Food preference measured by tokens as exchange tools in chimpanzees

Claudia Sousa and Tetsuro Matsuzawa

University of Coimbra, Portugal

Hominoids, humans and the great apes, use a variety of tools in contrast to other living primates such as monkeys and lemurs. Recent studies have established a long list of tools utilized by wild chimpanzees. However, while the nature of chimpanzee tool use is characterized by a one-to-one-correspondence between any particular tool and its target item, humans have developed a unique tool referred to as a "token". According to our definition, a token is an object that can be used for various goals and has a symbolic relationship to the target. In our society, there exist a large number of "symbolic tools", including coins, bills, tickets, and cards, that can be exchanged for different classes of desired items such as food, objects, comfort, and amusement. A token is unique in several aspects. It is easy to handle and to transport, and can be exchanged for various kinds of items. Values remain unchanged for extended periods such that a token can be saved. In this sense, it can function both as tool and as an effective reward. The purpose of the present study was to analyze the symbolic-tool aspect of tokens in a free choice situation in chimpanzees. Two adult female chimpanzees, Ai and Pendesa, served as subjects. They had already had some experience in the use of tokens. The subjects performed a symbolic matching-to-sample task involving associations between color and visual symbols. A token reward (a Japanese 100-yen coin) was given for each correct response. After each receipt of a token as the reward, subjects were allowed to exchange it for their preferred food by inserting it into a specially constructed "vending machine". A free choice situation followed the insertion of every coin, where two pictures showing different food items appeared on the monitor. As the chimpanzees indicated their favored choice, the vending device delivered the corresponding food item. We tested 45 possible combinations of 10 different food items. The subjects worked spontaneously for the tokens; Ai saved a few consecutive coins and Pendesa saved up to 15-30 before exchanging them for food. There was no significant difference between subjects in their total choice frequencies. The chimpanzees' order of preference was as follows. Ai: pistachio, blueberry, peanut, banana, chow, raisin, grape, potato, carrot, and apple; Pendesa: blueberry, pistachio, peanut, raisin, chow, grape, potato, banana, apple, and carrot. The rank correlation between our two subjects was r=.691. These results demonstrate that both subjects worked for tokens as rewards, and used them as tools to obtain different food items in various choice contexts. The use of tokens provided a sensitive scale of preference for food items in chimpanzees.

Sousa, Claudia

University of Coimbra

Largo D. Dinis - 3000 Coimbra, Portugal

caia@ci.uc.pt