Visual Identification following Inferotemporal Ablation in the Monkey

Nine monkeys took part in two experiments on visual learning and the effect of inferotemporal ablation. The first experiment contrasted lesion effects in two tasks given to separate groups of animals: serial visual reversal learning set with only two visual stimuli in use throughout the experiment, and discrimination learning set with a new pair of visual stimuli for each discrimination problem. Following inferotemporal ablation, the within-reversal learning rate recovered to the normal pre-operative level, while within-problem learning in discrimination learning set remained substantially impaired. This result suggests that inferotemporal ablation does not directly impair visual associative learning, but rather produces a defect in visual identification such that only a small set of familiar stimuli can be correctly identified and associated with reward. The second experiment replicated an earlier experiment in showing that monkeys with inferotemporal ablation could learn at a normal rate to choose between stimuli according to their spatial orientation. A relation between the two types of unimpaired visual learning in the operated animals was suggested and analogies with clinical syndromes were discussed. The experimental results were interpreted within a distributed-trace model of learning, and this interpretation was illustrated and supported by a computer simulation presented in the Appendix.