The visual perspective-taking ability of 4 chimpanzees (Pan troglodytes) was investigated. The subjects chose between information about the location of hidden food provided by 2 experimenters who randomly alternated between two roles (the guesser and the knower). The knower baited 1 of 4 obscured cups so that the subjects could watch the process but could not see which of the cups contained the reward. The guesser waited outside the room until the food was hidden. Finally, the knower pointed to the correct cup while the guesser pointed to an incorrect one. The chimpanzees quickly learned to respond to the knower. They also showed transfer to a novel variation of the task, in which the guesser remained inside the room and covered his head while the knower stood next to him and watched a third experimenter bait the cups. The results are consistent with the hypothesis that chimpanzees are capable of modeling the visual perspectives of others.

Four chimpanzees, Pan troglodytes, were individually trained to cooperate with a human partner on a task that allowed both participants to obtain food rewards. In each chimpanzee-human dyad, one of the participants (the informant) could see which pair of food trays on a four-choice apparatus was baited, but had no means of obtaining it. The other participant (the operator) could pull one of four handles to bring a pair of the trays within reach of both participants, but could not see which choice was correct. Two of the chimpanzees were initially trained as informants and adopted spontaneous gestures to indicate the location of the food. The two other chimpanzees were trained as operators and learned-to respond to the pointing of their human partner. After the chimpanzee subjects reached near perfect performance, the roles in each chimpanzee-human dyad were reversed. Three of the four chimpanzees showed immediate evidence of comprehension of their new social role. The results are discussed in the context of cognitive empathy and the potential for future research on social attribution in non-human primates. Genetics, learning and cognition have all been implicated as sources of human and animal behaviour. In particular, cognition has become an important focal point of research in most of the behavioural sciences (Gardner 1986). However, the study of one of the most widely investigated cognitive processes in humans, social attribution, remains relatively neglected in behavioural research with animals. Social psychologists developed attribution theory as part of an effort to understand how people perceive and explain the actions of others, and how these perceptions influence their behaviour (Heider 1958; Harvey & Weary 1981). Social and self-attributions can be thought of as perceptions and inferences about the mental states of others or oneself, respectively (Harvey & Weary 1981). The issue of attribution in animals has not been completely neglected, however, and a number of researchers have experimentally approached the issue in different ways. Gallup's (1970) demonstration of self-recognition in chimpanzees, Pan troglodytes, offered evidence of a self-concept, suggesting that these animals may be capable of some forms of self-attribution. Menzel's (1974) experiments with captive chimpanzees in large outdoor enclosures laid the groundwork for investigating social attribution of emotions and knowledge. More recently, Cheney & Seyfarth (1990a, b) have experimentally explored the

No abstract