Functional fixedness in chimpanzees

Abstract

Differences in the tool use of non-human primates and humans are subject of ongoing debate. In humans, representations of object functions underpin efficient tool use. Yet, representations of object functions can lead to functional fixedness, which describes the fixation on a familiar tool function leading to less efficient problem solving when the problem requires using the tool for a new function. In the current study, we examined whether chimpanzees exhibit functional fixedness. After solving a problem with a tool, chimpanzees were less efficient in solving another problem which required using the same tool with a different function compared to a control group. This fixation effect was still apparent after a period of nine months and when chimpanzees had learned about the function of a tool by observation of a conspecific. These results suggest that functional fixedness in our closest living relatives likely exists and cast doubt on the notion that stable function representations are uniquely human.

Figure 1

Procedure in Study 1. Participants either drank juice from a drinking container with a hose (experience group) or explored the hose with the empty drinking container present (control group). After this prior experience phase, participants were presented with a horizontal tube with two blockages close to its sides, and had to select the flexible, but rigid hose among three tools (hose, stick, string) to poke out a food reward from the tube. This test assessed if apes become functionally fixated on the tool’s previous function by their experience with it.

Figure 2

Procedure in Study 2. (A) The participant (on the left side) either observed how a conspecific demonstrator (on the right side) drank juice from a drinking container with a straw (experience group) or how she handled the straw without inducing a function (control group). (B) After this prior experience phase, the participant was presented with a U-shaped tube. She had to select the flexible straw among three tools (straw, stick, string) to poke out a food reward from the tube.

Figure 3

Results of Study 1. (A) survival time until success (or until the end of the session at 300 s; survival time is a compound of success and time passed), (B) survival time until target tool extraction and (C) latency until target tool contact across groups and sessions (median; boxes: 0.25, 0.75; whiskers: 0.025, 0.975). Session 1 and 2 were conducted directly after the prior experience phase, whereas Session 3 was conducted after a delay of nine months. Circles designate individual performance; the circle size is proportional to the represented number of individuals. ^aOne participant succeeded in second 300.

Figure 4

Results of Study 2. (A) survival time until finding the solution strategy (or until the end of the session at 300 s), i.e. inserting the hose into the tube without necessarily touching the reward, (B) survival time until target tool extraction, and (C) survival time until contact with the target tool across groups and sessions (median; boxes: 0.25, 0.75; whiskers: 0.025, 0.975). Circles designate individual performance; the circle size is proportional to the represented number of individuals.