The present experiment studied the process by which environmental events come to exercise some degree of control over the activation of contingent self-reinforcement. Through differential training, animals learned to impose performance requirements for self-reward in certain environmental contexts but not in others. Both contextual influences and periodic negative consequences for noncontingent self-reward increased adherence to response requisites for self-reinforcement.

Monkeys were tested for their relative preference for self-monitored and externally imposed systems of reinforcement. One monkey clearly preferred to self-reward his own performances. whereas a second monkey favored slightly the externally administered system but displayed a small gradual shift toward self-reward as the sessions progressed. Unlike previous findings. unmerited self-reward did not result in rapid discarding of self-imposed performance requirements. The level of work output and the interactive effects of multiple reinforcement control are possible operative factors in sustaining self-imposed contingencies long after punishment for unmerited self-reward has been discontinued.

Electrical responses evoked by clicks, flashes, changes in noise level, and changes in light level were recorded from the scalps of human subjects set to detect one of the stimuli. An early negative component of the evoked responses reflects selection between sensory modalities, whereas the later positive component reflects a more complex intramodal discrimination.