Anthropology IS an experimental science

One of my favorite quote from Geertz on anthropology as an experimental science:

The “natural laboratory” notion has been equally pernicious … because the analogy is false. … The great natural variation of cultural forms is, of course, not only anthropology’s great (and wasting) resource, but the ground of its deepest theoretical dilemma: how is such variation to be squared with the biological unity of the human species? But it is not, even metaphorically, experimental variation. (1973: 22) [more…]

By “favorite,” of course, I mean a statement so self-assured of its own common sensicality that it begs to be challenged.  So I thought about it again when, while preparing a class on ethnomethodology as “methodology” (in a methods class), I went back to Garfinkel’s recently published dissertation proposal (from 1949).  There he proposes to conduct experiments through which the construction of a social order might be observed.  The general model for these experiments is stated as:

Assuming Iσ, let there be meant a dyadic group made up of Aγ(x) c Bγ(x). When A is regarded by B (x)-wise, A’s treatment of B will be interpreted in such a way (x) by B as to encompass a change (x) in an element or elements (x) of B’s cognitive style, the change being of such a character (x) as to limit B’s alternatives of action (x) … [more …]

The technique to observe what B will do is simple:

To help us in “slowing up the process” of B’s interpretive activity, we shall use the device of cutting B off by facing him with incongruous material. (My emphasis. 2006: 206-7)

For the rest of his careers, Garfinkel kept imagining versions of the experiment he modeled in this passage.  The most famous (at least for teaching purposes—which is what I imagine I do in this blog) may be the following one:

Students were asked to spend from fifteen minutes to an hour in their homes imagining that they were boarders and acting out this assumption. They were instructed to conduct themselves in a circumspect and polite fashion….

In nine of forty-nine cases students either refused to do the assignment (five cases) or the try was “unsuccessful” (four cases). (1967 [1964]: 47)  [more …]

The less obviously experimental of these observations range from following Agnes through her sex change operation (1967 Chapter 5) to the research on a blind woman organizing her kitchen so that she can cook by herself—only possible if no sighted person helps her (2002: 212ff).  The best set of such observations is the ensemble of research in conversational analysis.  Audio-taping and videotaping does exactly what Garfinkel called for: a slowing down of social interaction so that one can observe the actual building of a social order.

A half century of work in that experimental mode has produced an ensemble of findings about sociability that should be presented more succinctly, and, I dare say, celebrationally.  These findings (laws?) range from the generality of indexicality as the mechanism through which communication is anchored in the here and now, the principles of “trust” (a generalization of the generality of “passing” as another fundamental principle), the “etc.” principle (communication does not proceed through full knowledge of the situation—thereby disproving all forms of cognitivism), and so on and so forth.

One thing that work has not produced is a formalization of the conditions under which a particular social order (this one) comes about and transforms itself.  In other words ethnomethodology and conversational analysis are, fundamentally, a sociology of social ordering.  But there has never been an equivalent anthropology of historical culturing.

Which brings us back to ethnography as, arguably and contra-Geertz, an experiment in “slowing down processes” (or perhaps, in fact, “accelerating” the passing of time).  Boas and others (including Geertz in the above quote) intuited (and hypothesized) that human variability is a fundamental principle.  How would one demonstrate that?  For Boas et al, the answer was simple: by examining social orders in human groups widely separated and, perhaps even more powerfully, by examining social orders in neighboring groups.  Eventually, the more fine grained the analysis, the more one could demonstrate that the same tasks of survival can be performed in all sorts of ways.  For example, middle aged women in graduate school can prepare for an examination just as well siting on the floor in veils (or in blue jeans, sitting on chairs).

Iranian women studying

I tried to formalize this in an earlier blog entry.

But we still need to figure out how specific social orders arbitrary to the “needs” they may appear to fulfill actually do appear in history.  And so, we need to devise experiments that might it possible for us to witness the process.

For a defense of cultural anthropology as science

Given any ordered social state (system, pattern, culture, …), this state will always re-order itself into any number of new states none of them being identical to any state ever produced in human history.

A scientific “law” derived from anthropological research?

I have been thinking for some time about the de-institutionalization of what we might call “anthropological authority”: the authority to speak about humanity from the point of view of an evolving discipline that has developed over more than a century a powerful and distinct way to discover aspects of humanity that other ways of knowing do not bring out.

I thought this movement was a product of the evolution of American political activity where the tendency to “know-nothing” merges with the hyper-expertise of a narrow cadre of techno-engineers convinced that “data-driven” research will necessarily produce “evidence-based” policy and lead to the oft-predicted “end of history.”  Well, there is a French version of that evolution leading, in good French centralized fashion, to the erasure of anthropology from university undergraduate education.

That is a radical threat if ever there was one!  And it leads to people rising in passionate defense.  For example, look at the following:
https://www.facebook.com/PromotionDeLEthnologieAnthropologie/posts/399679680141403

In summary, the petitioners present the major achievements of anthropology over the past century as consisting of efforts

  1.   “to stimulate social reforms necessary for a fairer and more equitable redistribution of produced wealth” (Mauss and The gift);
  2.   [to found] “a new humanism based on a more universalist and egalitarian framework” (Lévi-Strauss and Race and history);
  3.   to oppose attempts to anachronise or exoticise ‘non-Western societies’ in order to comprehend them (Balandier);
  4.   to understand all human societies as equally valid and contemporary (on going).

My question today: Are these the achievements we should celebrate at this time?  Are these the reasons anthropology should be kept as an undergraduate major in French universities?

I find striking that all these justifications are ideological and politically (and may be religiously) charged.  None claim “science.”  Is this the best we can do to affirm our contribution to those who make different political choices, or to those, particularly in the United States who are aggressively “non-political” (actually there is a version of that in France where a particular cadre of government officials, just below political appointees who come and go with governmental majorities, keep serving whoever has legitimate power.  They are “les hauts fonctionaires”)?

So, I’ll try my hand briefly at another kind of justification based on the contribution of anthropology to “basic science.”  Let’s start with Saussure as developed by Jakobson and Lévi-Strauss.  Saussure, on the basis on a century of basic research in historical linguistics established that while any state of language builds on earlier states, one cannot predict the next state.  This is fundamentally related to the “arbitrariness” (non-rationality) of the means through which meaning is achieved.  Anthropological research, particularly in the Boasian traditions, has confirmed that is arbitrariness can be generalized to all forms of human behavior in history (religion, myth, political ideologies, etc.).  This, of course, is also the contribution of Lévi-Strauss in his major works (Totemism, Savage mind, etc.)—among many others.
While all particular historical forms are tied to earlier forms, and must also fulfill various kinds of biological, ecological, demographic, etc., needs, the exact means through which these needs are met are fundamentally “arbitrary” (or, in more recent formulations, “playful”).  All this is true “cross-culturally,” across historical periods, and, as we are now finding out, “cross-“ the various new forms of differentiation produced “internally” within the new global society.

All of this has been established through various forms of detailed ethnographic-like research (including historiography, philology, conversational analysis, etc.) and debated within a small set of social science disciplines.  It may even be written as a “law” that cannot be broken any more than the second law of thermodynamics:

Given any ordered social state (system, pattern, culture, …), this state will always re-order itself into any number of new states none of them being identical to any state ever produced in human history.

The consequences of this general knowledge should lead to a radical challenge of “evidence-based” research to the extent that it is founded on the sense that the evolution of human societies can be predicted and controlled.  That is, as I understand it in the world of school policy I know best, researchers design complex experiments to establish that ‘y’ is function of ‘x’ (z, etc. through complex statistical means) and that this is not a historical, arbitrary, relationship.  There is however no evidence that any such research, in the past, has led to the prediction of even minor changes in the future.  I do not know for example whether the sociologists who developed the framework for “value-added-teaching” ever confronted the possibility that teachers might strike over it, that administrators might dissemble about test results, that state administrations would discover means of subverting the processes, etc (or that we can not predict what will happen to all this when new local or national administrations are installed).  And yet any anthropologist of the past half-century (whether they invoked Boas, Lévi-Strauss, Foucault, de Certeau, etc.) could have warned that such “play” would happen even if they could not predict what forms it would take.

Not accepting the “law” (however we may end up writing it) is placing oneself outside of science and too many of our colleagues are willing to do that.

As for us anthropologists (historians, sociologists, etc.) we must keep training students rigorously to explore implications, challenge, further specify our paradoxical laws.