Memory, serial anticipation pattern learning, and transfer in rats

Abstract

Three experiments tested hypotheses about whether rats respond appropriately to, or track, an orderly series of reward magnitudes terminating in nonreward by encoding the rule structure of the series, as a recent view derived from human serial-pattern learning models suggests, or whether they employ the memory of one reward event as a cue to signal the next reward event in the series, as animal partial-reinforcement investigations suggest. The memory signaling the terminal nonreward was called nonreward memory. The capacity of the nonreward memory to also signal reward was varied indirectly in Experiments 1 and 2, through stimulus generalization from other memories that were rewarded, and directly in Experiment 3, by rewarding it in an initial preshift phase. Consistent with a memory view, tracking of the terminal non-reward event decreased as the nonreward memory increasingly also signaled reward. These findings indicate that serial-pattern learning in rats has much in common with other, better investigated forms of instrumental learning, such as partial reward, and is not a unique form of learning, as suggested by the rule-encoding view, which was strongly disconfirmed.