Testes de simetria e transitividade em macaco-prego

Abstract

One way of exploring symbolic behavior and its prerequisites is to teach arbitrary relations with a conditional choice training model, with the purpose of verifying the emergence of new relations between pairs of stimuli used in training. The objective of this study was to test the emergence of symmetry and transitivity in a capuchin monkey, applying before tests teaching conditions used in studies that obtained emergent performance in non-humans. An adult male capuchin monkey from the Experimental School of Primates (NTPC-UFPA) with substantial experience in choice tasks served as subject. In phase 1 of the study, the arbitrary relations AB, B-A and A-C, established in a previous experiment, were retrained in both delayed and simultaneous matching-to-sample (MTS) procedures. In phase 2, the same arbitrary relations A-B, B-A and A-C were trained interpolated with identity trials A-A, B-B and C-C. In phase 3, the reinforcement probability was reduced, in preparation for the testing sessions that would involve trials with no programmed reinforcement. In Phase 4, the emergence of the relations C-A, B-C and C-B were tested. Performance was accessed in terms of the percentage of success across all trials. In training session, the performance criteria were 90% minimum of correct choices and, at most, one error per trained relation. In Phase 1, the subject performed with most precision in the delayed MTS (minimum 75% and maximum 100% of correct choices), probably due to its previous experience with this task; in simultaneous MTS, the performance varied between 64,44% and 97,78% and the number of sessions to criteria was higher than in the simultaneous MTS. In Phase 3, there were 23 sessions to criteria, with an average performance of 89,32% of correct choices in the delayed MTS (6 sessions total) and 87,75% in simultaneous MTS (11 sessions). The subject had difficulties with the simultaneous MTS in which two identical stimuli appeared on the computer screen. This training model induced to errors and was corrected by restricting the location where the sample was presented. In Phase 3, only 7 experimental sessions were enough to criteria, with an average performance of 91,66% correct in 4 sessions of simultaneous MTS and 91,10% in 3 sessions of the delayed MTS. In Phase 4, the tests were negative for the emergence of C-A relations. The average number of correct trials was 23,33%. A drop in the baseline performance and emotional reactions in the subject were observed, probably due to the absence of reinforcement. In B-C test, the average number of correct choices was 71%, demonstrating emergent behavior. C-B relations tests were also negative. Performance on these trials average only 12% of correct choices. The procedure was efficient to document the emergence of transitive relations (B-C). According to the literature, the study suggest that non humans seem more ready to present the emergence of transitive than symmetric relations. Future studies shall engage in training C-A relations and retesting the emergence of C-B relations