IPS-SLAPrim 2020 Information IPS-SLAPrim 2020 About Quito IPS-SLAPrim 2020 Program IPS-SLAPrim 2020 Registration ISP-SLAPrim 2020 Primates in Ecuador IPS-SLAPrim 2020 Tours IPS-SLAPrim 2020 Contact Us

Abstract # 1505 Event # 34:

Scheduled for Monday, June 26, 2006 03:20 PM-03:40 PM: Session 4 (Kama B)

A Cross-Modal Identity Recognition Task : How Chimpanzees (PAN TROGLODYTES) and Humans Match Voice to the Vocalizer's Picture

L. Martinez and T. Matsuzawa
Primate Research Institute, Kyoto University, Language and Intelligence Section, Kanrin 41, Inuyama, Aichi Prefecture 484-8605, Japan
     Previous studies have emphasized that the vocal behavior of chimpanzees is closely related to the complexity of their social relations, which involves frequent intra- and inter-group mixing among a large number of individuals. This spatial dispersion requires efficient non-visual recognition of the vocalizer. In this study, we attempt to investigate the relationship between auditory and visual categorization of identity through a cross-modal matching-to-sample task. In preliminary experiments, subjects were required to match playbacks of pant-hoot calls from 3 familiar chimpanzees with the corresponding vocalizers’ facial pictures. Among a group of 14 captive chimpanzees, one adult female named Pan, an expert in audio-visual tasks at the beginning of the study, obtained 90% of correct responses within the first two sessions, while a second female chimpanzee named Ai, with no prior experience in audio-visual task, scored 46%. Human subjects familiar to these chimpanzees (n=11) were tested under the same conditions. Their performances were located just between the two chimpanzees, with 73% correct on average (ranging from 54 to 88% among subjects). In ongoing experiments with inexperienced chimpanzees, we are manipulating several parameters of the stimuli and procedure to facilitate the acquisition process. Afterwards, we aim to test for any correlation with age or other individual experiences among subjects, and to assess how familiarity, “other-species effect” and social preference can affect this cross-modal identity recognition.