Abstract

It is well known that in two-way active avoidance learning, animals that can avoid footshock successfully after extensive training show reduced fear or no fear at all to the stimulus (i.e., CS) used in the training. However, recent studies reported that animals with extensive avoidance training still show significant fear to the CS. The present study examined whether this discrepancy is due to context discrimination. That is, if extensively trained animals can discriminate between fear-testing and avoidance learning situations, then they would know that no shocks follow the CS in the testing situation, and consequently, show reduced fear to the CS. Flats were divided into 2 groups: A preexposure group that received 24 presentations of CS alone in fear-testing box before avoidance learning, and a non-preexposure group that received avoidance training without preexposure to the testing situation. Both groups received 15 avoidance training trials on day 1, and were trained to the criterion of 27 consecutive avoidance responses on day 2. The fear-testing results indicate that while the non-preexposure group had normal level of conditioned fear to the CS, the preexposure group showed reduced fear to the CS. These findings support the context discrimination hypothesis and suggest that extensive avoidance training does not necessarily reduce fear to the CS.