In lieu of an abstract, here is a brief excerpt of the content:

Chapter Three The Study of Culture: A Framework for Theory and Methodology Everyone can sing and has songs inside. If you don’t think you can sing, go into the forest and start singing. Pretty soon the trees will start to respond —swaying and moving their leaves. —A Menominee elder IT IS HARD to do cultural research without having a clear definition of culture in mind. Serious questions come up whenever cultural comparisons are undertaken—for example, how to decide what groups are relevant to study; how to select samples of participants; how to measure whatever it is you want to measure; and how to interpret any differences or similarities found. Researchers’ specific notions of culture and cultural processes go far to determine the way they answer these and other questions. The logic of cross-group comparisons is very, very tricky. In the cognitive sciences most of the researchers have been trained on methods where there is an independent variable, which the experimenter varies, and one or more dependent variables, which the researcher measures. For example, the independent variable might be amount of caffeine given to research participants in some drink and the dependent variable might be their speed and accuracy at solving multiplication problems. Culture isn’t like that. You can’t randomly assign people to cultures in the same way you could assign them to groups given varying amounts of caffeine. Cultural variables are multifaceted. In short, culture is neither an independent variable nor a unitary thing. The peculiar properties of cultural comparisons create the dilemma that if you compare two groups, there are two possible results and both 23 24 Culture and Resource Conflict of them seem like bad news. If one finds no differences between the groups, the results may not be considered particularly newsworthy and you would have gone to a lot of work to show what other people had been taking for granted. But at least then the generality of the results would have been confirmed and would be on firmer ground.1 If, on the other hand, one compares two groups and finds clear differences , problems of interpretation quickly emerge. Which of the many ways in which the two groups vary are crucial? For example, Alejandro López et al. (1997) found that U.S. undergraduates and Itza’ Maya of Guatemala showed a different pattern of responding on a reasoning task involving mammals. This is an important finding in that it undermines the idea that the particular reasoning phenomenon is universal. So far, so good. But figuring out what causes these group differences is a considerable challenge, because the two groups differ in myriad ways, including age, education, literacy, livelihood, language, and world view. It is practically impossible to disentangle these various factors, because cultural groups cannot be found that represent comparisons of single variables holding other factors constant. A researcher who compares cultures may be confronted by the dilemma of finding either weakly informative similarities or uninterpretable differences. Here’s another problem. Suppose we could control for age, education, literacy, and other features when we compare Itza’ Maya and undergraduates . How do we decide which variables represent “culture” and therefore should not be controlled for, and which variables do not, and should be controlled for. The Itza’ Maya practice agroforestry and also hunt and collect plants in the forest. To be Itza’ is to be connected with the forest. This raises the question of whether intimate contact with the forest is a factor to be controlled or whether we should think of it as part of Maya culture. The answer is not clear. One could require that only American foresters be used in any Itza’–United States comparisons, but then one would be comparing Itza’s who have a typical Itza’ occupation with Americans who have a culturally atypical occupation, and possess lots of knowledge that is atypical for Americans in general. Or imagine if we controlled for occupation by contrasting a sample of Americans with whitecollar jobs with a handful of the very few Itza’ holding corresponding positions in Guatemala. That just won’t do. Suppose that we do control for every variable we can think of and still find differences. In this case, it seems that we may be more or less forced to reify or essentialize culture. That is, the only explanation of the cultural difference involves appealing to some abstract notion of “culture.” This leaves us caught between two equally undesirable possibilities: either to end up with a...

Share