Delayed matching in pigeons: can apparent memory loss be attributed to the delay of reinforcement of sample-orienting behavior?

Abstract

In earlier research using constant-delay matching with pigeons, there is evidence that delay of reinforcement of sample-orienting behavior may contribute to the decline in matching accuracy with increasing delay between sample and comparison stimuli. In the present research using this procedure, we found that a significant decline in matching accuracy between the first and second session can occur when delays are relatively long. This effect cannot be accounted for in terms of either additional memory loss or surprise (generalization decrement) associated with the increase in delay. Furthermore, the decline in matching accuracy occurred regardless of whether the delay was inserted between samples and comparisons (where it would be expected to affect the use of sample memory in making the comparison choice response) or between comparisons and reinforcement (where it would not be expected to affect the use of sample memory in making the comparison choice response). Thus, the decrease in matching accuracy between Session 1 and 2 following an increase in delay appears to be unrelated to sample memory at the time of choice. Instead, the results suggest that delay of reinforcement of sample-orienting behavior may play an important role in the negative slope of the retention functions obtained when constant- or mixed-delay matching procedures are used to assess animal memory.