Models and Social Reality: An Alternative to Caws

Abstract

Recently in these pages (AA 76:1-10, 1974) Peter Caws claimed that the objective social of the societies we study is dependent, in part, upon our own theories or models of them. is the scientist's representational (i.e., explanatory) model, the theory he constructs to account for the data and their interrelation, that confers objective on the system. And the use of 'confers' is deliberate, since it would be quite accurate to say that until the explanatory model was constructed the system had no objective structure (p. 7). This might come as something of a surprise to anthropologists. Most of us have probably not considered it part of our job to confer objective on the societies we study by theorizing about them, much as children in Peter Pan keep fairies in existence by believing in them. The more common view is that objective social is something a society has (perhaps not at the empirical level) before we get there-and is, indeed, one of the main things we go to learn about. It is not something the society receives from us. I want to argue, against Caws, for the more common view. At the same time, I want to acknowledge his contribution in constructing a compelling argument that challenges us to think carefully about some of the fundamental issues in anthropology: the nature of the phenomena we study and the nature of our knowledge of them. Before proceeding with my critique, let me summarize Caws' argument as I understand it. The relations studied by social science are mind-dependent. That is, while natural relations such as between the moon and the tides exist and operate whether men know about them or not, social or cultural relations can be nothing other than what they are construed by men to be. There is no universal intrinsic social relationship between father and son; proper behavior between them is just what it is defined in one or another kinship system to be. Often such definitions can be attributed to the natives: in their explanation and conceptualization of the system (Caws calls these representational models) and/or in their conscious or unconscious motivation for acting (operational models). Occasionally, however, anthropologists say things about social systems that cannot be attributed to native