The Field as an Experimental System

In Culture and Cultural Analysis as Experimental Systems Michael Fischer adopts Hans-Jorg Rheinberger’s notion of ‘experimental system’ to describe the history of anthropological theory and the culture concept. As a an update of the ‘long review essay as theory’ it is an interesting example of the genre, but it is problematic in other ways. The point of Fischer’s article seems to be that ‘culture’ is a concept that has morphed over time in different ways as anthropologists (and others) have used it as the lens through which their research problematic is inflected.

As a way of understanding the interconnection of the history of this concept with the epistemological, political, and ethical values analysts bring to the table this is an interesting idea. The problem is that it doesn’t seem to have too much to do with Rheinberger’s concept of ‘experimental system’. Rheinberger uses this term to foreground the practice of science, the actual artifactual nature of a laboratory’s set up and its influence in experimental practice. In doing so Rheinberger tries to move away from histories of science which document, on the one hand, intellectualist unfoldings of scientific ideas or problematics through time and, on the other, approaches which relegate the world to background ‘conditions’ or ‘contexts’ out of which scientific practice emerges. Rheinberger wants to talk about what happens when you get a new, more powerful centrifuge, or when you move the refrigerator slightly to the left. Fischer, on the other hand, seems to be using their term in a much more old fashioned (although stylistically more adventurous) way.

Now you may ask: Since when did the quality of a scholar’s work hinge crucially on how closely they hew to The One True Word Of Hans-Jorg Rheinberger? And of course this is a good point — rip mix and burn baby. But this did get me thinking, what is anthropology’s equivalent of an experimental system?

Its an interesting question, because what really makes scientists ‘scientists’ is the distinct form of knowledge and practice that crystallizes around a lab, which creates a sort of sandbox for experimentation for some variables. At first I thought the anthropological equivalent would be our fieldnotes, since these are purified, recontextualized bits of life that can be manipulated, sorted, and searched through in a way that creates an ‘archive’ whose affordances have a scarily large effect on what sort of research results we produce.

Then it struck me — maybe it seems stupid to you that it took me this long — that the equivalent for anthropologists really is the field site. A lot of doing fieldwork means transforming your situation in your fieldsite from just ‘being in the field’ to ‘doing fieldwork’ which means creating routines, instruments, methods, and relationships which allow you to do things (like census, interviews, transcription, etc.) which are more or less like embryonic experimental systems.

If this is true, then it seems to me that anthropologists differ from bench scientists in two important ways. First, we do a lot more ‘being in the field’ and a lot less ‘fieldwork’ than most of us would care to admit — and that includes the people who see ‘fieldwork’ as alienatingly objectivistic, scientistic, obsessed with a false standards of neutrality and objectivity etc. etc. I have the idea — totally unbased on any actual evidence — that through the past couple of decades the ratio of being to doing has grown greatly. This has implications.

Second, anthropologists rarely spend much time in the field. Even ones at ‘research universities’ like me are really paid to teach, and must show tremendous amounts of hustle to get the funding together for major time in the field. This has to do with lots of things (and of course bench science in unis involves juggling teaching duties too) but key among them is that for most of us the field is just not that easy to get to, and it takes time to get things set up when you arrive.

There are, of course, people for whom The Field is right next door. And indeed, often people so situated and so inclined do have the ability to produce short research notes on findings which are similar to the sort of stream of publications that come out of labs, including collaboration with ‘locals’ and grad student etc. This really is quite a different pattern than the usual ‘three years ago I spent two months in my field site and here is another article culled from the experience’ pattern you often see in some anthropological work.

Now, aspirations (and realizations) of nomothetic, experimentalist inclinations require more than just propinquity. You’ve got to ‘want to be scientific’ as well. But it would be interesting to examine more the way that abstract debates about anthropology’s status ‘as a science’ and what ‘science’ is were examined through the lens of our concrete research arrangements rather than abstract analyses of ‘what we do’ or biographical scrutiny of particular anthropologists in their particular fieldsites.


Alex Golub is an associate professor of anthropology at the University of Hawai‘i at Mānoa. His book Leviathans at The Gold Mine has been published by Duke University Press. You can contact him at

7 thoughts on “The Field as an Experimental System

  1. This is thought provoking, although a bit confusing, perhaps because you don’t allow science to be multiple here. On the one hand, I really like the idea of the fieldsite as the experiment–one of the key insights of rheinberger’s work is that what makes experimental systems work (not just experiments, but systems of experiments) is that they *generate surprises* and are deliberately designed to do so. To speak of the fieldsite as a generator of surprises is to recognize that we design the fieldsite, to some extent, to generate surprises–things we didn’t know before going there. On the other hand, this wreaks havoc with the naive realist position of fieldwork as objective observation of human behavior because in that idiom (also a kind of science, but not, I would say, an experimental science), the fieldsite represents the natural state of society, or of human behavior. I’ve always thought the experimental version of science more intriguing than the observation-of-behavior version of science.

    I’m also not sure what you mean by being vs. doing. Participant Observation has always struck me as the perfect label for the fact that the distinction is never clear– you can sit for weeks waiting for something to happen, and when it does, is that the only part that counts as fieldwork?

  2. Rex and savage minders,

    So I have a question: would it be less confusing if I just removed the early invocation of Rheinberger from the essay? The point was to make a connection to the contemporary conversation between anthropology and STS (which is where the essay ends, and its mate in the Nov. issue picks up).

    May I remind that Anthropology as Cultural Critique’s subtitle was “an Experimental Moment…” and as we explained there (and again in the 2nd edition), our use of “experimental” drew on the two meanings of experimental as it has been used in anthropology from Malinowski on (one aesthetic, writing, experiments, of which “being there” is one form Malinowski, Firth, et. al. perfected [see how Malinwoski’s students remembered reading Argonauts for the first time in the collection edited by Leach called Man and Culture; the other the scientific aspirations of anthropology. As a matter of principle I insist that anthropology has feet both in the humanities and social sicences, and that the social sciences do have scientific aspirations even if anthropology even in its social scientific mode tends to be anti-reductionist. (As STS has been teaching us [and many scientists all along], the “hard” or “natural” sciences also do not operate as mechanically as they sometimes publicize themselves as doing.) The Rheinberger reference is, as Chris Kelty rightly points out, to a notion of experiment generating surprises (see also Avital Ronell’s Test Drive for an expansive reading). In anthropology this can be done by varying the fieldwork as well as by varying the ways in which data are analyzed and presented.

    I quite agree that a lot of so-called ethnography is published while still quite thin. Any anthropologist (and many others) should be able to read for the ethnography and discount those that are data thin or that are analytically weak because based on thin ethnography. As Chris rightly points out, this does not mean you have to sit in one place for 18 months or more (though it always helps to have some temporal depth). Indeed the questions that we are addressing today require a more experimental approach to figuring out how to frame good research questions, identify relevant data, find analytic tools, and write, film, or savage mind the results or invitations to continuing dialogue across points of view, interest, etc. Much more complicated than in Malinowski’s day.

  3. Good post, Rex. I confess I’m always baffled by our obsession with _lab_ science when ethnographic fieldwork has more obvious affinities to natural science–ecology, ornithology, geology, ethology, etc. Field scientists have an explicit research program and methodology, yet all but the most dim-witted remain alert to unexpected events, perhaps only tangentially related to their ostensible project, that invite a reframing of questions and perhaps a profound reinterpretation of reality. Anthropologists have the advantage (and burden) that our subjects can respond to systematic queries by saying, “Gee, that’s a dumb question. You should be asking _this_.” That’s the primary reason why I became an anthropologist rather than a sociologist–the latter a discipline where, as C. Wright Mills pointed out long ago, “abtracted empiricism” still holds sway except among a small group of ethnographically and qualitatively oriented scholars.

    Eons ago, when I was still playing with ethnoscientific elicitation methods, I discovered that they were often more useful–and a heck of a lot more interesting–for the off-topic observations they generated than the official “data” they produced. They tended to take interlocutors in directions that might not have been encountered via non-directed observation, with results that sometimes surprised even my subjects.

  4. A supplementary thought about Rheinberger. One of the intriguing things about his work is the space of play between expansiveness in the use of Derrida (the claim for instance that we now write biologies rather than merely discovering them, meaning among other things, that we create biological entities and combinations that never previously existed), and on the other hand his narrow account of how the rat-liver system at MGH generated that surprises that led to a new program of charting out the biochemical cascades involved in protein synthesis. His new book Epistemologie des Konreten pursues the generalizing side (its epistemological implications) with accounts of Husserl, Berson, and others. A recent workshop at Irvine, organized by Kaushik Sunder Ranjan, with Rheinberger as the honored guest, attempted to explore this “space of play” as it might appear in the work of a variety of anthropologists of science and technology studies along with other interdisciplinary scholars in comparative literature, philosophy, and computer science. None of this seems foreign to the experimental side of anthropology at its best.

    Michael Brown is correct, anthropology is a field science, not a lab science. On the other hand, part of our field science now is lab sciences, both how they operate internally and how their various products (including power, prestige, authority, international linkages, epistemologties, medical technologies, pharmaceuticals, etc.) operate externally.

  5. Michael, I hope my comment wasn’t taken as an implied criticism of the ethnographic study of bench science, which I find utterly compelling, especially when the research moves beyond the logocentric study of scientific discourse to include scientific praxis and the interests that bear upon it. It’s an important social world that merits as much attention as any other–perhaps more.

  6. Thanks for these great comments all. This post is really intended to be a meditation about what fieldwork is, and how it might compare to lab work. The rhetorical hook at the beginning was the genealogy of my thought in this matter — comparing Fischer and Rheinberger. So its not meant to be a critique of Fischer’s article per se.

    The thing about lab work — to generalize on a topic that is not my speciality — is that the surprises which Chris so rightly mentions only come out of the system. As Brown points out, you’ve got to have a method in place to begin generating material that always escapes the method. And that of course is the interesting stuff.

    I think we can all take it as read that we are on the same page re: naive realism etc. etc. But what strikes me about the difference between ‘being in the field’ and ‘doing fieldwork’ (this is Harry Wolcott’s distinction, btw) is that, as it were, the later doesn’t really involve any sort of ‘experimental system’ as it were.

    Let me be speculative: over the past twenty years one unintended side effect of rethinking our notions of fieldwork and objectivity has been a withering of attempts to create ‘experimental systems’ in the field. The result is that not only our genre expectation, but our lived experience blur more and more with journalists, travel writers, parish priests with an interest in folkore etc. Except that typically they write better than us.

    I like talking to parish priests with an interest in folklore! But if this extremely speculative reconstruction of the state of the field is correct, then it seems to me that a culturally specific way of knowing the human that is worthwhile is changing in a way I’d prefer it not to change.

    But that really is out there.

  7. Rex,

    A lot of us were trained with “systems” of various sorts for fieldwork. Remember Malinowski sent his students off with institutional checklists (as well as crayons and rolls of shelf paper and instructions to buy tennis shoes so the water could run out when you cross streams — viz. Return to Laughter). Learning to do extensive genealogies and taking field linguistics were important systematic techniques of inquiry I was trained with, as were designing pilot questionaires (with which I mapped bazaars in several cities). Structuralist tactics of analyzing Galenic derived medical systems from Asia to Latin America were yet another of these systematic probes. This is why I still insist (probably with insufficient authoritarianism) my students think about data collection and how one does it relatively systematically even if we are now studying other arenas of life. And when various people throw aspersions on “postivist” methods, I reply that positive knowledge is one of the things we are after, and ask them to specify which of the many definitions of positivism they are really imagining and why that particularly one is relevant and who espouses it (usually no one). Often as you say what escapes the system is what is interesting, or in Rheinberger’s language, provides the surprise that (to vary the analogy) in information theory is called the meaningful bits. What’s discouraging, and this is perhaps what you are getting at, are those who get attention for writing things without doing the empirical work. (In science that’s called “philosophy,” in the newspapers it is called “punditry,” but I digress . . .)

Comments are closed.