The formation of learning sets by the chinchilla in an auditory discrimination task.

Abstract

The performance of 9 chinchillas (Chinchilla lanigera) in a completely computer-automated sound-discrimination experiment was observed using stimulus pairs from a library of 180 complex sounds. Each sound of a pair was 0.2-2.08 sec in duration, separated by an interval of 250 +/- 125 msec. A modified "go/no go" paradigm was used. Licking behavior at a water-delivery tube was occasioned by keeping Ss at about 85% of ad-lib weight by water deprivation. After habituation and blank-trials training stages, sound discrimination was required: If the sounds of a pair were identical, no response was required; if they were different, a fleeing response (crossing a mid-cage barrier) was required. Each problem set utilized a different sound pair, for a total of 90 problems. Therefore, Ss learned to perform general auditory discriminations, not sound-specific discriminations. The chinchillas quickly learned to respond, but not at a rate consistently higher than 75 %-correct. The results were compared to those of other studies of sensory discrimination in monkey and chinchilla. Suggestions were made for minor improvement of the paradigm for use in future sound-discrimination tasks.