Visual-visual associative learning and reward-association learning in monkeys: the role of the amygdala.

Three Cynomolgus monkeys (Macaca fascicularis) took part in an experiment on visual learning set in an automatic apparatus. Each new visual discrimination problem was solved using a visual secondary reinforcer consisting of a white line. If the monkey chose the correct stimulus (by touching it), the white line appeared over the correct stimulus. Primary food reward was delivered only after a new problem was solved to a criterion, and the problem was then replaced by a new one. Thus, within-problem learning did not rely on primary reinforcement but on the visual secondary reinforcer. The animals were trained preoperatively in visual learning set with this procedure and were assessed postoperatively for their ability to learn new visual discriminations with the same procedure. Bilateral amygdalectomy did not significantly impair the animals' learning ability in this task. Learning remained unimpaired when transection of the uncinate fascicle and of the fornix was added to amygdalectomy. The effect of bilateral amygdalectomy in this task was much less severe than in a similar task we previously studied, with auditory secondary reinforcers. The results show that the involvement of the amygdala in processes of secondary reinforcement depends on the sensory properties of the secondary reinforcer. From these and other recent results, we conclude that the sensory attributes of a reinforcer are easily associated with a discriminative stimulus when they are in the same modality and same spatial location as the discriminative stimulus and that this sensory-sensory association is independent of the amygdala.