Letter from the President (Society for Anthropological Sciences)


Science has a type of unity. We can distinguish biological, chemical, and physical theory from one another. But when they are applied to something they are mutually consistent. If an inconsistency appears, all concerned recognize this as indicating an error somewhere that must be corrected.

The same idea of unity underlies the “four field” view of anthropology.  This is why the four-field view of anthropology is associated with the idea that anthropology is a science. It is also why those who reject the idea of anthropology as a science reject the four-field view.

The Society for Anthropological Sciences was formed in reaction to the AAA deciding that it would no longer define itself as a scientific society. Since they dropped the burden, we picked it up.

The present state of the AAA, and by implication of anthropology as a discipline, is that there are multiple streams analysis focused on different topics. Leaving aside the sections based on age, ethnicity, and the like, some of the topical groups define themselves as sciences and some do not. Some are ambiguous.

In many conversations over the years, we have told ourselves that we should be acting as a forum to bring all of the scientifically oriented sections together, to advance our common interest.  It is also why the word “sciences” in our title is plural. We can do a better job of it.

Of course there is a difference between real unity and false unity. The “Unity of Science Movement,” founded it at the University of Chicago by Vienna Circle positivists who had fled Europe at the beginning of World War II, represents false unity. Their central claim was that science was a system of propositions.  So all one had to do to create unity was rewrite these propositions in a common propositional language. This was factually wrong. Science simply is not a system of propositions. It is a system of experimentally established facts, and it is in the linkages between these facts that we find the unity.

An experiment is a constructed situation, designed to produce an observable outcome that answers a question by eliminating possible alternative interpretations. It can occur in a cyclotron, a test tube, or an ethnographic interview.   There is also such a thing as an “thought experiment” in which a practical test is imagined but the outcome is so obvious that there is no need to carry out the actions in fact.   But in all forms, an experiment must involve interaction with nature in which nature can fight back if it is mischaracterized.

In order to construct an experiment, one has to have some idea of what one is looking for. In positivist hands, this led to the misleading formulation there could be no facts unless one first imposed one’s own analytic scheme, saying what kinds of facts are possible. This assumption was at the heart of the failures of componential analysis and Levi-Strauss’s structuralism that David Schneider pointed out in his historically important critiques, published from 1965 through 1994.

Schneider’s own work was based on the same positivist assumptions, but he did not realize how thoroughly he had accepted them.  In American Kinship: a Cultural Account (1968), he tried to strike out into new ground by talking about culture as a symbol system rather than kinship as a type of social organization. But the way he did it was the same, imposing his own analytic scheme on interview results rather than designing field experiments that let his informants explain their own ideas in their own terms. He described American kinship as a system of “cultural symbols” “refracted out of” the “central symbol” of incest.    He did not report that anyone who was interviewed ever said any such thing, and it is a safe bet that they did not.  In the process of writing it, he recognized that his criticism of componential analysis and the alliance-descent arguments applied to his own work as well and he could think of no alternative. His Critique of the Study of Kinship (1987) followed. It concluded that a “quartet” of types of organizations that social scientists had focused on for over a century–“kinship, economics, politics, and religion” (1987:181)–were nothing more than “metacultural categories imbedded in European culture which have been incorporated into the analytic schemes of European social scientists” (1987:184).   This is the conclusion that led our colleagues to abandon social organization as a topic, reject science is a hegemonic imposition, and adopt a view of anthropology as “cultural critique” based on subjectivism and Geertz’s interpretivism.   It also, therefore, led to the creation of the SAS.

I am not recommending that we try to come up with a description of the experimental method that will bar all possible misinterpretations. I do not think this is possible. Schneider was my dissertation advisor. The dissertation described my approach to the analysis of kinship terminologies and organizations.  It seemed to me that there was no way he could escape noticing that it was not subject to the circularities he pointed out in componential analysis and the alliance descent arguments, and that it provided a coherent analysis of a kinship system that was meaningful in indigenous terms yet objectively elicitable and verifiable.  Apparently, however, it made no impression.

Logically and probably practically, the best way to advance the unity of anthropology as an experimental science is by demonstration, by concentrating more on explaining carefully and critically the kinds of inferences in conducting field inquiries and subsequently that lead us to our conclusions.  This has not been a major concern of SAS-sponsored papers up to now, but it has very often been the focus of discussions of these papers after they are presented. The late Kris Lehman was especially good at stimulating such discussions.   I think it would be a good idea to put more of that kind of material in our papers to begin with.

The language of science takes its meaning from the way it is embedded in the experimental apparatus of science and not the other way around.

The deadline for submitting panels and papers for the 2018 AAA meetings is Monday, April 16. I urge you to do as much as possible. Submissions can be withdrawn after that date but cannot be added.  If you have an idea for a panel, submit it. If you have an idea for an individual paper, submit it and use this website and our email list to tell others what you are doing and invite them to join you. Later, the program committee will have the opportunity to arrange individual papers into panels.  We can also move papers between panels and adjust titles of panels.

I will submit either a paper on the experimental elicitation of kinship terminologies or the experimental elicatation of peasant farm budgets.