Concurrent schedule control of monkey's observing during vigilance.

Experiment I exposed three crab-eating macaques to a vigilance task which required them to detect signals by holding down a lever and to report their occurrences by releasing the lever. Only those reporting responses which occurred within 1 s after onset of each signal were reinforced and letting go of the lever in the absence of the signals was punished by a blackout. Signals were presented according to concurrent variable-interval schedules and subjects could change over between alternatives by pressing another lever. All subjects showed frequent changeovers and time allocation of observing responses showed under-matching. matching. Experiment II exposed the subjects to a choice task which required continuous lever holding. Although all subjects showed undermatching, one of the subjects alternated less rapidly than in Experiment I. Experiment III exposed another monkey to a typical choice task in which a lever-pressing response was reinforced by a primary reinforcer and a vigilance task used in Experiment I. The subject alternated less rapidly and time allocation roughly mathced the reinforcement ratio in the choice task, whereas time allocation showed undermatching in the vigilance task. These findings suggested that performances under vigilance situations might differ from those under typical lever-pressing situations.