| Literature DB >> 20971169 |
Bezawork Afework Bogale1, Masato Aoyama, Shoei Sugita.
Abstract
We trained jungle crows to discriminate among photographs of human face according to their sex in a simultaneous two-alternative task to study their categorical learning ability. Once the crows reached a discrimination criterion (greater than or equal to 80% correct choices in two consecutive sessions; binomial probability test, p<.05), they next received generalization and transfer tests (i.e., greyscale, contour, and 'full' occlusion) in Experiment 1 followed by a 'partial' occlusion test in Experiment 2 and random stimuli pair test in Experiment 3. Jungle crows learned the discrimination task in a few trials and successfully generalized to novel stimuli sets. However, all crows failed the greyscale test and half of them the contour test. Neither occlusion of internal features of the face, nor randomly pairing of exemplars affected discrimination performance of most, if not all crows. We suggest that jungle crows categorize human face photographs based on perceptual similarities as other non-human animals do, and colour appears to be the most salient feature controlling discriminative behaviour. However, the variability in the use of facial contours among individuals suggests an exploitation of multiple features and individual differences in visual information processing among jungle crows.Entities:
Mesh:
Year: 2010 PMID: 20971169 DOI: 10.1016/j.beproc.2010.10.002
Source DB: PubMed Journal: Behav Processes ISSN: 0376-6357 Impact factor: 1.777