1. Population variability

The existence of significant population variability is convincingly documented by the authors, who find it as soon as they look beyond the ridiculously narrow samples on which claims of universality have typically been based. Although in agreement with the finding, we have a problem with the authors' underlying assumption. This is the idea that it is possible to neatly sort variable features of human behavior and psychology from universal ones. As anthropologists, we have no problem in accepting that cultural, historical, and environmental contexts affect all the features discussed in the article, but this observation has no bearing on the question of whether such features are “universal” or “variable.” It is the variable/universal dichotomy itself (and the questions it generates) that is misleading. This is because human beings are affected simultaneously by processes of a different nature, among them phylogeny, history in its social and cultural instantiations, and ontogeny. But none of these processes is ever active in isolation, making it impossible to track its universal or variable effects. Searching in any human phenomenon for the clear signature of one of these processes in isolation is a wild goose chase.

2. WEIRD as outliers

The authors use three broad population contrasts in order to zoom in on the weirdness of the subject population used to generalize about human nature. The point is well taken, as is the call for more research among non-WEIRD populations. In their eagerness to condemn the reliance on WEIRD subjects, however, the authors end up presenting and conceptualizing population variability in terms of extremely dubious categories. Curiously, while they feel the need to clarify what they mean by the term “Western” and to acknowledge its limitations, they offer no apology for using “small-scale societies” as if the term referred to a unified, meaningful whole (a similar point could be made for “non-Westerner” or “East Asian”). This uncritical lumping together of a variety of disparate societies is particularly odd in a paper that denounces unsound generalizations. As clearly demonstrated by the results of the economic games, some “small-scale societies” can vary just as much among themselves as they do from the WEIRD population – a fact that should not be surprising given that “small-scale societies” are as caught up in the flow of human history as any other. One could argue that the extreme weirdness of the WEIRD population is partly the result of having lumped together other populations under too simplistic and under-theorized labels.

3. Our proposal

As anthropologists committed to the study of human nature (see Bloch Reference Bloch and Bloch2005), we welcome Henrich et al.'s critical appraisal of the behavioral sciences' comparative database. We feel, nonetheless, that the authors have not sufficiently taken to heart the fundamental implications of their analysis. One obvious conclusion they might have drawn is that behavioral scientists should pay more attention to the work of cultural/social anthropologists, since these are the scientists who have made human variability their main focus. It is striking, however, how little reference Henrich et al. make to anthropological research. This, of course, is no accident. It has to do with the kind of data that anthropologists have produced, which in turn has to do with the history of their discipline.

At the start, anthropologists went to the field with ready-made questions that were generated by a simplistic, yet highly influential, evolutionary theory, which is still the basis of popular understandings of the difference between “civilized” and “primitive” societies (the latter sometimes euphemistically called “small-scale”). But such outmoded theory had to be abandoned because, it was soon realized, human history does not proceed along a progressive and unilineal path. Because of the human capacity for culture, each human society is the unique product of a unique, albeit not isolated, history.

Ever since the recognition of this fact, anthropologists have faced a methodological difficulty: Questions formulated from within one historical context produce misleading answers when transposed elsewhere, as they appear weird, uninterpretable, or mean something else (arguably, this is what generates the weirdness of the WEIRD population, since what distinguishes it from all the others is that it is the one that generates the questions). The way anthropologists have tried to overcome this challenge has been to abandon, initially at least, all questions formulated outside the context under their investigation. Rather, through participant observation, they have allowed themselves to discover, from the inside, the terms and values of the people they study. This strategy is not without difficulties, as it generates a kind of data that appears impressionistic and anecdotal and which, crucially, precludes comparison and generalization – which is why such data is so often ignored by other behavioral scientists, such as the authors of the target article. We recognize that this is a very serious limitation, but we insist that behavioral scientists must acknowledge and never underestimate the equally serious and unavoidable problem that led anthropologists down this methodological route in the first place.

Therefore, the solution cannot be, as suggested by Henrich et al., to administer studies upon studies to the billions of (poor) people around the world who remain untapped by the behavioral sciences. The solution is far more complicated and costly. It requires an often uncomfortable compromise between internal validity and generality, and a lot more detailed ethnographic work than many seem to be willing to accept (see, e.g., Astuti & Harris Reference Astuti and Harris2008; Astuti et al. Reference Astuti, Solomon and Carey2004). Only in this way will data from non-WEIRD populations become a meaningful and indispensable ingredient of any general theory about our species.