IRBs and the ethnography problem: demarcating ‘research’, locating allies

The point of the November 2006 AE Forum I put together, “Anxious borders between life and work in an age of bureaucratic ethics regulation” (follow the link in Tom Strong’s post introducing me), was to identify the distinctive features of ethnography and of the IRB system (so-called ‘human subjects committees’) that set them up for conflict, and to explore the implications.

Demarcating ‘research’:

A key distinctive feature of ethnographic research is the fact that ethnographers–whether they work in Highland Papua New Guinea or New Jersey–typically embed themselves with their interlocutors: that is, both ideally and often enough in practice, ethnographers live where they work. Anthropologists have long recognized that significant ethical dilemmas derive from the fact that research (ethnographic ‘work’) isn’t demarcated from not-research (the rest of the ethnographer’s ‘life’). But it is a fresh headache on account of intensified IRB oversight. This is because the federal human subjects research regulations (what IRBs are set up to enforce) presume a clear distinction: ‘research’ is, after all, the regulatory object. (I’ll discuss this point in another posting.)

To exaggerate this problem so as to make it more visible, the AE Forum focused discussion on unfunded fieldwork. While ethnographic sociology is typically unfunded–a surprise to most of the anthropologists, I suspect–anthropologists who work “at home” (wherever that may be) also often do so without research grants. This may be especially true of research past the dissertation phase: I’m interested in Savage Minds readers’ experiences. We also focused on the necessarily open-ended, exploratory character of ethnographic discovery practices. While this is also obviously not a new insight, viewed as part of the IRB ‘research’ demarcation problem it helps clarify the distinctiveness of ethnographic work. This is because the federal human subjects research regulations are meant to be applied before ‘research’ begins. The rules are designed for ‘research’ understood as a distinct event, not as an emergent process (as is oftent the case in ethnography, and especially when it’s done at home).

Locating allies:

The focus of the AE Forum was on ethnography in the broadest sense of the term–that is, the research style that anthropology shares with several other fields, notably sociology. This is why I invited Jack Katz (a well known ethnographic sociologist from UCLA who has been doing important critical work in IRBs) to be part of this project.

There’s an important strategic point here. Anthropologists are in the habit of telling themselves that they are the inventors of ethnographic fieldwork and that others who adopt the approach are derivative or in some sense inauthentic. This is just not true. Fieldwork–participant observation particularly–was a co-creation of sociologists and anthropologists, whose methodological histories overlapped heavily during the 19th and early 20th centuries. (I’m happy to say more about this, if anyone is interested; for example, look at Rosalie Wax’s fascinating book, Doing Fieldwork.) While there’s no doubt that fieldwork is positioned differently in present-day anthropology (where it’s the default approach) and sociology (where it’s a marginal or minority approach), I think that recognizing this common history may help anthropologists cultivate allies in their IRB struggles. It’s in our strategic interest to make common cause with ethnographers of all sorts both locally (in our respective institutions, on our local IRBs) and nationally, given the expansion and intensification of ethics regulation.


I’m Professor of Anthropology at Princeton University. I’ve blogged here about the disconnection between ethnographic field ethics and ethics regulation compliance, and advocated hither and ton for humanistic social studies in regulatory ethics contexts. Otherwise, I’ve done fieldwork in Highland Papua New Guinea (sociality, gender, exchange, historicity) and on knowledge cultures in the US academy (anthropology, historiography, sociology, experimental psychology).

15 thoughts on “IRBs and the ethnography problem: demarcating ‘research’, locating allies

  1. “Fieldwork—participant observation particularly—was a co-creation of sociologists and anthropologists, whose methodological histories overlapped heavily during the 19th and early 20th centuries.(I’m happy to say more about this, if anyone is interested..)”

    I’d be happy to read more indeed. I have wondered already earlier why qualitative sociology so seldom gets cited in the context of anthropology’s attempts to reconceptualize its subject to turn the discipline attention on cultural phenomena that are close-by (“at home”) and constitute “the own culture”. Thinking of Robert E. Park for example there seems to be not only the anthro buzzword sociology to be a barrier in the way but also the one of journalism that may prevent anthropologists from citing those early qualitative urbanologists.
    But maybe I have gained a wrong impression from my reading and talking to anthropologists so far..?

  2. Your post made me wonder about multidisciplinary projects and IRB approval. One the one hand, we can look for allies in sister disciplines in order to be able to ‘push back’ against a narrowing of the idea of ‘research’ in the context of bureaucratic ethics regulation; in that case, it might be an advantage to retain disciplinary difference in order to multiply discrete positions for supporting ethnographic research.

    On the other, allies may cooperate across disciplines within a single project. (My own experience in this regard pertains to applied research related to HIV risk and prevention.) In those instances, ‘ethnographic methods’ may be positioned vis-a-vis other approaches within a single project seeking IRB approval. I am wondering how the management of ethnographic research in that context differs from the management of ethnographic research in stand alone anthropology projects.

  3. Does anyone know of IRB type restrictions outside of academia and government-funded research?

    I ask because to the best of my knowledge commercial market research firms don’t go through the contortions that IRB imposes on academic research. This could, however, reflect nothing more than my own ignorance.

    I ask, too, because, in a politically conniving, brainstorming kind of way, it occurs to me that a movement to impose similar restrictions on market researchers would lead to a massive industry reaction and support for loosening restrictions on a scale that academics are unlikely to achieve by themselves.

  4. Rena– Thanks for the interesting discussion and links. Your discussion of the IRB-related difficulties posed by the lack of clear boundaries between `research’ and `not research’ in much ethnographic research made me wonder to what degree these problems are regulatory or legal in origin, and to what degree they arise from preconceptions about the nature of research, stemming from the particular `cultures of research’ in which the IRB members participate.

    I ask because I just had to update my `ethics training’ for renewal of an IRB approval, and so had the opportunity to slog through lots of regulatory prose on IRB-related matters. One of the things I came across was a definition of `research’, according to the `Common Rule’, as “systematic investigation … designed to contribute to or develop generalizable knowledge.” Pretty bare-bones, and there does not seem to be much in that definition that necessarily says anything about the demarcation of `research’ from `not research’. It seems that this definition could easily accommodate ethnography if IRBs so chose.

    Having been party to a number of cross-disciplinary discussions regarding methodology, I wonder if the problem that ethnography poses for IRB-types is not so much the absence of a clear demarcation between `research’ and `not research’, but rather the opportunistic nature of much ethnography, which may be perceived as lack of *systematicity*. This perceived lack of systematicity in turn makes regulatory types nervous, not for any well-founded regulatory or legal reasons, but simply because its does not square with the way they do *their* research.

    Maybe this distinction is not important, but it seems to me that different strategies are called for if the basic problem is regulatory and legal or if it stems from perceptions fueled by cross-disciplinary differences in research practices.

  5. I wonder whether there are resonances between IRB attempts to control and delineate the messy flux of social interaction and intellectual propety’s attempts to control and delineate the circulation of knowledge and information?

  6. I think it would be interesting to look at the opposite end of the spectrum. Instead of focusing on those examples which are furthest from the framework within which IRB was intended to function, look at those examples which are supposed to be paradigmatic.

    My guess is that in a controlled medical experiment with human subjects the issues are just as difficult and just as far removed from the bureaucratic requirements of the IRB as they are when doing fieldwork among friends in your own backyard.

  7. A couple of things struck me as a researcher and IRB chair reading the post, related posts and comments. Let me try to articulate them:

    1. I didn’t realize that people working in anthropology and sociology have an animus towards IRBs to the degree that is expressed here (and reflected in hunkered-down, military metaphors like ‘allies’). I am more familiar with the crude ‘us versus them’ mentality of biomedical researchers, and perhaps naively thought that people in the social sciences would be more sympathetic towards the aim (if not always the means) of protecting research participants in the pursuit of knowledge. I guess not! Here it sounds like IRBs are regarded as a pointless bureaucratic nuisance, with no redeeming qualities. This does not ring true to my experience, though, where researchers are often happy to get a second opinion on their research projects that can (in the best scenario) help the ethics and the science: as long as the review does not take too long!

    2. I was fascinated by the opinion that IRBs have little sensitivity to the nature of ANY research, biomedical or otherwise, because the former is (supposedly) dynamic and emergent, and the latter is (supposedly) a dead and fixed bunch of rules. One thought I had is that the same lack of sensitivity can be said of a lot of stakeholders in the research enterprise, including funders. Don’t researchers always feel misunderstood (lord knows I do)? Another thought is that the only way that an IRB could be truly sensitive to the demands of any particular study would be to join the research team as an ’embedded observer’, something both impractical and (for researcher and IRB member alike) undesirable.

    3. The above point brings me to a series of questions. There is little discussion here of what IRBs should do to improve themselves in the eyes of social scientists: so what would make social science and IRBs better bedfellows? Or I am missing the point, and the real issue is that the entire project of ethics review is misguided? And that as far as ethics is concerned, social scientists do not need any guidance or regulation, and can be trusted (blindly) to do the right thing? Some of the comments seem to be arguing in this vein. For example, take the argument that ethnographic research is not research (as defined by the Common Rule), therefore cannot be ‘human subjects research’, and therefore should not be reviewed by an IRB. Or take the arugment that if other realms of inquiry (journalism, pharmaceutical research, etc.) do not require ethical review, then social science research should not be reviewed either. Bottom line: no ethical review needed or desired. Is this really where you want to go?

  8. In reply to Stuart, above, I have two answers to his questions:

    1. “What would make social science and IRBs better bedfellows?”

    I suggest making IRB review voluntary. If IRBs are good at their jobs, and “researchers are often happy to get a second opinion on their research projects,” IRBs will not need federal or university rules requiring researchers to consult them.

    2. “Am [I] missing the point, and the real issue is that the entire project of ethics review is misguided? And that as far as ethics is concerned, social scientists do not need any guidance or regulation, and can be trusted (blindly) to do the right thing?”

    Stuart conflates the concepts of IRB review and ethical review. Ethical review could take many forms, including departmental ethics committees, national networks, and more attention to ethics by journals and publishers. Review by local IRBs, which generally lack expertise in the issues raised by a particular project, is a particularly poor form of ethical review. Mandatory review by local IRBs, which absolves IRBs of any need to persuade researchers, is a catastrophic form of ethical review.

  9. Thanks for your comments.

    In regard to (1), I think that making IRB reviews voluntary would just mean researchers of all stripes opting out of having their research reviewed. Most would probably do a rapid informal ‘cost/benefit’ analysis in their heads and decide that, even if being having their research reviewed brought some benefits, in the end they would rather have less paperwork and get out into the field quicker. In this way, the proposition ‘make IRB review voluntary’ is really code for: do away with IRBs.

    In regard to (2), I am not sure what is meant by my ‘conflating’ IRB review and ethical review. On the one hand, I see IRB review as one form of ethical review, perhaps dysfunctional and apparently unloved, but a form of ethical review nevertheless. (Even if the judgments of an IRB were mindless applications of the Common Rule, the latter is shot through with ethical considerations.) On the other hand, I agree that IRBs exhaust all possible forms of ethical review of research. I also agree these alternative forms should be explored. But I am skeptical that national networks (some developing countries have these already) and departmental ethics committees (they exist all over the USA already) would be immune to the kinds of criticism leveled at IRBs today, and I wonder if having journal editors review protocols would be a workable alternative, when they are at an even farther remove from research activities than IRB members are.

    Question linking (1) and (2): would the alterative forms of ethical review also be voluntary? If so, why should anyone take them seriously? Again, the bottom line seems to be: no ethical review needed or desired. I don’t want to sound dramatic, but is this a call back to ‘pre-Tuskegee’ days of scientific research?

  10. “On the other hand, I agree that IRBs exhaust all possible forms of ethical review of research.”

    I meant to say that they DO NOT exhaust all possible forms of ethical review of research. Sorry about that!

  11. Stuart writes that he ‘naively thought that people in the social sciences would be more sympathetic towards the aim (if not always the means) of protecting research participants in the pursuit of knowledge. I guess not!’

    This strikes me as a big distortion of what Rena and others are saying here, which is in fact that the bureaucratic regulation of ethical behavior in the context of ethnographic research may work at cross purposes with its ostensible aims. CKelty very nicely summarized this in his quick take on consent/copyright overlap. Certainly, the point is not to avoid the subject of ethics in general, but to be able to make ethnographic sensibility heard in the context of its ‘management.’

    You may know that in fact anthropologists are almost paralyzingly concerned about the ethics of our research practices and even the ‘morality’ of our whole discipline. The idea that we simply want to avoid the subject strikes me as preposterous.

  12. In reply to Stuart (and thanks for your comments):

    1. You write that most researchers “would rather have less paperwork and get out into the field quicker.”

    What happened to all those researchers who were happy to get a second opinion? Wouldn’t they be clever enough to return, and advise their colleagues and students to do so? Or perhaps they aren’t that happy after all?

    A voluntary IRB would be in the position of a university librarian or pedagogy expert. They only get customers when they do their job well.

    2. You write that “departmental ethics committees . . . exist all over the USA already.”

    I would like to learn more about this. Why can’t these committees perform the tasks now delegated to IRBs? This is hinted at in the 2006 AAUP report, “Research on Human Subjects: Academic Freedom and the Institutional Review Board.”

    3. “Is this a call back to ‘pre-Tuskegee’ days of scientific research?”

    What is the Tuskegee of anthropology? What is the Tuskegee of oral history? Inapplicable analogies from medical research are at the root of this problem.

  13. Strong and Zachary,

    Many thanks for your comments — I am learning much about the anthropological take on these issues, and I am grateful for it.

    In regard to Strong’s remarks, I could cut and paste the portions of text that led me to my original impression that IRB’s were seen as a alien imposition of a regulatory regime insensitive to the nature of biomedical or behavioral research. But I don’t want to flog a dead horse. If, as C Kelty’s post suggests, anthropologists are working creatively to try to make regulations make sense in the field, all the while driven by ethical concerns about those involved in their research, what more could/should an IRB want?

    In response to Zachary’s points:

    1. Like I said, many researchers are happy to get a second opinion via the IRB, but there are many other things that make researchers happy, and getting out into the field quicker and dealing with less paperwork are two of them. Constructive criticisms and comments are often welcomed — especially by grad students neglected by their own advisors. But I didn’t say they were ecstatic, and driven to IRBs like bees to pollen.

    2. I don’t have any firm figures at hand, but in my experience, many academic institutions are a mixed bag in this regard. They have some IRBs and a number of departmental ethics committees, with all sorts of relationships with one another. (It has to do with the haphazard growth of ethics review committees over the last 20 years.) One relation is a screening relation. In our institution, the departmental ethics committee (say, journalism or sociology) initially reviews a proposal, then it goes to one of the IRBs (say, behavioral). The double review doesn’t seem useful, and there is a struggle over who should go: the departmental review or the IRB. You seem to favor the departmental review over the IRB, so I’d be interested to hear your opinion on that.

    3. As far as the Tuskegee example goes, I was talking about the ‘pre-regulatory’ days before Tuskegee came to light, followed by the Belmont Report and the rise of IRBs. To ask ‘What is the Tuskegee of anthropology or oral history?’ is to ask other questions: have there been abuses to research participants in anthropology or oral history similar in magnitude as Tuskegee? I think the answer is ‘no’. Does anthropology or oral history pose no significant risks to research participants, such that no more regulatory requirements are needed than when Malinowski was doing his work? I think the answer to that is ‘no’, too, especially when anthropology is getting more entangled with health research (I am thinking of genetic research on native populations.)

  14. If, as C Kelty’s post suggests, anthropologists are working creatively to try to make regulations make sense in the field, all the while driven by ethical concerns about those involved in their research, what more could/should an IRB want?

    I assure you that, for the most part, historians are every bit as hostile to the regulations as you imagine. The regulations make little sense, and OHRP’s guidance often makes none. Read Carome’s three examples at Do you find significant ethical or legal distinctions in the three hypothetical projects he describes?

    The current regime allows pre-screening by a member of the IRB. If that person rules the project to be exempt from review, the process ends there. It sounds as though the departmental committees you describe do not enjoy that privilege, leading to the useless double review. Giving departments the power to exempt projects–by individual review or by crafting policies that describe which projects require review–would, in contrast, help a great deal.

  15. The exchange between Stuart, Zach and Tom is interesting. I have a couple of comments.

    First, Stuart needs to recognize that, as my very first post suggested, what I’m writing here presumes some familiarity with at least my main contribution to the Forum published in the November issue of American Ethnologist. While I don’t expect most people actually to read it, I do expect folks to be aware of that background. Which is to say: none of my posts will be complete, complex positions. For something more like that, please do read the published article.

    Second, I’m to blame here for assuming that most of the folks who actively participate in this conversation are anthropologists. Silly assumption! If I have one point to make about IRB discussions generally, it’s that disciplines are diverse in certain key respects bearing directly on the central mission of ethics oversight work.

    That is, the human sciences — by which I mean all fields in which humans are studying humans — have all of the same ethical issues that the natural sciences do. In all fields, relationships among researchers can be exemplary or go awry (plagiarism, abuse of power). However, the human sciences add to this the fact that the research itself involves social relations between the researcher and the researched.

    The key point is that, from biomedicine to literary studies, the research relation differs among fields in ways that are intimately intertwined with their respective “knowledges”. While their respective ethical conventions do overlap (as I tried to sketch in the AE paper), these fields’ ethical configurations are also mutually distinctive.

    In my view, perhaps the key challenge that researchers and IRBs face is how to deal with the diversity of disciplinary ethics cultures (if I can use that metaphor). Presently, the tendency is to assume that we’re all speaking the same language, with some of us more fluent than others. In my view, that’s not the case.

    Ok, so I didn’t realize until today that Zach isn’t an anthropologist: he’s an historian with quite a bit of expertise concerning IRB matters, and a presense in the AHA’s recent oral history discussions. This may have some bearing on Stuart’s sense of who he’s hearing from and what he’s learning on this site.

    Along the same lines: really to understand this conversation, I need to know where Stuart is coming from. Can you tell us something about your own field of study? It also may be relevant to know whether you chair an IRB that reviews medical protocols, or social/behavioral science protocols, or some combination of the two. Those of us who have been studying how IRBs work have been learning that differently configured boards may have significantly different experiences reviewing what I’d call “outlier” methodologies.

    My questions aren’t meant to be personal: the discipline question is, in my opinion, absolutely central in any thinking about research ethics oversight and its various possible alternatives.

Comments are closed.