The effect of electrical shock and warning on cooperation in a non-zero-sum game

It has been generally accepted that if both players in a two-person have a means of retaliation they will use it to their mutual detriment (Deutsch and Krauss, 1960; Komorita, Shepesh, and Braver, 1968). The present authors suggest that the loss of points incurred by the use of the gate (Deutsch and Krauss, 1960) and the point reduction used as threat (Komorita, Shepesh, and Braver, 1968) have important differences from the real life situation to which they are usually generalized. It is believed that point reduction makes the more complex and puts the emphasis on competition (Kelly, 1965). In real life punishment is not in the same mode or currency as the game being played. This experiment therefore tests the effect of punishment outside the point structure (electric shock) on the amount of cooperation between players. In an attempt to make the results even more generalizable to real life situations, the effect of being able to warn the other individual of one's readiness to punish is also studied. Two rival hypotheses are possible concerning the effect of warning on the number of electrical shocks given. One hypothesis, coming from animal social behavior suggests that a warning signal serves the function of reducing the amount of actual physical combat occurring within a group. Failure to learn these warning signals has been repeatedly used as an explanation for the increased aggression shown in the isolate-reared and the zoo-reared animals. An alternate hypothesis which comes from the human literature suggest that repeated warnings will cause the subject to actually go through with the threatened electrical shock as a means of saving face by remaining true to his word. To investigate these questions, we used a version of the Prisoner's Dilemma game.