We investigated the acoustic individual discrimination in Japanese monkeys. Two experiments were realized. Two female Japanese monkeys were trained to discriminate vocal individuality on the coo call. For the first experiment, we used for the stimuli 5 individuals with 8 calls per individual. A discriminant analysis applied to those calls indicated that it allowed correct identification of the caller in 100 % of the cases (93 % of correct classification for the cross-validation) and that the best individual discriminating variables were duration of the call and start and end frequencies of the fundamental. In that experiment, one monkey succeeded in individual vocal discrimination on new call exemplars from the 5 stimuli individuals (i. e., response rate to the comparison stimulus was significantly higher when the call belonged to a different individual than when it belonged to the same individual). However, the other monkey failed to perform the discrimination on the new calls. A second experiment using as stimuli 3 individuals with 30 calls per individual was realized. A discriminant analysis applied to those calls indicated that it allowed correct identification of the caller in 98 % of the cases (100% of correct classification for the cross-validation). The best individual discriminating variables were the same as before. For this experiment, the two monkeys succeeded in performing vocal individual discrimination on new call exemplars from the 3 stimuli individuals. In the present study, we demonstrated that individual vocal discrimination on several unknown individuals is possible for Japanese monkeys.

CEUGNIET, Miyuki
Cognition and Learning Section
Primate Research Institute, Kyoto University
Inuyama, 484-8506, Japan
miyuki@pri.kyoto-u.ac.jp