"We refer to culture - or tradition - as all group-typical behavior patterns, shared by members of animal communities, that are to some degree reliant on socially learned and transmitted information". To investigate this very controversial phenomenon, ethologists, up to now, have relied upon what they call the "ethnographic method" (an ethnographic method that is closer to that of the Chicago School than to modern anthropology). In a nutshell, the method (as exemplified in a famous 1999 Nature paper) is the following : repertory all behavioural differences between groups, make sure that no genetic difference explains them, then explain away the environment, and you get a handful of behavioural differences that cannot be accounted for by genes or by the environment, and behold ! You just discovered animal cultures.

The problem with this approach is that it cannot directly assert that the "cultural" differences spring from cultural learning. All it can do is excluding genes and the environment from the analysis, and underline the importance of a third factor. But this third factor might not always be cultural learning. When Michael Tomasello reexamined the data from the '99 Nature paper, he found that in-group behavioral repertoires where not homogeneous enough to have been produced by direct cultural learning. In other words: although it is true that levering is found only in Gombe and not in Mahale, a particular individual in Gombe is not more likely to learn about levering than one in Mahale. So just because a behavior is cultural in the ethnographic sense does not mean it was spread by means of social learning. You'll find many other examples in Laland and Janik's very comprehensive review.

According to the authors, the ethnographic approach fails because it cannot see the way genes, environments and public informations can interact. In a pure "Triple Helix" fashion, Laland and Janik advocate a thoroughly integrative approach. Their moto is : just because a behaviour is genetic or environmentally induced, does not mean it's not cultural, and conversely, just because it's cultural does not mean the environment has nothing to do with it. They blame the ethnographic method for masking gene-culture-envioronment coevolutions.

This is far from surprising on behalf of niche-constructionists. But what's new and exciting is the change in methods they push for, that is : translocation. Translocation means transporting genetically similar animals in similar environments, and observe the behavioural outcome : if you reproduce the data from your group of origin, then culture is explained away. Same thing with genetically similar animals in different environments, culturally similar/different environment, culturally similar/genetically different (why not?), etc. Any combination is possible. Their model is a beautiful study in blue fishes, which they sum up. Translocation allows one to observe finely the respective effects of genes, environments and culture, in a quantitative and relative way. This is much better than simply ascribing a behaviour to a 'tradition'. This is simple, it seems efficient and it's really exciting. The downside is that, for obvious ethical reasons, it can only be used on the less complex animals (i.e. not in primates).

This makes me think of the approach that Bruno Latour advocates for the human sciences : instead of explaining away cultural features by just saying that they are cultural, we should try to explain culture itself. That means a searcher should not be satisifed when he has managed to attribute such or such phenomenon to culture. Cultures cannot be taken for granted : one must follow their means of transmission, the artiofacts they produce, their interactions with the environment. That's what Latour calls "relocating culture".

What the authors ask of "animal anthropologists" is a move away from descriptive ethnography towards experimental testing, something non-ethologists have long been asking for. If empirical research follows, the change in methods advocated by Laland and Janik might prove paradigm-shifting.