Early data and theories suggested that the properties of rewarding electrical stimulation of the brain (ESB) differ from those of natural rewards, such as food or water. It is contended here that such differences are solely due to differences in deprivation and training conditions, rather than to inherent differences in the rewards. Therefore, it is proposed that ESB has the same properties as other rewards and that its motivational properties are inherent in its reinforcement properties in the form of incentive motivation, rather than in the induction of drive energization. Data testing the model are reviewed and are found to support the incentive position. It is further suggested that to the extent that brain stimulation can maintain behavior in the absence of deprivation the laws for incentive independent of deprivation conditions can be found using brain stimulation.

Rats, previously trained to bar press for electrical stimulation of the brain (ESB), received ESB on an FI schedule for either relatively fast or relatively slow heart rate levels while their skeletal musculature was completely paralyzed by d-tubocurarine chloride. 15 of 19 Ss rewarded for fast heart rates increased their heart rates during a training period, while 15 of 17 Ss rewarded for slow heart rates decreased their heart rates. Rats yoked to experimental Ss in either reinforced group failed to show similar changes. The results suggest that instrumental learning of the heart rate is possible even when the possibility of skeletal muscle mediation is eliminated.

No abstract

Rats were trained to lever press for electrical stimulation of the brain (ESB) and ultimately were assigned to either a fixed interval 1 min. (FI-1 min.) or a variable interval 1 min. (VI-1 min.) schedule of reinforcement. All Ss easily attained and maintained responding on the schedule to which they had been assigned. Patterns of responding during training and extinction were similar to those observed when conventional rewards, such as food or water, are used. Fixed-interval Ss demonstrated scalloped responding; variable-interval Ss demonstrated steady rates of responding. The implications of these results for understanding ESB as a reward are discussed.

No abstract