摘要：When someone is described as memorizing a phone number, part of what is being asserted is that the person is capable of dialing the number without looking it up in the directory. Such responses, which may decrease and stop occurring as training increases, can be interpreted as nonrequired precurrent behavior. In different experiments, participants could look up an auxiliary screen to see the numbers (Experiment 1) or arbitrary characters (Experiment 3) corresponding to different shapes. In Experiment 2, a typing task with a covered keyboard was used, in which participants could look up an auxiliary screen to see key positions. Duration of precurrent response, divided by correct current responses, decreased as a linear function of the logarithm of trials in all three experiments. In Experiment 3, the complexity of the task was changed, by altering the number of responses to be learned per pair, per position, and in the total task. Results indicated that these variables produced systematic effects on performance and are compatible with an interpretation of task complexity based upon the quantification of the programmed contingencies of reinforcement.
