ARC: By “critical rectification” we mean: we can discuss our concepts and findings, and we can have an argument in which — through the application of shared standards and understandings — we can figure out that some formulation/idea/distinction was wrong, in need of greater precision, or in need of reframing in relationship to other contexts and distinctions.

Here is an extended example that might provide a sense of how this happens in terms of the “internal workings” of the collaboratory, per your question above. There are also some citations to give you a sense of how this process has related to the production of more traditional products of research (published articles and working papers):

When we began our current project, we defined security, following Niklas Luhmann, as the transformation of uncertain dangers into calculable risks (Collier, Lakoff, Rabinow, 2004). Thus, in our initial formulations (in our NSF grant proposal, for example), we proposed to study security “initiatives” that sought to transform emerging biothreats into calculable risks.

In his initial fieldwork, Lakoff began to examine one such initiative: a collaboration between the Cold-War era think task RAND and a firm that specializes in modeling catastrophic events, Risk Management Solutions. The goal of the collaboration was to integrate terrorism expertise with probabilistic models of catastrophic events in order to make possible a market in terrorism insurance. The case presented an interesting twist on our initial formulations. Lakoff was struck by how this collaboration employed scenarios as non-quantitative techniques for approaching security threats whose likelihood and possible impact could not be calculated. He observed that these security techniques of “preparedness” could be distinguished from Luhmann’s understanding of security as risk-management, in that they did not necessarily involve quantification (Lakoff 2005). This distinction — of risk versus preparedness — seemed consistent with a large theoretical literature on risk (that includes central contributions by Beck and Ewald), which claims that contemporary society must deal with threats that are incalculable. But Lakoff’s initial work on scenarios suggested a direction of inquiry one might pursue to fill in the lacunae in this theoretical literature, identifying the kinds of techniques used to manage incalculable threats.

This initial distinction seemed worth trying out more broadly as an element in the toolkit of our collaboration around security. Thus, we began to look collectively at some documents in which “preparedness” was articulated as a normative rationality for dealing with security problems — including a DHS national preparedness plan that had just been released (these conversations were ongoing in spring and early summer 2005). An argument arose in the collaboratory. On the one hand, the DHS document relied heavily on a set of scenarios of catastrophic events, and thus did not seem to be engaged in “risk-based” calculation. On the other hand, it drew on techniques of quantification and calculation, for example, in risk-based budgetary distribution formulae that closely resembled similar formulae that Collier studied in his work on social welfare in Russia (Collier 2004). In other words, “calculation” and techniques of quantification seemed to be an important part of the DHS approach. And this framework for “preparedness” was quite explicitly engaged in risk management, apparently confusing our initial distinction between risk and preparedness. A further specification and conceptual refinement, we ultimately agreed, was required in order to adequately characterize how this form of security functioned.

Here is where an empirical “sounding” (discussed below) played an unexpected role. We have a research assistant, Lyle Fearnley, who has been doing terrific work — through interviews and documentary research — on “syndromic” surveillance systems for detecting outbreaks of infectious disease. Government agencies charged with “security” questions have become very interested in the ability of such systems to detect a health “event” in a population in real time, particularly given the increased concern with the threat of a bioterror attack since 9/11 and the anthrax letters.

As we were engaged in this discussion of preparedness, Lyle was pursuing a historical chunk of this research (Fearnley 2005), which focused on the moment of “epidemiological transition” in the United States. One feature of this transition was a shift in the focus from diseases regularly occurring in a population to disease “events,” that is, diseases whose dynamics were not known. Experts thus identified a need for real-time identification of health events in a population. Their efforts are one important genealogical precursor to contemporary syndromic surveillance (and, in a certain way, to contemporary ‘preparedness’).

Fearnley reported that first order actors, in this historical context (immediately after World War II), identified a distinction between the “archival” knowledge required for the management of epidemic disease and the real-time knowledge of populations required to deal with disease events. This distinction helped to clarify the dispute in the collaboratory, and pushed the process of conceptualization forward. We agreed that the initial distinction — between preparedness and “risk” — was not exactly the right distinction, and that the distinction between calculability and non-calculability advanced in the “risk” literature was also not quite the right distinction. Rather, the salient contrast was between a certain risk technology — insurance — and preparedness. Insurance is based on actuarial analysis that draws on archival knowledge of populations. Preparedness, by contrast, draws on techniques of what we later called “imaginative enactment” (Collier and Lakoff, 2006) to deal with low-probability, high-consequence events about which no archival knowledge exists.

The payoff of this clarification can be identified on a couple different levels. On the one hand, it led to a critical intervention into discussions of “risk society” by people like Beck and Giddens (and responses to this work by scholars of “governmentality”). Given our research, we see two very fundamental problems in this literature. First, they have identified as important the emergence of events to which insurantial mechanisms don’t apply. But they have not yet found a way to investigate what comes next, or what techniques are used to manage such events. Second, we think that they have the basic distinction wrong. It is not a question of “calculability” per se but of the kind of calculability, the techniques of quantification, and the purposes for which they are used. It is not a question of “risk” but the techniques of risk assessment, how they change in relation to different sorts of objects, and how they are articulated in certain apparatuses. We now have a sophisticated vocabulary for thinking about these things. The theoretical literature does not.

On the other hand, we have been able to confirm and build on this distinction in a number of other areas — for example, in the work that one of our research associates, Dale Rose, has been doing on the CDC and smallpox vaccination strategies. It is now also informing PR’s research into how life scientists and government regulators are proposing to regulate synthetic biology, as well as further historical work that SJC and LA are doing on preparedness and the political logics with which it has been associated.

ARC: This distinction between “fieldwork” and “investigation” is not clear to us, and we would be interested to know when something counts as “fieldwork” — whether, for example, the activity of a political scientist who spends two years in the field counts. Certainly, the term is used in that context. Or, to take a case from the collaboratory: Fearnley’s work has involved looking at documents, attending a couple conferences, and talking to a handful of experts. He has produced penetrating conceptual insight. And indeed, all of us are combining things like intensive discussions with experts, attending security-related events, and documentary research. Does this count as fieldwork?

A better approach to this question, however, might be to take a step back. An important impetus of our work is to think more about the status of “fieldwork” or “ethnography” in relation to the problem of methodology in anthropological discussions. We have been saying for a while (see Collier and Lakoff, 2000) that anthropologists tend — mistakenly — to limit discussion of “method” to discussions of fieldwork and writing. In our view, fieldwork is a technique — or, perhaps, a set of techniques — but not a methodology. There may have been previous configurations of the discipline in which it is was the central technique of an anthropology that was committed to a rather holistic version of the culture concept, but it seems those days are gone.

We would prefer to deconstitute the idea of “fieldwork” and to ask what it is, more concretely, that is being talked about. It seems that in anthropology fieldwork can refer to interviews, observations of (and participation in) meetings, informal discussions; and also, close reading of documents produced by actors. Our main point is that it is good to reflect on these techniques, both individually and in their interaction, but that is not the same as reflecting on “methodology” — which concerns, among other things, how these techniques of data-gathering interact with concept formation and the establishment of collective standards, norms, and conventions to yield meaningful claims, and meaningful progress in thinking. So one way to put this point is that we are not focused on fieldwork per se, but on the process of interaction between concept-work and fieldwork. So the collaboratory has a number of people doing “fieldwork” of various kinds in various areas — syndromic surveillance, vaccination, synthetic biology, civil defense, strategic bombing, and so on. The question, then, is how to generate a process in which this collective work feeds into broader conceptual issues, and how, in turn, these conceptual issues generate specific questions to be approached through fieldwork.

ARC: Both of these are helpful points of reference, and seem useful to us. We agree that developing techniques to generate unexpected findings is important (although we think that it is in need of further specification and can’t be claimed uniquely for anthropology; after all, Rheinberger is talking about how natural scientists use research design as systems for generating surprise). What we think might be missing in both cases is what is often missing from discussions of method in anthropology: analysis of orientation, of significance, and of problem formation. These are the questions posed above: “How does one decide to do fieldwork?” “How are significant problems identified?”

So Strathern argues that an important aspect of fieldwork is its indeterminacy — that one goes into the field and one collects more than one “needs,” because, presumably, one does not yet know what is significant. This seems right, and resonates with our process in the current project. For example: a year ago, we had no way of knowing that we would be working on theories of strategic bombing or the history of exercises in the military! And the only way to “discover” that these things were relevant to our project was to go into the field with at least some indeterminacy, and a sense that there was a period of thrashing about without knowing what is going on that had to be part of the process. But we would want to add that one does not do this in an unstructured way, or in a way that is not guided by an understanding of what significance is. Where does one go? Whom does one talk to and what about? What books does one read? To describe this simply in terms of a method that collects more than it needs, that emphasizes indeterminacy is just mystification. There are reasons that we go one place rather than another: we are interested in bio-power, we are interested in expertise, we are interested in the interaction between security and social welfare, etc. So we think that the na ¯ve “entry into the field” stories are just bad accounts of what anthropologists actually do. So we want to describe how a process of relatively open-ended searching might be linked to a rigorous process of concept-formation.

We would say the same thing about Rheinberger’s understanding of experimental systems in the sciences as techniques for generating surprise. Again: on one level this seems right. As PR has often argued, the point of studying emergence is that one does not know what one is going to find. But this way of being “experimental” is not avant-gardist. It does not simply try to undermine existing norms or to create “surprise” for its own sake. A scientist needs norms, conventions, and shared understandings about interesting or significant problems to make an experimental system meaningful. These norms, conventions and understandings will be different for a natural scientist and for an anthropologist, and we are suggesting that more reflection on what they are in anthropology is needed.

ARC: We want to be provocative, at least in the sense that we want anthropologists to be less complacent, and more critical and reflexive, about a range of taken-for-granted assumptions about method that have taken shape over the past couple decades. But we don’t think that this is a throw-back. It seems to us that these are quite contemporary debates that are being thrashed out in anthropology and elsewhere.

In asking questions like “how might move thought forward” we are not talking about a na ¯ve objectivism. Rather, it is a pragmatist epistemology that is consistent with Dewey, Rorty, etc. It acknowledges, as Paul has written (2004) that any reasonably coherent theory of scientific knowledge acknowledges that it is based on concepts that are constructed. But this does not mean that one does not seek to advance understanding, successfully and productively reframe problems, make defensible claims about the world, or clarify concepts in a fashion that can be subjected to critical rectification through impersonal norms.

GM: ARE REFINED CONCEPTS THE ACTUAL FINDINGS IN/FROM ‘THE FIELD’ OR ARE THESE FINDINGS SOMETHING ELSE? AGAIN, FIELDWORK SEEMS TO BE DISTINCT (AND INDIVIDUAL) BUT I AM NOT SURE. ALSO I AM NOT SURE WHAT THE FIELDWORK IS? INTERVIEWS ONLY? BONDING WITH PARTICULAR INFORMANTS/ENTERING THEIR WORLDS IN A SUSTAINED WAY?

answer 5: ARC

ARC: This is a helpful question. As you know, a major part of Paul’s approach (at least since French Modern, but really it is very central to Reflections on Fieldwork) has been the study of observers “in the field” who are also engaged in concept work. And this is definitely part of our past and current projects. We are interested in studying experts, or, better, what Hacking calls “styles of reasoning” that experts employ, and in understanding how these styles of reasoning give shape to institutional responses that are part of new problematizations.

But we should emphasize that — unlike those who have been trying to draw an analogy between anthropological knowledge and the knowledge of the experts we study — we don’t think that all forms of knowledge are structured in the same way. It seems to us that there is a clear distinction between the kind of concept work that we do and the kind of concept work done by those in the field — the “first-order” observers.

Let us give an example. We have been following (in the sense of reading the writing of, and listening to the talks of) a guy named Stephen Flynn. Flynn is one of our exemplars of a “vital systems security” expert. He is defining a very distinctive conceptual and practical position in contemporary debates around how to deal with security problems. So, for example, he has been engaged in a debate with another expert — one James Carafano, who is, in our terms, a “sovereign state security” guy — over the problem of port security. Flynn wants to see it as a vital systems security issue (we need to have systems assurance technologies that allow us to identify rapidly the source of a security breach in the international shipping system so that “auto-immune response” does not shut down the system). Carafano thinks that shipping is a highly unlikely channel for terrorist attack, and that it would be a waste of billions of dollars (and an unacceptable burden on international commerce) to install such a system. Much better, he argues, to find the bad guys and kill them — a classic case of what we call a logic of “interdiction.” So these “first order” observers argue, try to convince politicians, influence spending, marshal evidence, etc. Those are the stakes of their game.

For us, something very different is at stake. We are interested in the broader rationality that they are working with, and in arriving at a conceptualization of it that points toward critical diagnosis: i.e. that suggests where it comes from, and what is at stake in its current formations. We are open to the possibility that a productive exchange could develop with this kind of conceptually-oriented first order observers, but we are quite clear that the aims and norms of their knowledge practices are distinct from ours.

ARC: Per the above, experts in the field are asking themselves what the threat is, and how to best approach it; we are analyzing how they do this — how they pose problems of security, how security becomes a certain kind of problem. Of course, this distinction is not ours. This is what Foucault said he was doing in all his methodological writing (how did madness become a certain kind of problem? How did criminality become a problem? How did sexuality become a problem?), although his description of the object of analysis shifted over time (episteme, discourse, apparatus).

ARC: This is discussed a bit above: It seems to us that #1 subscribes to a kind of na ¯ve empiricism that suggests everything one needs for understanding, or diacritical analysis, or inquiry, or whatever, is waiting for you in the field. All you have to do is “be there” with the proper “anthropological” ethos of indeterminacy, interest in finding surprise etc. As noted above, this seems to us both methodologically unsupportable and a bad account, in any case, of what actually happens in anthropological inquiry.

We gave one example — above — of how we think that a dynamic process of conceptualization might relate to “fieldwork.” Perhaps it would be helpful to give another that indicates the costs of the first approach noted above:

There has, as you know, been a huge amount of writing in the “critical social sciences” — including in parts of anthropology — about security in recent years, especially after 9/11. In reading this literature, it’s astonishing to what extent it frames the basic problem of security today as one of “militarization” of the civilian sphere. If you want an example, check out something written recently by Ann Stoler (an article in the Radical History Review, for instance), who is working on some of the same documents produced by DHS that we are working on (in other words, the “field” is the same). For her, the story of DHS is a story about militarization. What is striking from our point of view is that she has not even posed the question: “what is security?” One of the advantages of making this an object for critical reflection before diving into a research site is that we now understand that security does not always mean guys with pressed uniforms and shiny boots. What is more, the history of preparedness in the United States specifically includes a very deep concern with the relationship between “military” and “civilian” affairs, and the concerns with militarization have most powerfully come from conservatives who are against the federal government, and for the system of free enterprise. So at least one thing that #1 can get you is confused, both empirically and politically.

ARC: A sounding means dipping a fathom into the water to test the depth. — so, metaphorically, this means dipping into the water to get some quick empirical orientations that can feed into problem-formation and the refinement of concepts. In the example we gave above, the work that Lakoff and Fearnley were doing early on gave us some empirical soundings to check the depths: are our concepts right? Have we framed the scope of our domain of interest properly? An advantage of using such soundings is that they allow us to assess a field without actually committing ourselves to long-term fieldwork in the old sense, in part because we don’t know what the problem is.

It is also worth noting that this is an example of an area in which collaboration, along with some hierarchy and some division of labor, offers something essential that an anthropologist working individually might not be able to do. What we call a “sounding” in Fearnley’s case was actually six months of intensive work on a project that also was his senior honors thesis at Columbia. So it was a big project. But for us it came at a step in our broader project that we could use it in the way we described. And, given the authority relationships, this did not pose any problems (although it did take due care with respect to credit, etc.).

ARC: The distinction you are drawing here between anthropology and history is not clear to us. In the sense noted above, we would think of historical research as a technique of inquiry, rather than a methodology. (The fact that history is a “discipline” is another problem that probably deserves some reflection.) We feel — and the practice of anthropologists would seem to justify this feeling — that this technique is appropriately employed by anthropologists in various contexts. What we are trying to talk about is not on the level of technique but on the level of methodology or mode of inquiry. So, from this perspective, history versus the anthropology of the contemporary seems like the wrong distinction.

As to Popkin: we could read him and try to give a better answer. It is entirely imaginable that a sensitive political scientist could come up with the distinctions we have come up with. There are certainly (Weberian) traditions in political science that do terrific work around concept formation (the literature on democracy and democratization, for example). That said, the literature in political science doesn’t convince us that the discipline has produced the kind of insight that interests us. But this is not because we are “anthropologists”, and, to the extent that we have found scholars working on related problems, they often aren’t anthropologists. We’ve had productive engagements with critical geographers and cultural sociologists, for example.

ARC: Per the above, we are observing their practices ofconceptualization — how they develop and operationalize concepts. But as noted above, our aims are different.

ARC: By “this” you mean the moments of recombination? They are, of course, “actual” in the sense that they happen as events in the organization of things, institutions, etc. What we can do in the context of the lab is to come up with concepts that allow us to identify moments of significant recombination. So, for example, SJC and AL are now tracing the various political configurations of techniques associated with vital systems security. The raw material is constituted by things like: Experts identify a new problem in a given analytic and conceptual frame, and endeavor to formulate a response.

ARC: A quick answer with respect to Giddens is that he is producing “theory” about modernity. So, modern subjectivity involves a technologized, calculative, relation to the self. You can go study that, but you don’t really learn anything new — you just confirm over and over again what you already theorized was the case (we moderns have calculative subjectivities, etc.). ARC is producing concepts rather than theory. Their value is to help one see problems, tensions, motion in a given situation and to try to understand it. To identify and find significance in singularities. So our concepts, per Weber, help us identify the significant features of individual situations. So, for example, in the security project, we are not diagnosing a general state of security today. Rather, we have an analytics that lets you see how particular elements are in motion in particular sites.

ARC: Can you say what you mean by “the conventional idea of social science”? We do think that there is much to learn from the other social sciences, even if we don’t ultimately share their aims and techniques (this point is expanded upon in our “What is a Laboratory” paper (Collier, Lakoff, Rabinow, 2006).

We have certainly learned some of the lessons of the reflexive turn, but feel that it needs to be taken in a different direction. Obviously we are not interested in doing na ¯ve, objectivist social science. One difference from at least a clich ©d understanding of conventional social science is that we don’t think we are producing objective truths about the world but rather analytics that identify significance. But here is where we think that more reflection on the kinds of claims anthropologists make would be helpful. After the “reflexive” turn, anthropologists have not stopped making claims about the world. Our question would be: To what extent are these claims accompanied by a structure of accountability or responsibility: that is, what are the norms of adequacy, conceptual coherence, or adherence to something like “the empirical” that make these claims valuable contributions to thinking about the present?

We have tried to outline above how the concept work and empirical soundings lead to progressive conceptual clarification and reframing of significant problems. That feels to us like a lot — indeed, since we don’t believe in “theory” it seems like the most one can ask for. Another example would be the importance for us of the three-fold analytic that we developed between vital systems security, sovereign state security, and population security. For us, this opened up and organized a vast field of empirical problems whose significant interconnections were not clear to us before. Although we are still working on how to explain this distinction to various audiences, we think it is extremely significant and diagnostic. We have outlined some of the ways above.