intraverbal relations (transfer tasks). Training results indicated rapid, errorful responding on example identification tasks and slow, accurate responding on exemplification and definition tasks. Test results indicated rapid, errorful responding on example identification extension tasks; slow, accurate responding on exemplification extension tasks; and slow, errorful responding on definition extension tasks. In testing, differential responding occurred on transfer tasks as a function of the kind of intraverbal training received, and substantially lower levels of performance were obtained on transfer tasks than on extension tasks. It appears that the intraverbal can be subdivided into more specific categories of operants.