AFTER BEING PAIRED WITH ANOTHER SUBJECT, EACH SUBJECT IN THE EXPERIMENT WAS EITHER PROVOKED OR NOT PROVOKED BY A CONFEDERATE'S ELECTRIC SHOCK EVALUATION. THE SUBJECTS THEN OBSERVED A FLASHING WHITE LIGHT THAT WAS ASSOCIATED WITH EITHER THEIR FORMER EVALUATOR'S PAIN OR A NEUTRAL EVENT. NEXT, THE SUBJECTS ADMINISTERED ELECTRIC SHOCKS TO A NEW VICTIM, WITH WHOM THEY HAD NOT INTERACTED, AT THE FLASH OF BOTH THE FAMILIAR WHITE LIGHT (THE CONDITIONED STIMULUS) AND A NOVEL BLUE LIGHT. POSTSESSION QUESTIONNAIRES PROVIDED EVIDENCE OF PROVOCATION. FINDINGS SHOW THAT PROVOKED SUBJECTS ADMINISTERED MORE INTENSE SHOCKS TO CONDITIONED STIMULUS WHEN IT HAD BEEN ASSOCIATED WITH THEIR EVALUATOR'S PAIN. UNPROVOKED SUBJECTS ADMINISTERED LESS INTENSE SHOCKS TO THE LIGHT THAT HAD BEEN ASSOCIATED WITH THEIR EVALUATOR'S PAIN. THE RESULTS INDICATE THAT UNDER CERTAIN CONDITIONS, THE SUFFERING EXHIBITED BY THE VICTIM OF AGGRESSION IS REINFORCING FOR ANOTHER INDIVIDUAL, EVEN THOUGH THAT PERSON DOES NOT INFLICT THE PAIN. STIMULUS ASSOCIATED WITH SUCH EVENTS MAY ACQUIRE THE ABILITY TO MAINTAIN THE OBSERVER'S AGGRESSION UNTIL A LATER TIME WHEN HE IS PUNISHING SOMEONE ELSE, EVEN THOUGH THE AGGRESSOR IS NOT ANGRY WITH THE NEW VICTIM. TABLES ILLUSTRATE THE DATA, AND REFERENCES ARE PROVIDED. (AUTHOR ABSTRACT MODIFIED--KJM)