Effects of Punishment as Positive Reinforcement for the Punisher 1

Abstract

The hypothesis that punishment provides a positive reinforcement for the user was tested, by making the opportunity to punish contingent upon staying on target on the pursuit rotor during learning. Groups given the opportunity to "punish" another S's mistakes did not spend more time on target than did control groups; therefore the hypothesis that punishing had acquired secondary reinforcing properties prior to the experiment was not supported. There was some evidence that switching from receiving to administering shock produced an initial performance increment, but this effect was short-lasting. There were no effects of prior shock, or shock administration, upon subsequent performance under normal conditions.