Do Chimpanzees Have Culture?

Blog || Politics || Philosophy || Science || Fiction || Quotes

"Culture": a nebulous, ubiquitous word meant to capture something essential to human existence. We Homo sapiens sapiens have traditions that are passed on through generations, socially-transmitted behaviors that transcend not only genetics but environmental pressures. That is, not only do we have shared practices, but they are socially learned, rather than being pre-programmed in our DNA or prompted merely by the environment we live in.

(While shared practices of throwing rocks at predators might develop in a location where the environment has rocks available - even if no one learned to throw rocks from each other but picked it up individually - that would not be considered cultural behavior).

Culture, like self-recognition, language and altruistic behavior, has been suggested as a trait which humans alone possess, something that sets us apart from the other animals. These claims follow in the wake of historical attempts to cast animals as purely mechanical entities (incapable of sensation, only able to make simple pre-programmed responses). However, we now know that animals do experience sensations like pain and pleasure, they do learn and adapt their behavior over time. Likewise, mirror-tests have demonstrated self-recognition in some other species (including chimpanzees). Altruistic behavior has been demonstrated in countless other creatures, spawning copious ethological literature. Even language may not be completely our own domain, as relatively recent ape language research studies hint at. So will culture too be found in nature?

Field Studies

Jane Goodall, Andrew Whiten and others, in 1999, combined the data from a large number of influential longitudinal field studies of chimpanzees to explore whether the chimpanzees had behaviors which could be classified as traditional. They identified dozens of behaviors as customary or habitual at some sites while being absent elsewhere, even after apparently ruling out ecological and genetic factors. For example, nut-cracking behavior as well as its absence were observed in groups of the same genetically similar sub-species in similar habitat on opposite sides of a river - they shared the same genes and environmental stimuli, but their shared practices differed. They claimed that this provided hard evidence for culture (that is, tradition) in non-human primates.

A more specific example attempting to establish the existence of tradition is the "leaf-pile pulling" behavior of chimpanzees at Mahale (Nishida et al., 2003). The behavior involves an individual turning around when traveling down a slope, walking backward and raking dry leaves loudly with the hands for attention. It is almost always done in the presence of an audience (i.e. troupe procession), which presents an opportunity for social learning. The behavior is observed in most immature (and many mature) chimpanzees in the Mahale group, but not in chimpanzees in ecologically similar locations. They perform this behavior rather than other alternatives which exist but are not customary (e.g. "leaf-pile pushing", which lacks an audience). There is plenty of opportunity for observation of others doing it, and a long history of the behavior among the group. The authors suggest that this provides strong evidence that leaf-pile pulling is a cultural tradition.

The central point to notice in these examples of alleged culture is the attempt to control for genetic and ecological factors in order to supposedly leave only social learning as an explanation for the difference between groups. Can researchers in the field really rule out ecological explanations as they claim? The Whiten article apparently did so on consensus of the experts, but it is not hard to imagine ecology coming into play in unobvious ways. For example, some critics (Laland and Hoppitt, 2003) point out that when one group performs nut cracking and another does not despite the same nuts and 'hammers' being available, it could be that there are simply better alternative food sources for the group that did not crack nuts, so that is why they do not do it; in a different situation they might do it for ecological and not cultural reasons.

Likewise, Galef and Laland (2005) cite a famous alleged example of culture in chimpanzees at Gombe (East Africa) and Tai (West Africa), where the animals use distinctly different ant-fishing techniques and tools. They mention a study which showed that the tools and techniques used by the chimpanzees were in fact predicted by the behavior of the ants at each location - that is, ecological variables sufficiently accounted for the chimpanzees' behavior. Since ecological factors could not be ruled out, social learning could not be proven. Thus, the claim that this was a cultural tradition does not stand up based on the evidence presented.

This highlights a general problem with field studies of cultural tradition. Even when environmental variables appear to be ruled out, they may still come into play in subtle ways. Hence, the control of experimental laboratory conditions may be necessary to fully control for genetics and ecology in order to demonstrate that social learning actually leads to differences among populations. Researchers in the UK and at Emory University in Georgia attempted to do just this in an elegant experiment with chimpanzees.

Experimental Demonstration of Culture

First they designed a simple puzzle box to put just outside the mesh that surrounds the chimpanzee enclosures, out of their reach but within reach of a wooden stick provided to them. The box could be opened one of two ways (a lifting maneuver and a poking maneuver), releasing a food treat from inside to reward the opener.

Next, they separated the chimpanzees into three groups: a control group, a "lift group", and a "poke group". All three groups of chimpanzees were kept separate from each other during the study so that they could not influence each other. One dominant chimpanzee from the lift group was taken out of the group and taught to open the box using the stick in a lifting motion. A different dominant chimpanzee was taken from the poke group and separately taught to open the box using the stick in a poking motion. These individuals were placed back into their respective groups, while none from the control group was taught how to open the box.

The chimpanzees in the control group never figured out how to open the box. This suggests that the chimpanzees were not genetically predisposed to solve the problem using either a lift or a poke, and exposure to the problem alone did not lead to puzzle-solving behavior. However, the really interesting results came from the other two groups. Those chimpanzees in the lift group soon picked up how to solve the puzzle box using the lift technique. Those chimpanzees in the poke group quickly learned how to open the box with the poke technique. They had the same environmental setup: the same sort of home enclosure, the same box, the same stick available. The only thing that differed between groups was the behavior of one of the other individuals in the group, and that alone was enough to account for the different behavior of the groups. They had established shared behavior patterns by learning socially from another member of their group.

However, the results are more impressive than just that. A few of the chimpanzees spontaneously figured out how to open the box using the opposing method. For example, an occasional chimpanzee in the poke group would learn to solve the puzzle box with the lifting technique (and did it more than once, so it was not just a fluke accident). Amazingly, though, such chimpanzees always ended up going back to the style of their group, despite showing mastery of the opposing method. That is to say, the chimpanzees not only learned the task from another individual (it was not genetic, and the control group showed that they would not have learned to solve it without social learning), but they showed a conformity bias, suggesting a mechanism for preserving socially learned behavior through time. In other words, the experiment appears to have demonstrated using rigorous experimental controls that tradition exists outside of humans.

Shared Cognitive Bases

So culture, like many other traits once thought distinctly human, appears to be present in the animal kingdom. It is no accident that as humans-are-special theorists propose ever more specialized functions as potential dividing lines between us and animals (self-recognition, language, altruism, culture), that our closest phylogenetic relatives are those in whom we most easily find counter-examples to demonstrate that humans are not completely special. Thus it is in great apes that we find, if not all the abilities of humans (for there is no reason they should share everything with us; we have indeed adapted since branching off), then at least predecessors to those abilities.

Chimpanzees will never be taught to use human language in as complex a manner as we use it, and perhaps their socially-learned traditions are not as involved as our own memetic inheritance, but the basics are there, the cognitive stepping stones exist. We humans did not appear out of the blue with our special powers of reasoning and culture - we evolved from a shared ancestor with the apes, and as such we still share much of their underlying neural toolkit. We have simply specialized that toolkit, and in doing so produced full-fledged language, tools to make tools, books to preserve words, and a varied philosophical spectrum that allows us to ponder and debate the mental capacities of other animals. But we are forever to be reminded of our beginnings by the surprisingly 'human-like' behaviors and capacities found in our primate cousins.

Originally Written: 03-07-06
Last Updated: 03-07-06