Directed forgetting in rats, to elucidate active control of memory rehearsal processes while controlling for nonmemorial artifacts, was examined using an eight-arm radial maze. To-be-remembered and to-be-forgotten items were presented at different arms in the same trial. A trial consisted of learning and test phases. Rats needed to remember win or loss of a food pellet presented in the middle of the arms, signaling presence or absence of a large reward there in the subsequent test phase. Two other qualitatively different foods placed at the end of the arms served as remember (R) or forget (F) cues, signaling whether those arms would be presented in the test phase. Compared with the normal test, rats' performance deteriorated significantly if the arms previously marked by F-cues in the preceding learning phase were actually used in the test phase, showing reliable directed forgetting in rats. Rats were also tested in a condition in which F-cues were not presented at all, and thus rats had to remember all the arms. Although positive evidence of reduction of memory load in working memory by utilizing F-cues was not demonstrated, analysis of individual data suggested that utilization of R-cues and F-cues interfered with the main task of remembering win/lose information.
Keywords: Active memory control; Directed forgetting; Memory load; Radial maze; Rats; Working memory.