Cumulative culture in the lab and chimpanzees

At the recent EHBEA conference held April 6-8 at Saint Andrews, I saw presentations by both Andrew Whiten (a primatologist who specializes on nonhuman cultural traditions, especially in chimpanzees) and Christine Caldwell (who examines cumulative cultural evolution in the lab). It was interesting to see the question of cumulative cultural evolution from these two very diverging perspectives.

It is now generally established that nonhuman animals, including chimpanzees, macaques and a variety of bird species, display a socially transmitted behaviors, which in humans are termed cultures. However, to date the evidence for cumulative cultural evolution in nonhumans remains sporadic. For example, in the case of nut-cracking chimpanzees in the Taï forest, there is little variation in how nuts are being processed, i.e., cracked by means of a hammer and anvil, and whereas some individuals have learned to use auxilliary stones to stabilize the anvil, this innovation has not spread to the entire population. The question is: why not?

a typical use of anvil to crack a nut from a Taï forest chimpanzee

Andrew Whiten recently co-authored a study in which chimpanzees were confronted with an optimal and cumulatively built technique for extracting honey from an artificial device. Whereas the individuals learned the simple 'dipping' technique with ease, they did not master the more complex 'probing' technique, which built on elements of the dipping technique they already mastered. The chimpanzees thus got 'stuck' at a simple but suboptimal technique, although control tests showed that the more difficult technique was not beyond their cognitive capacities. Why would this be? Whiten tentatively suggested in the paper that chimpanzees may be 'conservative', unwilling to try a new technique if the one they already knew was good enough.

Another line of reasoning, which has garnered much attention, is that of Tomasello.

Tomasello, for example in this paper together with Carpenter, thinks that it is shared intentionality, such as joined attention and active instruction, which are uniquely  human capacities, that allow for cumulative cultural evolution. In children, but not chimpanzees, we can witness the development of triadic interactions (i.e., sharing attention with someone over an object, and knowing that you both share this attention).


Paper airplanes differ considerably in their capacity for flight. Subjects who can build on earlier designs do better.

However, Caldwell's study on cumulative cultural evolution in laboratory context (see this pdf for an earlier experiment from her team) suggests that shared joint attention is not necessary for cultural evolution. She let people construct a paper aeroplane which flew as far as possible, and in the other, they were instructed to construct a tower of spaghetti which was as tall as possible. Importantly, people did better if they could build on the solutions of earlier participants (in agreement with the intuition that cumulative cultural evolution allows for more optimal solutions). However, in her talk on the EHBEA conference, Caldwell indicated that it was not necessary for participants to see the others actually make the planes or construct the towers. Of course, it is intuitively obvious that humans don't need to see an artefact actually being made to copy it – we can just backward engineer the solution. This suggests that several cognitive capacities, rather than one silver bullet explanation, might underlie the capacity for cumulative cultural evolution.

1 Comment

  • comment-avatar
    sophie litas 11 May 2009 (15:26)

    I may sound off tune but it seems, concerning the question why the chimps did not want to learn the “probing” technique, that it can not properly be answered without first addressing the assumptions behind this question. What is assumed when we do not understand why they do not switch to the more “productive” behaviour is: 1- that they understand this technique is more productive in terms of honey 2- assuming that they do somehow get clues about the fact:”I can get more honey”, why do we jump to the conclusion that he would want to get more? This leads us to view them as “conservatives”, a too anthropomorphic (and connoted) concept in my opinion. I obviously acknowledge the fact that, seeing the world through our brain, we can only but analyse it with human concepts, but isn’t it too far fetched here? In a nutshell, words such as suboptimal, productive and the framework implicitly linking higher cognitive capacities with the understanding that more is better, seem quite problematic in this kind of research. Probably even impeding meaningful conclusions. What do you think?