Chimpanzees' flexible targeted helping based on an understanding of conspecifics' goals

Abstract

Humans extensively help others altruistically, which plays an important role in maintaining cooperative societies. Although some nonhuman animals are also capable of helping others altruistically, humans are considered unique in our voluntary helping and our variety of helping behaviors. Many still believe that this is because only humans can understand others' goals due to our unique "theory of mind" abilities, especially shared intentionality. However, we know little of the cognitive mechanisms underlying helping in nonhuman animals, especially if and how they understand others' goals. The present study provides the empirical evidence for flexible targeted helping depending on conspecifics' needs in chimpanzees. The subjects of this study selected an appropriate tool from a random set of seven objects to transfer to a conspecific partner confronted with differing tool-use situations, indicating that they understood what their partner needed. This targeted helping, (i.e., selecting the appropriate tool to transfer), was observed only when the helpers could visually assess their partner's situation. If visual access was obstructed, the chimpanzees still tried to help their partner upon request, but failed to select and donate the appropriate tool needed by their partner. These results suggest that the limitation in chimpanzees' voluntary helping is not necessarily due to failure in understanding others' goals. Chimpanzees can understand conspecifics' goals and demonstrate cognitively advanced targeted helping as long as they are able to visually evaluate their conspecifics' predicament. However, they will seldom help others without direct request for help.

Fig. 1.

Tool set consisting of seven objects that were supplied to a potential helper. Only one of them (a stick or a straw) was needed for a conspecific to solve either a stick-use or straw-use task in the adjoining booth.

Fig. 2.

Helpers’ first tool selection and offer to their conspecific partner. Each condition [(A) first “can see”, (B) “cannot see”, and (C) second “can see” condition] presented participants in the recipient booth with one of two tool-use situations (“stick” or “straw”). Graphs are based on the data from three participants (Ai, Cleo, and Pal) who completed all of the conditions based on an A–B–A design. For the statistical analysis, see Table 1.

Fig. 3.

Photo showing Ayumu standing up and assessing his mother's situation by peaking through the hole in the opaque panel wall separating the two booths. He was the only chimpanzee to assess so actively his partner's situation and to select and transfer the appropriate tool to his partner in the “cannot see” condition.