• Open Access

Public secrets in public health: Knowing not to know while making scientific knowledge



Unknown knowns—or “public secrets”—may play an integral part in publicly funded medical science. In one large transnational field research site in Africa, such unknowing pertains to vital material inequalities across the relations of scientific production. These inequalities are open to experience but remain often unacknowledged in public speech and scientific texts. This silence is not usually achieved by suppressing knowledge but through linguistic convention and differentiation between places and moments of knowing and ignorance. Switching between known and unknown according to situation and interlocutor is an important, largely implicit skill that maintains relations necessary to conduct clinical research—linking bodies, lives, institutions, and technologies across differentials of resources, expertise, and power. Unknowing, then, facilitates research; and it shapes the resulting work and perpetuates the political and economic contradictions that pervade the context and the research endeavor itself. Unknowing thus poses a challenge for conventional anthropological modes of critique and engagement.

Es gibt kein richtiges Leben im falschen.

—T. W. Adorno, Minima Moralia

“Unknowing” can be a significant dimension of scientific medical research: Those involved in advancing important scientific knowledge know certain aspects of the reality they work on and in and yet do not know, do not want to know, should not know, or actively unknow them by way of oversight, ignorance, discursive conventions, and alternative terminology. Drawing on an ethnographic study of transnational science in Africa, I analyze the contribution of unknowing to public health research. “Public secrets” (Taussig 1999), unknown knowns, pertain here, I argue, to vital material inequalities across the relations of scientific production, which are open to experience but absent from public speech and scientific texts.

This unknowing is not usually achieved by suppressing knowledge but is actively produced through linguistic conventions, irony, and differentiation between places of knowing and ignorance. Switching between known and unknown according to situation and interlocutor maintains relations necessary to conduct clinical research—productively linking bodies, lives, institutions, and technologies across differentials of resources, expertise, and power. Unknowing is, then, a condition of scientific knowledge production under given material conditions; in turn, it shapes this scientific work and sustains and exacerbates some of its inherent political-economic contradictions.

African laboratories

The trope of Africa as a laboratory has been popular in European thinking about Africa for over a century, first, among

colonial doctors and administrators who liked to describe their work as experiments and, later, in the writings of historians, sociologists, and anthropologists about these endeavors (see, e.g., Bonneuil 2000; Lachenal 2010; Tilley 2011).1 The metaphor draws our attention to the prominence of scientific, notably, medical, research in imaginaries of Africa and, at the same time, to the tentative, exploratory nature, and often failure, of external governmental interventions on the continent. These two features have recently gained renewed salience, driven partly by the exceptional attention bestowed on the HIV/AIDS epidemic, which for a while appeared to be the continent's main problem, and partly by growing transnational flows of funding in response to this and other health problems.

Medical issues are high on the agenda of external interventions in Africa today, and, because of the nature of the most prominent health problems and of the organizational and economic structures by which they are addressed, scientific investigations remain central. Indeed, research and intervention are often simultaneous, intertwined endeavors rather than experiments preceding interventions as pilot studies or succeeding them as evaluations—as evidenced by the current excitement in the global health field over “evaluations” and “implementation science” (e.g., Padian et al. 2011). This landscape, in which the boundaries between governmentality and scientific investigation are porous, has led some anthropologists—giving the old African laboratory trope a new lease on life—to designate “experimentality” as the dominant social–political paradigm of Africa today (Nguyen 2009; Rottenburg 2009).

Crucial nodes of this transnational order are the handful of large-scale medical research sites that—especially since the end of the last century—have emerged across the continent.2 These sites usually link a local scientific institution, such as a parastatal research institute, a university, or a teaching hospital, to a European or U.S. scientific agency, university, or charity (sometimes to several). Drawing on cutting-edge scientific infrastructure, these “centers” or “field stations” access specific territories—sometimes constituted as “demographic surveillance areas”—and health facilities. Employing hundreds, sometimes over a thousand qualified staff, they involve hundreds of thousands of inhabitants in their surroundings in scientific surveillance, trials, and innovative interventions—with significant positive health impact as a side effect. These programs’ annual budgets run in the tens of millions of dollars, derived from the Northern partners, and their publications advance medical knowledge of Africa and shape national government policy and international guidelines, technical standards, and interventions.

The concentration of scientific medical research in Africa in these “field stations” associated with parastatal national science bodies—rather than, say, based in national universities or government ministries—is driven, on the one hand, by changes in science: notably, increasing cost and rapid turnover of high-tech apparatus (e.g., in immunology and genomics), the rapid pace of technical innovation, and rising quality standards set by, for example, the U.S. Food and Drug Administration (FDA) and the International Organization for Standardization. All of these changes require some concentration and protection of substantial investments (see Geissler in press). On the other hand, the evolution of collaborative African research stations has been shaped by declining standards and scientific possibilities in ordinary, entirely state-funded university departments, hospitals, and laboratories across Africa (see, e.g., Okeke 2010), caused by economic and political crisis and privatizations since the late 1970s, which makes technical and financial contributions of outside partners indispensable. The combination of these factors—capitalization and acceleration of global science and deprivation of local scientific and medical landscapes—results in the contraction of valid scientific inquiry into enclaves.

If African health science once dreamed of expanding modernity's reach across national territory (as in imperial colonization or postcolonial nation-building—“Africa as a laboratory”), using large cadres of public health staff, networks of field stations down to divisional level, and closed-circuited interministerial action (see Geissler 2011; Malowany et al. 2011; Ombongi 2011), the contemporary global configuration of African laboratories reconstitutes technical modernity as an archipelago of a few high-powered and well-resourced islands of global science, connected to one another and to the centers of scientific excellence and policy by long-distance flows of knowledge, materials, and resources (for this geography, see, e.g., Ferguson 2006; Mbembe 2002).

This peculiar geography of medical science in Africa—sharp boundaries between research sites and their surroundings and hopping relations to distant overseas centers—makes for marked confrontations with inequality for those working at these sites. This is, then, the underlying theme of this article: the inequality in transnational public health research and the scandalon—the “stumbling stone,” disturbance, or trigger for thought or ethical impulse—that it may entail. Specifically, I propose that, as a countercurrent to the scientific project of making the unknown known—rendering a dangerous landscape of disease legible and navigable—certain facts about the world, including vital inequalities, are here “unknown” or, rather, handled—in Michael Taussig's (1999) terms—as “public secrets.”

The anthropology of unknowing

To anthropologists, “knowledge” is a central concern. Earlier distinctions between knowledge and belief—with scientific truth as standard—have given way to the relativist multiplication of local “knowledges” and emphasis on “indigenous” knowledge (see, e.g., Hobart 1993; Horton 1967; Worsley 1997), but knowledge remains the positive substance of our scholarly efforts. Accordingly, scientific research is often examined by anthropologists as generating, exchanging, or contesting knowledge (e.g., Latour and Woolgar 1986; Traweek 1988; Woolgar 1982) or, in a relativist vein, as dominating “other” knowledges with a universal scientific one (e.g., Turnbull 2003).

Not knowing has often been treated in a commonsensical manner, as lack of knowledge—equivalent to ignorance, as absence of a specific kind of knowledge, or as a lacuna filled by erroneous belief, for example, by “false” consciousness. Studies of “ignorance” have occasionally turned analytical interest to the substance of not knowing itself, for example, describing the ignorance (vs. indigenous knowledge) produced by allegedly reductionist development practices (e.g., Hobart 1993) or even intentionally promoted by (nefarious) actors such as the tobacco industry (Proctor and Schiebinger 2008). However, in these approaches, nonknowledge is also measured against the standard of true knowledge, and the constructive contribution of not knowing to social associations and processes is only partially recognized.

The creative side of unknowing has, of course, emerged in earlier anthropological analyses of secrecy (e.g., Barth 1975), which drew attention to the social importance of uneven distributions and flows of knowledge, but even there, the positive effort of making things unknown and the meanings and experiences attached to not knowing within regimes of secrecy did not attract much attention. Mostly, “not knowing” has remained tied into the duality implied by the negative prefix and has rarely been appreciated as a creative social form in its own right.

Preceded by Murray Last's (1981) early article on “the importance of knowing about not knowing,” more recently, anthropologists have, from different angles and ethnographic fields, drawn attention to nonknowledge, gaps in knowing, concealment, and ignorance (Green 2009; Mosse 2005; Riles 1998, 2006; Strathern 1999; Wagner 1984), suggesting that nonknowing, like knowledge, can link (and separate) actors and produce, shape, and cut networks, collaborations, and social processes—contrary to what its conception as a nonentity would suggest. In particular, awareness of ignorance or conscious not knowing—the “present absence” of knowledge—has been shown to be a procreative force (Højer 2009). Social studies of science have in this regard especially thought about the importance of not knowing in relation to scientific experiments as a source of validity—as in randomization and blinding—and as a performative tool that may serve particular interests or bolster the authority of those deemed capable of overcoming ambiguity and equipoise, for example, in the context of pharmaceutical industry information politics (e.g., McGoey 2009).

Describing this as “unknowing”—rather than as, say, intuitive, affective, habitual, or embodied kinds of knowing—is a choice, motivated partly by its contrast to (scientific) knowledge—explicit, peer reviewed, printed, and imbued with superior truth status. “Unknowing” delineates the contours of a problem, which, on further scrutiny, is as much about exclusion and the maintenance of the boundaries of legitimate knowledge as it is about the inclusion, in social practice, of other invisible, silent, or practical ways of knowing (and about shifts, and translations, between these). Unknowing is, then, not the opposite of knowing; the pair of terms helps, instead, to describe the work invested in, and the effects engendered by, maintaining this politically salient division.

In relation to Africa, nonknowledge, in the forms of vagueness, dissimulation, elision, and mirage has also been evoked by anthropologists with regard to contemporary “millennial” political-economic regimes that reconfigure the relation between knowledge or truth and their various others (e.g., Comaroff and Comaroff 2000; Ferguson 2006; Ferme 1999; Geschiere 1999; Mbembe 2001; Meyer 1999). In these new regimes of knowledge, nonknowledge is actively produced and maintained, sometimes serving, although not necessarily directly, specific political and economic interests, at other times proliferating in less determinate ways. While epistemic ambiguity, crucially in the form of rumors, has been part of colonial regimes (White 2000; see Hunt 1999), it seems to have proliferated in postcolonial African politics (Cohen and Atieno-Odhiambo 2004). The line between known and not known becomes blurred in a postrelativist fashion: The two do not exclude one another, and yet they cannot be collapsed into “alternative” knowledges. Power can draw nurture from uncertainties and fear generated by the unknown, or it can legitimate authority with superior knowledge.

Emphasizing the political import of unknowing, Taussig (1999) argued that what is known but must not be articulated in a given societal arrangement—what he called “public secrets,” notably pertaining to hierarchy and domination—is constitutive of social order through a double bond with power: making domination unspoken, silencing critique and resistance, and exacerbating power differentials, since the force of making violence unknowable exceeds that of the violent act itself. Power rests thus not just in knowledge; “unknown knowns” are the apotheosis of power. This applies to Melanesian secret cults as much as to European fascisms (see Pitt-Rivers 1971), where Taussig (1999) finds his key cases. In this article, I explore the contribution of unknowing to social order in a much more benevolent and self-consciously progressive situation than the conservative or reactionary regimes referenced by Taussig, in an order that is not bent on maintaining social order and inequality but, on the contrary, pursues epistemic and social transformation, amelioration and justice: public health research.

Nonknowing in knowledge making

I take orientation from these anthropologies of unknowing—of its productivity and political and economic import—to examine a large transnational bioscientific collaboration in Africa, for which—notwithstanding its aim to make scientific knowledge—nonknowledge is foundational and unknowing an active practice within knowledge making. My guiding question is how (and why) a group of people that is brought together by the scientific pursuit of truth with a view to transforming both knowledge and, through it, the world, retains—indeed creates—zones of unknowing at the core of its practices.

I draw mainly on several years of fieldwork in an African city with a high prevalence of HIV and a corresponding intensity of HIV research, conducted collaboratively by national medical research institutions of a North American and an African nation, which I refer to as, respectively, the Government Health Agency (GHA) and the National Clinical Research Organization (NCRO).3 Many of my observations are general in nature, drawing on comparative experiences in diverse large field sites in Africa; some more-specific descriptions are drawn from a particular clinical trial (below, “the CT”) that involved a large group of urban mothers and babies for several years in a successful intervention and clinical, lab, and behavioral data collection and that also gave participants excellent medical care and support. (See Figure 1.)

Figure 1.

Clinical trial participant and doctor at research clinic in Africa. Original painting by research staff member J. A. Ondiek. Courtesy of the artist.

The African partner in my main site, NCRO, is a parastatal organization, founded, like many comparable organizations across Africa, in the late 1970s and consisting of several regional centers. The NCRO center where my colleagues and I worked is one of the best-funded and most productive of these sites, largely because of the successful engagement of overseas partners and, in particular, the collaboration with the GHA. Since its foundation, it has expanded steadily, becoming one of the best-equipped sites for field- and laboratory-based bioscience in Africa. By 2009, it had over 1,000 staff members (including only a good dozen expatriates), drawing on an annual budget in excess of $30 million to conduct research involving hundreds of thousands of participants in trials and surveillance.

The collaboration is the biggest employer apart from the civil service in the wider region; its staff benefit from good salaries, and although, for budgetary reasons, collaborative contracts can only run for a maximum of one year, many staff have enjoyed continuous employment for many years, even decades, on consecutive projects. Jobs are highly coveted and competitive also because they offer educational opportunities and scientific exposure. The same applies to research participation, which entails substantial health care benefits for study volunteers, relatives, and communities and which is regarded by most people as highly desirable.4“Being with GHA” is something few if any local residents would decline.

The GHA–NCRO collaboration consists of a main field station in a rural location outside a city, where administration, data processing and analysis, and laboratories are enclosed in a modern campus with catering facilities, seminar rooms, and so on, which, to protect considerable material resources, is clearly demarcated and well protected. The collaboration maintains two custom-built, well-equipped research centers adjacent to government hospitals, which conduct hospital-based clinical research and coordinate field research in surrounding areas; most of the daily data collection and some clinical procedures are conducted in smaller government health facilities (which, in turn, benefit from infrastructural improvements, staff, and supplies) as well as in participants’ homes; a minor semipermanent installation in specially designed containers deployed next to a government dispensary serves as temporary housing for one trial.

“Unknown knowns” pertain in this site to experiences of material differences across the scientific “trial community” (Geissler 2011)—the study participants and research workers, scientists of different origin, government officials and doctors, academic institutions, health facilities, and state bodies who are connected by medical research work. Inequalities regarding the conditions of life, which are the raison d’être of transnational public health research, are actively unknown in the pursuit of scientific knowledge. In public speech, professional conversations, and scientific reports, these appear at times not to be known, whereas—inverting the logic of a double-blind randomized trial—most (but not all) of those involved do “know,” see, and experience the allegedly unknown materialities and do speak about them or respond to them in certain situations. As a result, material differences are simultaneously known and not known, constituting a “public secret” of sorts (Taussig 1999), in which knowing what and how not to know constitutes a foundation of social organization and process.

In the first part of my discussion below, I lay out some of the basic rules of engagement that characterize present scientific work in this site (representative of other African sites I have visited in my research). I emphasize the stress placed on autonomy and material independence, or immaterial equality—as expressed in idioms of voluntariness, choice, partnership, and so on—that shapes both expectations about research participation and practices of scientific collaboration. I suggest that this historically situated rendering of scientific work in Africa makes differences inherently problematic. In the remainder of the article, I observe concrete engagements between scientific workers (from international scientists to locally recruited fieldworkers and volunteers) and study participants. I show how, while producing scientific facts, people invest effort in “unknowing” difference. In doing so, they neither deny, hide, nor ignore it, yet they do not establish it as explicit truth. Rather, they oscillate between knowledge and nonknowledge, exempting certain areas from the wider pursuit of epistemological closure and contradicting scientific commitments to a correspondence theory of knowledge. Unknowing serves to make scientific collaboration feasible; to link bodies, lives, institutions, funding, and technologies across wide differentials of resources, expertise, and power; and to produce, under given political, economic, and moral conditions, valid scientific knowledge. But, I argue, this feasibility has ethical and political costs; short of an easy solution, closer scrutiny of unknowing and what it does to resulting knowledge might be of use.

The prominence of unknowing in this situation poses a challenge for anthropological analysis: Classic forms of enlightenment critique—iconoclasm and revelation—which privilege knowing over not knowing, reach their limits. Neither “telling the truth,” replacing alleged ignorance with knowledge, nor the familiar anthropological celebration of diverse knowledges has much purchase if people can tell truths from untruths and habitually negotiate both registers. Instead, attention to the work of not knowing itself—neither opposed to nor like referential knowledge of social reality—is required.

(Un)Knowing difference

Research in Africa: A science of difference

Medical science—claims to universal truth, based on standardized and reproducible methods, negotiated through open debate and scholarly consensus, and aiming for transformations of knowledge and the world—has been produced in Africa for over one century, and since its inception, it has involved material disparities. Indeed, differences in the conditions of life and well-being, including disease distribution, nutrition, and health care standards—are at the core of what used to be called “overseas” medical research: They give it moral legitimacy as a social justice project, determine its focus on “tropical” diseases, and allow investigations of diseases and bodily states that are found in Africa but not (or less commonly) in Europe or North America.

Such research involves associations between scientists and institutions in and outside Africa—not only because all science engages networks of consensus production but also because certain forms of expertise and technological, organizational, and financial resources were and are not widely available within Africa. African science has thus always involved relationships across material difference, not only the epistemologically charged one between cosmopolitan scientists and their African study subjects but also those among scientists, employers and employees, technologies, academic and health care institutions, and their nations of origin.

These differences are a source of potential friction (see, e.g., Hoppe 2003; Hunt 1999; Leach and Fairhead 2007; Packard 2007; Vaughan 1991), as underlined by rumors about abuse of power and economic exploitation (see, e.g., Geissler 2005; Geissler and Pool 2006; White 2000). The challenge of working together across material differences can be addressed in different ways, depending on whether scientific work is conceptualized as, for example, imperial reconnaissance, colonial welfare, socialist internationalism, or development aid, to mention but a few of the earlier moral frameworks of tropical medicine. These older approaches explicated difference by establishing a hierarchy between Europe and Africa, and between science and its subjects, through idioms of, respectively, scientific racism, paternalism, or more recently, solidarity and aid. By contrast, contemporary descriptions of medical research in terms of “collaboration” or “partnership” are more careful in addressing inequality.

The age of collaborative science

Epidemiological difference and, notably, “diseases of poverty” remain the focus and justification of transnational medical research collaborations. However, while older regimes of tropical medicine solved the problem of difference by hierarchical encompassment, today—and maybe particularly in the North American–African relations discussed here—hierarchy and dependence are distinctly unpopular ways of rendering difference, partly because of their connotations of colonial hegemony but also because such hierarchies can imply responsibilities and lasting commitments for the dominant party and moral entitlements for its subjects. Instead of paternalistic inclusion, the present regime is premised on freedom, emphasizing the autonomy and independence of the involved entities. The associations that produce scientific knowledge—relations with study subjects as well as among colleagues, institutions, and nations—are here not to be construed in terms of material interdependence, comparability, and, thus, recognized inequality. The explicit demarcation of material difference—such as in notions of race, poverty, underdevelopment, and justice—is replaced by indifference to materiality. For scientific work to be ethically sound, “partners” and “participants” in collaborative research are to meet as equals.

The postulate of indifference and equality—not in a concrete material but in a narrower legal and communicative sense—applies to the individual human under study (the autonomous participant or “volunteer” who signs consent forms) and to the research staff employed by the collaboration (working on comparatively attractive short-term contracts). And it applies to institutions (and nations) engaged through memoranda of understanding or collaborative agreements. On different scales, associations that constitute the scientific network are imagined as voluntary associations between free, independent entities. Collaborative agreements between institutions, contractual relations with staff, and informed consent given by research participants are premised on voluntariness and choice. Autonomy has become foundational for the scientific endeavor; this is particularly evident in bioethics.

Bioethics and the ethos of collaboration

“Bioethics,” one contemporary way of defining what is good scientific work, have over recent years crystallized in successive guidelines and in the mid-1990s in the precepts of Good Clinical Practice (GCP), the global standard for pharmaceutical research, which regulates work with study subjects, now termed “participants” or “volunteers.”5 GCP is a contemporary equivalent of older moral constitutions of science in Africa, such as imperial self-interest, colonial responsibility, or postcolonial solidarity—translated into detailed rules. While remnants of the previous moralities persist in the contemporary formulation, its core value—autonomy—which has gradually risen to prominence in medical ethics over the past half-century, implies a shift in emphasis: The research participant is not a subject of control, compassion, or responsibility, not defined by a relation of inequality, but a free rational agent, whose participation in medical research is not derived from a logic of responsibility or care but of choice.

The shift toward a logic of choice is epitomized by the prominence of “informed consent” as the ultimate measure of ethical medical research (Kuczewski and Marshall 2002). Fully informed, the rational individual is expected to choose to volunteer in meaningful research, balancing potential benefits and risks (see, e.g., Emanuel 2005). The voluntariness of his or her choice relies on the absence of “undue inducement,” that is, the offer of satisfying needs in return for participation—money, food, or medicine.6 Vital material needs, and thus material difference, should not be part of research engagements.

Although “poverty” remains central to the lexicon of global health research, poverty is here not construed as a material, comparative relation but as a propensity that makes an individual (or group) “vulnerable”—a related key trope of bioethics discourse, referencing weakened autonomy. Material justice is here outside the remit of bioethically good research practice. Responding to the (relative) poverty and need of a research participant, in this logic, compromises the person's autonomy and challenges the morality of scientific work. In turn, bioethicists should, according to a leading proponent, “ignore money”—and, by extension, wider material differences—to focus on central ethical concerns such as autonomy (Emanuel 2005:13). The (fewer) ethicists for whom justice is central focus on global funding and policy (e.g., Benatar 1998; Macklin 2008; Meslin 2008). I am more concerned here with the ethics of concrete engagements and the localized experience of inequality arising from material confrontation in one place.

While indifference to material difference is inscribed in the rules of working with research participants, the same values of autonomy and voluntariness pertain to associations between scientists and institutional engagements, as they are conceived in the logic of “collaboration” and “partnership.” The different parties are here described as independent entities who engage with each other out of free and rational choice: This applies to African staff employed with overseas funds on temporary, project-related contracts, and it applies to the larger collaborative agreement between an African national, parastatal institute and a North American government institution (or any other Northern partner). All partners choose to engage or not to engage independently, and the relative inequalities between them, notably, differences in resources, do not entail responsibilities or entitlements and should not affect good working relations.

Irrespective of actual material differences, the ethics of research participation and scientific collaboration is premised on equality in two important dimensions: legal rights and information and communication. This tenet of autonomy and equality makes collaborative science production feasible (just as moralities of colonial domination or the paternalism of aid did in previous regimes, premised on acknowledged inequality). It allows one to work together with others and preserve dignity, stabilizing productive associations against the tensions and frictions that material inequality would risk producing between scientists and subjects, or among colleagues of different wealth and influence, and between their respective, differently resourced institutions and nations.

Yet the exclusion of material difference from the order of knowledge making cannot resolve the concrete practical problems that material difference raises: Making science, and medical science at that, requires close material engagements and cooperative social relations. In the situated daily work of science, then, material differences are seen and experienced—if often differently by different protagonists.

This knowledge is often not explicit, not verbalized and communicated or written or published, but intuitive, habitual, and embodied in skills. And, as it is in potential conflict with the constitutional narrative of equality-in-difference, such experiential knowing of inequality is often unarticulated, sometimes silenced, excluded from public conversation and official record, and restrained to particular situations—“unknown.”7 In the remainder of this article, I explore this movement between knowing and unknowing, drawing on observations around the CT trial and the wider NCRO–GHA collaboration, examining diverse locations across the hierarchical space of the trial community, and crossing the epistemological divide between researchers and researched.

Contradictions (un)known


Unknowing is most obvious around relations with study volunteers, recruited from economically deprived groups in the city, for whom material need—hunger and lack of health care—is directly linked to survival, health, and well-being. Especially in HIV treatment research, economic and medical realities intersect, affecting intervention outcomes and data validity, as nutrition affects drug effects and adherence to therapeutic regimens. Despite the importance of hunger to people's health and the primacy of seeking sustenance in participants’ everyday life, hunger is little spoken about in trial documents and procedures (except for studies that have a specific digestion-related interest like drug absorption studies; see Kalofonos 2010; Prince 2012a, 2012b). When participants were interviewed for our CT, their struggle to procure food did not figure in the questionnaires used during “nutritional counseling,” which individual participants were given when attending study visits and when presenting with signs of malnutrition or sickness. On these occasions, the nutritionist weighed and measured the participant and administered, first, a “food frequency form,” which detailed how regularly each food item from a five-page list, from maize to ice cream, cabbage to wines, was consumed, and, then, a “24-hour recall diet form” of food and drink consumed for three regular meals and up to four intermediate snacks. While the questions, if correctly answered, produced detailed data, there was some discrepancy between the forms’ range of conceivable foods (some relatively high cost) and those actually available and between the forms’ assumption that respondents consumed regular meals and participants’ everyday realities of few food options and skipped meals. Yet the participants in data collection encounters did not usually speak about this gap (for related observations in HIV care, see Prince 2012a, in press). (See Figure 2.)

Figure 2.

Encounters at the research clinic in Africa. Original painting/drawing by research staff member J. A. Ondiek. Courtesy of the artist.

We observed participants, while waiting for study procedures on the benches outside the clinic, speaking among themselves about these realities of their lives; staff, meanwhile, talking among trusted colleagues, sometimes expressed sympathy with participants’ misery. Only occasionally, “off the record,” in moments of personal conversation not planned by study documents, did participants admit to staff that they were hungry and tried to establish a relation between their own needs and staff's presumed resources and responsibilities (not in the sense of legal liability but as response abilities resulting from different resources). Hunger and other material needs were more likely to rise to the fore when staff met participants in their homes and established relations beyond their engagement as researcher and research subject–object (or research program employee and unemployed volunteer). This allowed for the articulation of need, evoking local idioms of sharing rather than making claims or begging, and for affective responses; and staff commonly gave small amounts of their own money or brought food or old clothes to participants who were obviously needy or had young children.

Such gestures suspended assumptions of autonomy and equality, establishing relations of difference. More patron–client relations than disinterested connections between independent agents, these relations were described by CT staff and participants as “friendship” or referenced through kinship terms—“My sister, I brought this for our baby”—enacting local patterns of extended kin responsibility. Practices like private gifts were not foreseen by the written standard operating procedures (SOPs) that operationalize trial practices, and they potentially infringed on regulatory rules. Staff therefore (and because their own money was involved) perceived such donations as ambiguous and did not report them in project meetings; in the words of one of the scientists, “Even a bar of soap can be interpreted as coercion by the ethics review.”

Participants’ deprivation, therefore, does not commonly feature in the public language of research, in conference papers or publications.8 Instead, it is spoken about in informal chat among staff as they sympathize with participants and think about possibilities to assist or reflect on their inability to do so, due to their place in organizational hierarchies, to limited resources, and to infinite needs. Local staff are especially familiar, from their own and family experience, with participants’ economic conditions and know that participants are hungry, but they cannot include this knowledge in their survey forms or systematically act on it.

The only written records of such experiences are the practical documents shared among staff to coordinate work, which are not distributed more widely. In the CT, the minutes of weekly fieldwork planning meetings thus detailed individual cases’“social issues”: Confronting poverty and hunger, staff provided “counseling” and “encouragement”; in cases of medical malnutrition, measurable as weight indexes or blood values, they sometimes prescribed pharmaceutical nutritional supplements for a given period; and, occasionally, one-off support, such as advising a participant to seek support at an NGO's nutritional project, was noted in the minutes.

More concrete entries on social conditions were found in field staff's personal notebooks, used in their daily work to remember details about participants that have a bearing on subsequent visits. In the CT, these notes were collected in an “encounters log” on preprinted, hand-filled sheets, through which field staff communicated with each other about practical concerns. Here, participants’ lives were jotted down in handwriting: depleted food stores and skipped meals, hungry or sick children, marginal incomes, leaking roofs and unpaid rents, participants’ fears of disclosing their HIV status, and hostile husbands. Observations ranged from mundane—“participant is fine; raised a concern of not having a watch so she is not consistent on [drug taking] time” and “had social problems with husband, which is resolved now”—to more medical—“she is malnourished, BMI 18.1; HB 8.5; commenced on Saferon capsules and folic acid”; from emotional and moving—“participant complained of dizziness; had so many issues e.g. lack of food, financial support and kind of rejection … needs serious support/counseling” and “participant was emotional, tears running down her cheeks, threatening to commit suicide”—to moments when professionalism could not hide profound personal experiences—“participant found dead early this morning … SAE and change of status filled, and verbal autopsy.”

This log was continuously updated and consulted; field staff checked it before setting off, or when phoning from the field to obtain details about a participant they could not find or did not know how to approach, or to update the document. Yet this knowledge did not constitute “data”—the term that here marks the limit of recognized knowledge. Its exclusion as data was not intended to hide hunger and poverty; rather, such knowledge is not deemed crucial for the scientific and ethical quality of research. Participants’ chronic hunger was thus dissociated from the knowledge practices of research and from the resources and responsibility of those involved in it. At the same time, knowledge about poverty, hunger, and other material conditions was obviously essential for the functioning of the CT: had staff not known which of the volunteers had particular needs—who could not feed a child before giving drugs, who had no permanent home, no family support—and if they had not shared such information, they would not have been able to find the participants during their restless peregrinations in search of sustenance and could not have supported participants’ adherence to medication and clinical procedures. Neither would they have known where to look for potential participants without a clear idea of the city's and region's economic geography or been able to convince people to join and stay in a trial. The unknown knowledge was vital to maintaining relations, without which research would be impossible, and yet it could not be spoken, lest it raise regulatory concerns and potential disciplinary action, rupturing the collaborative texture and possibly causing moral unease. It had to remain in the known unknown space of daily engagements to do its work.

If one asks staff and scientists involved in these situations why hunger is not discussed more prominently, one hears several valid responses, the most important being that hunger is a background reality, outside the remit of research—a matter of development aid programs and not of scientific institutions: “That's not our mandate.” One cannot fund measures against hunger through a clinical trial budget, except for certain individuals under strictly defined circumstances. For instance, when it presents as acute, clinical malnutrition—a medical “adverse event”—hunger can be treated with supplements or fortified foods, but only until the patient has regained weight or lab values have been restored. Or it can be addressed when diet directly affects a trial intervention, such as in drug absorption studies, in which intake of a modicum of food must be ascertained. Yet, in either of these cases, the food supplied is strictly for the individual patient's use, which ignores the shared nature of food supplies and commensality and the hunger of children or siblings. Underlying the nonknowledges and silences around hunger and other effects of material inequality, beyond such matters of mandate and funding, is a comprehensive sense of impotence among medical practitioners, the recognition of one's inability to respond effectively to the expanding societal and medical crisis and to alleviate the conditions of survival in which public health research in Africa is situated.

While vital inequalities are not much spoken or written about in public, professional engagements, they are keenly observed and spoken about, at times passionately, in private situations, outside the research setting, and in homogeneous groups within which people identify with each other, such as among the female field staff or between these women and female CT participants in their homes. Thus, although they do not become part of the scientific text or of public debate—and are thereby removed from political contestation—they do shape, in private discussions and through silences as well as in concrete mundane practices, engagements within the trial community, contributing to the successful production of scientific research.

Unrelated to study procedures and often driven by personal moral commitment, staff often provide lunch to participants during long study visits or extend other similar occasional hospitality. This is appreciated by participants as a gesture of goodwill and care but not perceived—at least by trial managers—as an integral part of the trial itself. Similarly, outside the trial, GHA scientists at times use their institutional proximity to nonresearch aid programs to ensure some nutritional support for needy participant groups. To go beyond such valuable ad hoc measures and, for instance, propose a systematic nutritional program not only would exceed the research mandate and budget and mean taking on a responsibility beyond the collaboration's capacity but also would create dependencies and hierarchical relations of care (and inequalities among recipients), contrary to the idea of equality and independence that underlies voluntary trial participation.


A similar exclusion of the materialities of survival from what is known pertains to participants’ health care needs. For the duration of the trial (usually a few years), research projects like the CT provide free consultations with good doctors and cover the costs of private hospital admission, drugs, and diagnostics, offering a quality of care that otherwise is unavailable in the region. This care excludes chronic conditions and referral beyond the local hospital (unless illness is directly and unambiguously caused by the trial intervention), and it probably falls short of the kind of health care that privately insured research staff and their expatriate colleagues are entitled to. But it constitutes an enormous improvement over public facilities’ poor and costly services, which participants usually have to rely on, and which they often are recruited from and referred back to after a trial.9 This is particularly the case for HIV-positive research participants, who, as part of trials, receive intensive, personalized treatment follow-up and free care for opportunistic infections. Trial resources thus enable survival for participants and their offspring, and participants sometimes describe clinical trials or the GHA–NCRO clinic as a “health care system.”10

Some dimensions of trial-related health care are defined in research protocols and budgets—increasingly clearly in recent years and, in particular, when health care responds to requirements of the trial. For example “passive surveillance” of diseases, an outcome measure in some studies, makes it necessary to invite participants for paid-for visits to the hospital for all outpatient care. Similarly, care for trial-related “adverse events”—vital for GCP standard evaluations of new drugs—may require paid-for inpatient admissions; some protocols include “emergency care” as a Hippocratic obligation, whereas others explicitly exclude accident and trauma admissions unrelated to the trial intervention.

Yet the broader contribution of clinical trials to the health of participants and communities can only partially be anticipated by trial documents, and protocols commonly describe care beyond trial procedures “on a case-by-case basis” (although in recent years, health benefits are sometimes more clearly defined in research protocols stipulating specific monetary amounts for inpatient and outpatient care per participant). This openness regarding health care benefits may partly be due to the silencing effect of ethics prohibitions against “inducement”—quality health care alone is an important incentive for potential study participants (e.g., Fairhead et al. 2006; Geissler et al. 2008). But, above all, it reflects the complexity of real-life health care, which requires medical doctors’ clinical judgment and triage (since even well-funded trials cannot cover all health care needs in their purview); in response, the health benefits, unlike clinical trial procedures or laboratory work, must often be modified in the process of research rather than being “set in stone” in SOPs (see also Geissler et al. 2008). Any a priori attempt to explicate the rules of trial care would inevitably be defined by its limitations and thereby draw attention to the persistence of unequal suffering (reawakening the genies of the 1997 standards-of-care debate; see N. 12).

Unknowing around health care applies also to the poor standards of care and lack of medicines in government health facilities as compared to the research clinic's near-global standards. These shortcomings are well known to researchers, especially local medical staff, trained in and recruited from government facilities. Despite this, research projects usually describe their own medical care as a mere complement to allegedly existent public services, which are assumed to take over when a trial, because of its mandate, budget, or limited duration, cannot extend care (most obviously in the posttrial referral of participants). In official research documents, discrepancies between trial care and public care in the government facilities (or between official policies and real care provision in such facilities) are often not detailed. The language of collaborative partnership, compounded by the fictions of official government policy and those of the global development partners, insists that these different medical institutions and the clinical practitioners inhabiting them engage as equals. To mention or put in writing that one party is less able or willing than the other to achieve its medical aims would break the collaborative consensus.11

However, differences are known in other ways: During trials like the CT, volunteers are referred for inpatient treatment to the government hospital; even if they are treated on the better-resourced “amenity” ward that incurs higher fees, the clinical observations generated during their stays (essential for data continuity) are often not up to the standards required by the trial protocol, not least because actual hospital procedures often fall short of regulative guidelines; thus, vital measurements such as temperature and cardiac and respiratory rates are not collected when they should be, or, as was repeatedly noted in the CT encounters log, “participant's charts cannot be found” in the hospital. To remedy this problem—an issue of data validity as well as patient care—dedicated and well-trained NCRO–GHA nursing staff and clinicians closely monitor “their” patients.

Having previously worked in government clinics, they possess informal knowledge of local medical institutions and know where and how to intervene, causing as little friction as possible (because, formally, patients in government hospital care are under the authority of government clinicians): checking patient charts and alerting nurses to amend them, ensuring food and feeding implements are available on time, carrying nutritional supplements distributed at the government hospital from there to the private clinic or special equipment from the private to the public hospital, making small payments for diagnostic procedures and supplementing pharmaceuticals that are temporarily out of stock in the hospital pharmacy, and guiding patients through hospital procedures such as X-ray and payments, which they know are unclear and riddled with officialdom. Their familiarity with the setting, and the experiential knowledge arising from their research work, is reflected, again, in notes in the log such as, “participant had a problem at discharge when invoice went missing from her file. It is suspected it was used to sneak out another patient. Hospital administration said they would investigate the matter. The mother was eventually released.” And it is because of the staff's informal familiarity with health care differences that some procedures may be conducted in more expensive private hospitals that are better equipped and staffed. Moreover, patients may be moved to such hospitals temporarily to receive better care, for example, because staff in the government hospital did not consider it necessary to admit a participant referred to them by GHA staff or because the government hospital ran out of intravenous fluids or certain drugs.12

This knowledge—much of it acquired while working in the CT itself, by tracing research participants through public hospital care—helps staff bridge the multiple gaps between the transnational standards of their protocol and the realities of the government hospital, ensuring valid trial data and ascertaining a modicum of care for their patients, enhancing patients’ chances of clinical success.

Yet this detailed knowledge of public health facilities also reveals the acute problem of referring study volunteers back to these facilities after the end of a trial or if they suffer from chronic ailments (e.g., cardiac problems, asthma, or cancer) or injuries and trauma that exceed a trial's time span or budget. If it were a recognized fact that government hospitals do not, or do not reliably, provide the services necessary for patient survival, and if this knowledge were written into research documents or publicly debated within trials, the final referral of volunteers “for further care” to these government hospitals would require more careful consideration.13 While some lead scientific staff admit to “not knowing” exactly which treatments relevant to ex–trial participants are available in the public hospital—and given the almost daily changes of such conditions, this is not easy to know—no initiative is taken to systematically acquire this knowledge, and the experiences that are gathered by staff working across the institutional divide are buried in their notes—“anecdotal evidence” but not “data.”14

Doubts about posttrial referral are left implicit and are partially responded to through ad hoc support for individual patients. The CT's doctors were occasionally consulted by participants after their “exit” from the trial and always seemed willing to advise or even examine these patients, whose personal doctors they had practically been for years, or to assist by contacting their new care providers. This care relies on personal clinical commitment and infringes—like personal gifts in response to hunger—on regulations specifying that interaction with trial participants must end after their “exit” from a study, even if the doctors do not draw on diagnostic or pharmaceutical resources of the trial. Thus, these engagements remain undocumented; again, crucial knowledge about the materialities of survival remains simultaneously known and yet unknowable, unspoken of, at the heart of a medical research endeavor, although without this experiential knowledge, clinicians and nurses would have been unable to collect valid data and ensure adequate care.

Among diverse explanations for this unknowing are, apart from the wish to avoid alterations or infringements of the approved protocol and complications with regulators and funders, the diplomatic avoidance of confrontations with the Ministry of Health and with colleagues working within public hospitals, whom one respects and on whom one relies for access to participants and research settings. Just as in the case of hunger, the unknowing of intimate knowledge of difference in the official script of the clinical trial facilitates collaborations with government clinicians and the Ministry of Health, and both unknowing and knowledge are critical to the practical conduct of everyday research practices.


The same strictures apply to small monetary payments that accompany and facilitate participant recruitment. These so-called transport reimbursements range from £3 to £5 ($4–7) per clinic visit, exceeding actual transport costs by 30–100 percent. They are equivalent to a day's wage in an economy with scarce labor opportunities, where many participants do not know in the morning what will be available to eat in the evening. Such transport reimbursements form part of a wider economy attendant to clinical trials and research programs, which also provides opportunities for monetary income through involvement as “peer-educators,”“recruiters,” and other “volunteers,” who are compensated for transport but not paid for labor.

In trial documents and dialogues, these vital transfers are never referred to as net “payments” but are persistently called “transport allowance,” covering participants’ alleged expenses to make participation cost free. Mentioning “payment” in study documents can cause regulatory problems and delay ethics approval, as the assumed voluntariness of participation, and subjects’ independence, would be compromised by transfers of value.15 Thus, in the CT, a trial form asking participants for “potential concerns” arising from the informed consent process explicitly asked, “Why do we reimburse fare?(Explore for the possibility of confusion of fare reimbursement with payment.)” There is thus a gap between participants’ experience of transport reimbursements as a net gain and their description in the public discourses of research as zero-sum transactions. (See Figure 3.)

Figure 3.

Medical research fieldwork in an African city. Original drawing by research staff member J. A. Ondiek. Courtesy of the artist.

Everybody concerned—participants and staff—is aware of this “misunderstanding” and yet agrees to maintain it. If one asks scientific staff how much they “pay” volunteers, the answer is “we don't pay,” sometimes accompanied by an ironic or wry smile. The opening between terminological rigor and laughter maintains the oscillation between known and unknown; the mirth might also reveal a certain pleasure the speaker takes in the ability to define knowledge and to unknow.

The unknowing of the material realities of monetary transfers is achieved not only by excluding them from written documents and public discourse but also by the use of alternative idioms: reimbursement instead of payment or gift. Like hunger and medical neglect, marginal incomes are not part of the knowledge that the research program generates or explicitly recognizes. As in the previous examples, this unknowing must remain incomplete because the unknown known is important to maintain productive trial relationships: To mobilize prospective participants, “reimbursement rates” must be made known in the “community” (following ethics guidelines, they are often even stipulated on participant information sheets). To offer competitive rates, scientists must know their value in a local economy and how much other research projects offer, and sometimes rates of transport reimbursement are negotiated between different trials competing for participants. In turn, experienced participants, aiming for higher “reimbursement,” know how to justify claims with “raised fares,”“fuel prices,” or “increased distances” (rather than living expenses or minimum salaries). This agreement helps to avoid openly renouncing the fundamental value of autonomous choice. Were the nature of these value transfers to be made known as incentive payment or gift, for example, in protocols submitted for ethical approval, this might invite institutional review board criticism for “undue inducement.” Yet, if, in response to such critique, no money at all were given to participants, it would decrease participant recruitment and retention. Either way, making the unknown known would throw a wrench in the workings of the trial. (See Figure 4.)

Figure 4.

Fieldworker in an African village homestead. Original drawing by research staff member J. A. Ondiek. Courtesy of the artist.

Relations among researchers

Nonknowledge of inequality is particularly striking with regard to poor study participants. Yet the practice of making material differences unknown applies also to relations among research staff, between local and foreign scientists, and between institutions. These actors are not directly threatened by sickness and death, but the conditions of “mere survival” among the poorest also affect life in the precarious new “middle” or, as they locally are referred to, “working” classes—as an omnipresent threat, and through extensive kinship responsibilities. Yet, although no less obvious than inequalities between research workers and participants, material differences among research staff working in scientific collaborations are rarely articulated.16

Income differences between staff are reflected in observable everyday inequalities concerning housing, cars, security and evacuation schemes, health insurance, and children's schooling—that is, much as they do in the relations with study participants, they touch on obvious inequalities in the conditions of life, not always in the sense of physical survival but in terms of pursuing dreams and projects and assuring one's children's futures. At the workplace, differences are associated with managerial rank, access to information and resources, and decision-making power, in which those staff who are employed on “international” contracts often act as research leaders and principal investigators (PIs; the latter issue being targeted by funding agencies’ calls for “capacity building” and institutional policies in favor of African PIs). But, despite their obviousness, they are not generally spoken about in professional engagements, although they might shape personal relationships and attitudes in everyday collaboration.

Many concrete everyday life concerns are not known, or are only partially known, between different groups—an expatriate scientist is unlikely to know much about the intricacies involved in local colleagues’ business activities, agricultural investments, or building projects, crucial for long-term security but sometimes interfering with other commitments; and a local scientist, in turn, has little idea about tax-free mail-order services and the origins of the North American colleagues’ muesli bars or the same colleagues’ mortgage payments and educational saving needs. This may well be because such issues are private, or irrelevant to the other group, but, as undiscussed observations, these facts feed into mutual perceptions.

Indirectly, these differences figure in casual conversations among social peers (rather than in groups involving hierarchical or institutional differences). Although the inequalities are about class and employment rather than geographical origins, they are often informally discussed among local staff by contrasting national citizens and expatriates (or employment category acronyms that distinguish, among others, North American officers, other citizens on North American contracts, locals on short-term contracts, and locals on national government permanent contacts).17 In this context, everyday conversations about “colonialism” among local staff at times substitute open debates about contemporary inequalities and index the sensitivity of these relations, as colonial has negative connotations for everyone involved in this collaboration. Thus, a lunchtime discussion about Christian and traditional practices, a propos participants’ (pharmacologically problematic) herbal self-medication, provokes some young men to declare themselves “traditionalist” and “pagan” and to denounce missionization as integral to colonial domination. A conversation after work about male circumcision, the new weapon against AIDS promoted by GHA, among others, calls forth, half jokingly, hard-line ethnic positions against this supposed assault on “our culture,” which here is rendered as part of a long history of colonial antitraditionalism.

Expatriate staff, especially if not from a former colonizing nation, generally see their endeavor as a rupture from the colonial tradition. Rather than tracing colonialism's historical extensions, they respond with understandable indignation to the term. This applies to progressive scientists, who considered their national history as untouched by African colonial history, but also to the expatriate administrative staff, who cited NCRO's permanent, pensionable employment contracts as an example of the “colonial legacy” that the collaboration had thankfully liberated itself from.

Because of short-term postings in Africa, mundane practical continuities, such as where one lives, shops, or goes to school, are not charged with particular significance by most expatriate visitors. By contrast, such material realities and historical continuities are not lost on local staff, and comments about “them and us” and material differences between the two are common in informal conversations among peers. Thus, when I set off to drive home after one of our first days at an urban research clinic and offered a lift to two junior staff members with whom I had spent the day, they responded, “You are heading that side [pointing toward what in colonial times was designated a “white” area], we are heading there, towards the ghetto [pointing in the opposite direction toward the colonial “African” and postcolonial informal areas of town].”18 This moment of speaking inequality at the end of a shared workweek, and outside the work setting, opened the way for further conversations. It is indicative not only of the unknowing of inequalities but also of the importance of making inequality explicit in private, during moments of conviviality or in socially homogeneous settings.

The episode also points to anthropologists’ ambiguous role as far as inequality and its unknowing are concerned. The practice of participant-observation, dear to our discipline, remains, if not a fiction, a struggle under conditions of extreme economic inequality, and itself involves unknowing. A. L. Epstein's (1981:9–10) description of urban ethnography under conditions of late colonial racial segregation acknowledges the near impossibility of participating in the spaces of the urban poor or obtaining an “insider” perspective (Epstein finding, ahead of his discipline, remedy in his own movements between unequal fields). My own fieldwork was, like that of any urban Africanist ethnographer, riddled by similar spatial effects of class. Maybe cities constitute a particularly problematic situation in that regard, but one might equally argue that the city makes for more straightforward anthropological engagement with factual inequality since unknowing—as the story above illustrates—is easily unveiled there. By contrast, an earlier constellation in which I was involved over several years, in the quintessential village mud hut, could, like most classic Africanist ethnography, be described as relying on a localized fiction of equality, a known unknown (Geissler and Prince 2010). Ours was a carefully calibrated balance of knowledge, which was broken whenever we ventured beyond familiar environs or met strangers who, ignorant of this established narration, honestly if bluntly asked for a share in our wealth. The unknowing involved in individual ethnographic work—intertwined with manifold everyday relational practices and deriving from extended shared presence—is different from unknowing as part of a large institutional structure, but there can be no doubt that the unknowing of inequality, and its effects on knowledge making, would also be a fruitful subject for (maybe especially Africanist) anthropological reflection.

In the NCRO–GHA context, speech about categorical inequalities rarely traversed the boundary of the unequal groups, as in the exchange above. Drinking and other private convivial moments marked by joking and banter and the intimacy of friendship created situations when inequality in professional collaboration could be raised. These moments were the equivalent of the situations of intimate, domestic relatedness that occurred between staff and participants, which, as mentioned above, allowed for conversations about hunger and acknowledged the equality of needs in contrast to differential access. Here, one could at times hear critical comments, sometimes concerning inequalities between those present, more commonly about those affecting others who were not. Income differences were then raised to justify unequal buying of drinks rounds or to negotiate openly about how to split a large bill in reflection of these realities. However, even on those occasions, frankness was an exception. More characteristic was awkwardness, as when a waiter handed the bill to a senior expatriate among the drinkers, common in locations frequented by expatriates, less so in the old civil servants’ clubs predominantly patronized by African officials: the irritation of local staff that the waiter took unspoken differences for granted; the embarrassment of the one who received the bill about this unwelcome rupture of a moment of communion across inequality.

That inequalities are not made explicit by those who have less is understandable; having less power and fewer resources is embarrassing in a competitive context marked by economic pretense, and it is risky to talk about because the idiom of collaborative partnership requires one to perform equality: Speaking out would be to expose one's weakness and to accuse one's (more powerful) interlocutor of speaking untruth about inequality and, thus, to rupture the agreement that sustains one's existence. But inequalities are not discussed more commonly by those in a superior position, such as senior and expatriate staff. One reason for this is politeness and “diplomatic” adherence to the unwritten code of collaborative partnership—to avoid laying bare contradictions between the egalitarian values that support one's personal and professional commitments and material realities. Underlying this is, as suggested above, a sense of impotency, which makes it seem futile to address issues—however personally uncomfortable one might find them—that one thinks one cannot do anything about.

Inequalities among research staff do not figure in public speech and scientific writings, because such explication would—just like drawing attention to participants’ vital needs and researchers’ resources—infringe the postulate of “equality-in-difference” (as opposed to the recognition of unequal conditions and same human needs): Persons are (legally) equal but belong to different places and economic (and medical) orders, limiting the possibility of comparison. For local scientific workers, the difference between expatriate staff and themselves is tangible, historically rooted, within the shared space of the city: hence, the quip about the “ghetto” above. For short-term “expatriate” staff like myself, by contrast, the situation is less obvious: I am not compelled to compare my relative wealth with that of my local colleague, because on some level I do not consider myself part of the city, nor do I necessarily experience the city as a shared polity (unlike my colonial predecessors might have done); instead, I measure myself by the living standards of my “real” home, say, a middle-class neighborhood in a distant European city, where I pay off my mortgage. Thanks to this fiction, I can stick to the idea of scientific collaboration as a link between different equals and between distant places—rather than situating it within a shared, contested, social space. The segmentation within the city, and in the trial community, is framed within larger, “global,” geographies, which, on account of their scale, are perceived as beyond the reach of powerless scientists.

The institutionalized aversion to articulations of inequality and diversity of interests is underscored by connotations of the word politics or political. Referring to something in this context as “political” implies that it is “tricky or dangerous,” charged with power issues, and, thus, best handled diplomatically by seniors; in the context of meetings, it might imply that it is best discussed in a smaller circle. By contrast, an individual or group among the local community or among staff that is deemed “political” may be seen as “troublesome.”19 Although the collaboration's senior staff, in principle, wishes for open debate and encourages junior staff and community members to “speak out,” this rendering of politics suggests that the conflicts of interest that inevitably arise from material inequalities are not necessarily considered a legitimate object of political contestation.

Not speaking about obvious inequalities between researchers of different origin is a way of unknowing them. This silence is accompanied, just as in the examples pertaining to research participants, by peculiar ways of speaking. The equivalent of “transport reimbursement” as a seemingly valueless transfer of value is here the “per diem” allotted, for example, to research staff for travel out of the station or to international conferences (see, e.g., Jordan Smith 2003; Ridde 2010).20 The designation of daily payments from the GHA to individual staff as “per diem” suggests that these allowances are equivalent to the costs of bodily sustenance for a day. However, for many lesser-paid local research workers, these are vital salary components—with per diems exceeding actual salaries. While their senior colleagues stay and eat in international conference hotels, local junior staff compete among themselves over who saves most of the per diem by staying with relatives or in less expensive accommodations and eating and drinking outside the conference venue. The tangible inequality in the daily life of these two kinds of conference participants—and, presumably, in their conference experience—appears here as “choice,” because, receiving a per diem, the local staff could choose to eat and drink in the international conference hotel (the same “voluntariness” as implied by the notion of “transport reimbursement”). Wider inequalities and dependencies are thereby silenced for the duration of the conference.

The conference situation also illustrates the diversity and degrees of (non)knowledge in this order: An expatriate Ph.D. student whose grant pays for the conference hotel feels uncomfortable that local colleagues with whom he works and socializes “in the field” stay far from the hotel and spend their evenings in different venues; a visiting expatriate donor representative, by contrast, sees nothing of this and makes, on the contrary, a favorable comment on the large proportion of African participants in attendance at the meeting; a resident expatriate might know who stays where but takes it for granted, focusing on the meeting's scientific content; likewise, local staff members see no reason to question the situation and are satisfied with the additional income at the time of annual school fee payments; the only sarcastic remark about these differences, uttered in a poolside conversation with the foreign Ph.D. student, comes from a very senior African scientist and university professor, who has been in this trade for a long time and has nothing to lose from an occasional joke about its blatant contradictions.

Performing equality

These ways of unknowing inequalities are complemented by performances that make equality visible, conjuring up the vision of partnership. In scientific conferences, researchers participate on equal terms. Seating arrangements, conference meals, access to presentation technology and refreshments create a situation of scientific equality, notwithstanding the different personal circumstances. The design of PowerPoint slides and styles of presentation are carefully rehearsed, and the many names of scientific authors on the program present a mixture of local and expatriate scientists, displaying collegiality. During breaks, participants mingle over free pastries around the pool, and little in dress or comportment would suggest significant inequalities. While this performance of science as level playing field certainly conceals differentiations—encoded in long and carefully weighed (and interpreted) authorship lists on the papers or in decisions about who presents which finding when and where—it is important to recognize how much this situation, nevertheless, is appreciated, especially by African scientists: For the older ones, the contrast to formal and informal segregations inherited from colonialism is a personal achievement, and, for younger ones, the appearance of equality on the conference stage holds up a promise for their global professional future.

Even more explicitly performing equality are employer-sponsored parties and team-building exercises that large-scale research programs commonly arrange. Staff parties in relation to successful projects, retreats to luxurious tourist destinations, sporting competitions to call forth the spirit of unity or to celebrate global health days, and events with specially hired team-building consultants organizing games and catering have the explicit purpose of making all employees feel like they are part of the larger collaborative whole and of bringing about togetherness and partnership across difference. Inequality is, on such occasions, made invisible, notably by the “equal” terms of sports. Physical proximity and shared bodily activities, albeit open to varying cultural interpretations, are meant to produce the experience of equal rights and shared purpose.

These performances of equality are often accompanied by some muttering about the inequalities, about the alleged stinginess of the organizers, about the choice of an expatriate caterer rather than a local women's group, and so on, by those who feel less equal than others in the trial community. Irony, “seeing through” realities and intentions behind such exercises, and self-conscious laughter while participating in them are common features of such events and not limited to any particular group; at the same time, people enjoy the opportunity to have fun together. Only few senior local scientists may abstain from these events, sometimes deploying their absence as deliberate critical statement (not always noticed by the addressees), emphasizing differences contrary to the aimed-for inclusiveness. Yet, despite their seeming triviality, such enactments of equality are more than just make-believe. While a day of group games cannot undo experiences of inequality, shared time, sports, eating, drink, and laughter—including ironic laughter over the obviousness of the exercise—have effects.

The alternating and contradictory layers of experience—known, unknown, and known again—are exemplified by a politically particularly conscious NCRO colleague who showed me with pride as well as irony his new screensavers, which depicted him with famous international researchers in luxurious venues on the occasion of international conferences around the globe. While he “saw through” the veneer of equality that such events conjure up, he did experience close social relations, scholarly exchanges, and the fun and privilege of exploring new countries and unfamiliar places. The same ambiguity applies to less centrally organized kinds of team-building exercises, such as scientists sharing meals and drinks during work, throwing private parties, or engaging in joint weekend expeditions on a low budget, adjusted both to those on a relatively low salary and to those who prefer simpler, more “human” or more “real” experiences.


This article originated, like many an anthropological text, from an awkwardness during fieldwork: the feeling one has when talking to someone who, one is sure, knows that what is spoken is at odds with what can be seen. An uneasy situation that is hard to rupture, because that would entail admitting to speaking the untruth and, worse still, accusing the interlocutor of doing the same. Yet this situation is not about “lying,” which entails that one party knows while the other does not; neither is it appropriately captured by a term like ideology, which, again, entails that some of those concerned do not know the true nature of reality. It is neither false consciousness nor conscious falseness. Both of these interpretations are based on the idea that knowledge and the world are or are not in a relation of correspondence. This is not at issue if everyone, as if following a tacit agreement, speaks as if they do not know what they know, do not see what they see. An odd situation, especially if one finds it among scientists.

Working in Africa with public health researchers—sharing goals, respecting science, enjoying company—I was time and again struck by our faculty to unknow our daily confrontation with inequality. As noted above, this unknowing is a concern for all metropolitan knowledge production in postcolonial settings (or across class divisions), including ethnography. For the context studied in this article, public health science in Africa, which is both motivated and enabled by vital inequalities, the question arises, what does this unknowing do for, and to, science work?

“Unknown knowns” perform double work: They maintain consensus over the ontological premises and principal values that underlie research collaboration—such as autonomy—and thus sustain working relations and procedures. At the same time, what is unknown and yet known provides flexible, malleable operational guidance, without which the research would be impossible: Were the fieldworker to acknowledge the volunteer's hunger, the research nurse to respond to working conditions in public hospitals, or the junior staff member to be aware of resources and possibilities provided by transnational collaborations—and if all of them at the same time were to disregard the dividing line between what can be known for fact and what knowledge should remain less definite—then their collaborative effort at revealing new knowledge about disease and health would be disturbed. Unknowing has a constructive place in contemporary arrangements of scientific knowledge production in Africa.

The consensus and stability established by this public fiction, then, allows for private efforts by those who produce science to try doing good according to their own morality, within the limitations of what is and what can be known: applying for additional funds that can be used to renovate a building, which also benefits the public hospital (albeit without stipulating this motive on the grant application), or trying to ensure staff contract extensions and training; setting up a corporate responsibility program or providing free water supplies to one's neighbors; examining and treating patients outside one's protocol, using program transport to refer a nontrial patient for treatment, or personally caring for participants after the end of a trial; setting up charity events or paying school fees. Going beyond what they are obliged, or even permitted, to do within budgetary and legislative constraints (and health system dysfunction), these actors provide help in a pragmatic, tentative manner and, sometimes, slowly extend possibilities within the lasting boundaries that separate research work from everyday human engagements and public, formal responsibilities from private commitments. Maybe, then, silences and evasions provide as stable a ground for effective social relations and actions as “hard facts”?

Yet unknowing has also less desirable effects. It affects the workings of social relations within the trial community: Issues that are unknown are harder to negotiate among concerned parties. This can prevent factual improvements—of care, pay levels, work practices, and lives. Moreover, in the absence of open debate, speculation about partially known disparities can translate into concealed misgivings and animosities, which can disturb research work or collaboration, through negative public attitudes or covert obstruction, or affect work morale and data validity. Unarticulated tensions can be converted into rumors of occult activity and accusations (see, e.g., Leach and Fairhead 2006) or may seek an outlet through labor tribunals and legal proceedings (for a case unrelated to the NCRO–GHA collaboration, see Nordling 2012 and Daily Nation 2011:10). Thus, in some situations here, differences and contradictions that (for diverse reasons) are unknown should simply be made known for the benefit of political articulation and contestation.21 This is, then, a task for doctors, scientists, and participants in medical research—guided by progressive ethicists, activists, or journalists.

For the anthropologist, the question remains what this collusion to unknow does to science and scientists. Partaking in this public secret, I suspect, affects individual motivation and direction of those who spend their lives in public health research, blocking the opening, the distant horizon of action, that the confrontation with injustice evokes. An unflinching gaze at the scandalon of unsatisfied vital needs, in the sense of what philosophers discuss as an “ethical impulse,” can provide orientation for action, even though—or precisely because—the discrepancy between the observed need and one's own abilities, and between one's own and the other's resources, remains.22 By contrast, the exclusion of this gap between is and ought from public text and from spoken acknowledgment depletes motivation and affects orientation on an individual level. The limited vista of the presently doable curtails one's view, imagination, orientation, and impetus. At best, injustice features then as something beyond this prospect-in-lieu-of horizon, as “not our mandate,” or “mixing politics and science,” or plain fatalist “unchangeability” (the negative connotation of politics, above, is a case in point).

Individual effects accumulate, then, to shape a particular contemporary organization, economics, and, importantly, temporality of “global health” science, which the historical anthropologist Guillaume Lachenal, in a recent critique of speculative hypes in “global health,” pointedly designates as “nihilist” (in press). The expectations of ever-new technological fixes for human suffering, powered by global charity and public–private partners that dominate contemporary “global health” conceal, for Lachenal, a sense of standstill beneath the hyperprogressive rhetoric—more and more of the same, ever huger leaps, on the spot. Maybe the open confrontation with inequality as scandalon, or stumbling stone for thought, suspicious of solutions, would reopen and concretize horizons of the scientific search. More of Camus's Dr. Rieux then, than of charitable billionaires’ technological fixes?23

I am in no position to predict where science would move instead, if it did face up more directly to inequality as a deontological signpost. On the largest level of global strategies and structures, proposals have been made by ethicists like Ruth Macklin (2004, 2008) and Solomon R. Benatar (1998) and some by activists (e.g., de Cenival 2008), who argue that medical research should address the problem of global justice and the right to care through financial commitments. As for the content of scientific work—hypotheses, evidence, data, method—the medical anthropologist Steven Feierman (2010) suggests—on the basis of his ethnography of his university's collaborative medical research in Africa—that recognition of inequalities in transnational health research should lead to a diversification of models of evidence beyond randomized clinical trials and to greater interest in local parameters and standards and health systems research, instead of fixation on pharmacologically driven vertical intervention. From a different theoretical and methodological angle, the geographer Simon Reid-Henry (2010), drawing on his ethnography of the Cuban medical research sector, points to the possibility of a different bioscientific epistemology, premised on recognition of material needs and scarcity, combined with an emphatically open-ended horizon of possibility. The virtue of these reflections from diverse disciplines lies not in offering ready solutions—proposing a new fund, more social science, more local capacity building, and so on—but in contributing to the debate on the possibility of another kind of science.

Crucial to this debate is that material inequality—between individuals, groups, territories, organizations, or nations—translates into positioned interests. While it may be easy enough to agree across lines of inequality that the world would be a better place with less illness, hunger, and suffering, propositions to engender social and scientific progress are rooted in material positions and ownership. That inequality produces conflicting interests seems obvious enough, yet it diverges from the underlying narratives of “global health” (and similar 21st-century socioscientific endeavors like humanitarianism and environmentalism), according to which a universal human cause, driven by objective science (or reason), would seem to inevitably be shared by an ever-widening inclusive “we”—as in the various “causes of our generation,” from producing HIV vaccines or eradicating malaria to combatting climate change or poverty. Making inequality and exclusion explicit instead of unknowing them, underlines that these challenges are ultimately matters of contestation, struggles about material boundaries, rights, and claims. Progress, then, is as much about disagreement as agreement, and collaboration is a site of conflict. I note above my sense of discomfort at the smoothness of a model of collaborative partnership that often unknows the presence of difference; under given material conditions, collaboration is, and must be recognized as, uncomfortable.

Anthropologists as critics?

Unknowing raises questions about the stakes involved when one “speaks truth”: What risk does it entail, for the speaker's relations and position and for the social order and practices that that individual and his or her interlocutors are part of? What is the anthropologist's ethnographic and political contribution? How can one critique a social constellation predicated on actively unknowing inequality? Classic forms of enlightenment critique—revelation (of reality) and iconoclasm (of false representations)—have less purchase under conditions of unknowing. Nonknowledge of the sort described here differs from older regimes of the untrue, such as (e.g., religious) superstitions, false (e.g., class) consciousness, or (e.g., nationalist) ideology, as well as from classic modernist knowledge–power compounds embodied by disciplinary colonial tropical medicine (see Packard 2007). Nonknowledge does not suppress truth and thus cannot be remedied through better knowledge (pace the “agnotologist's” promises; see Proctor and Schiebinger 2008). Neither does the once fashionable multiplication of “knowledges” of equal value adequately counter unknowing. If everybody involved in a collaboration can tell truths from untruths and engage both registers simultaneously, anthropologists add little by “telling the truth” and even less by relativist denial of such a distinction. Only by attending ethnographically to the work of unknowing itself can we begin to grasp the social processes that unknowing enables and then proceed to discern, and engage in, their contradictions.

The openness of responsibility

Beyond such ethnographic contributions to medical ethics, anthropology's way of knowing itself—and the different ethics this entails—might have something to offer to overseas “trial communities.” Defending ethnography against the “imperialism” of bioethical regulation (Schrag 2010), Rena Lederman and colleagues discussed in this journal (see “AE Forum: IRBs, Bureaucratic Regulation, and Academic Freedom,” November 2006) how unregulated, unfunded “informal” fieldwork “at home,” in the anthropologist's everyday world, like all “real” fieldwork, abandons work–life, formal–private distinctions that regulatory ethics presume. Rather than “human subjects,” such work engages “social beings in historical circumstances, including the ethnographer” (Fassin 2006:524). Gaining knowledge by being-with implicates researcher and researched, individually and together, in the politics of the field; the “informality” of fieldwork conjoins persons, who are always already relational and intertwines biographies as it transfers knowledge about lives (see Strathern 2006). And if intersubjectivity is the foundation of knowledge of the world—and anthropology's claim in this regard is meant to have wider purchase—then (multisided!) responsibility is inseparable from the epistemological endeavor. An anthropological ethics, then, is not concerned with regulatory “protection” of subjects but with “taking care” of and with relations with multiple others (see Association of Social Anthropologists of the UK and Commonwealth 1999).

From here, rethinking scientific work in the field-station situation could commence, decentering “(bio)ethical imperialism” with insights from the academic periphery: If station and field are simultaneously coinhabited and studied, professional and domestic, public and private, then knowing and being, epistemology and ethics, experimentation and responsibility cannot be separate. Knowing a field, then, begins by acknowledging that one is present in it and that one's shared presence with others is the inevitable condition of knowledge. From this recognition, the undoing of unknowing begins. And as Strathern notes in her contribution to the above debate, since such responsibilities are “embedded in social situations not defined by the parties themselves,” they require openness, especially toward unresolved contradictions: “Closure should not come too quickly” (2006:533).


  • Acknowledgments. I am grateful for the courageous support for this ethnographic project from members of the NCRO–GHA trial community. For this article, particular thanks go to the leadership and the scientists of both organizations, who welcomed me and my colleagues to study their work and constructively engaged in debates and disagreements throughout research and writing. The penultimate draft was presented publicly at the NCRO–GHA center, and comments and critique were included in this version. Many of those present on that occasion were in agreement with this article's argument about “known unknowns,” and my observations caused little surprise. Yet the ensuing debates sharpened my attention to a different question, which, alas, remains underexplored here: how well-intentioned and eminently well-trained people try—despite adverse conditions and the limits imposed by organizational structures, conceptual frames, knowledge, and sensitivities—to do the right thing; how they realize (if not always publicly) mistakes and failure and pursue incremental improvements; I take up this crucial issue—more suited to extensive ethnography—in a forthcoming monograph. While this text has been improved by the clarifications, criticisms, and advice of colleagues from GHA and NCRO, the views expressed are mine and do not reflect (indeed, sometimes differ from) those of NCRO, GHA, and their staffs. This article is offered as a contribution to, and extension of, our discussions.

  • I was privileged to conduct extensive periods of fieldwork together with my colleagues Philister Madiega and Gemma Jones. I have greatly benefited from conversations with our London research group's Ph.D. students studying clinical trials in different African sites. Colleagues at workshops in Halle, Germany; Kilifi, Kenya; and London, where earlier drafts were discussed in 2009–10, provided valuable critique, as did my coworkers in the research group “Anthropologies of African Biosciences.” I am most grateful for continuous discussions with and uncompromising critique and advice from Ruth Prince, during our shared fieldwork and ever since. Thanks, finally to the anonymous but generous AE reviewers, whose nuanced advice I only began to heed here.

  • 1

    I use experiment in this article as a value-neutral reference to scientific hypothesis testing, which is central to improving scientific medicine; controlled trials, which in recent years have emerged as the gold standard of medical, as well as, increasingly, social experimentation are one instantiation of the more or less formalized forms of experiments that drive medical knowledge. I do not mean the term to connote the negative associations sometimes carried by public discourse on “human experimentation.” I am grateful to my medical colleagues from the “Government Health Agency,” for whom the term evoked the latter, for alerting me to the potential for misinterpretation.

  • 2

    This characteristically African configuration of transnational science differs from that described by Adriana Petryna (2009; see also Petryna 2002) in her pioneering work on international epidemiology and trials outside Africa, in that organizations and funding in African research sites are largely public—commercial interests being less prominent—and in that local clinical, diagnostic, and research facilities, equipment, and expertise are not readily available on-site (unlike in India or Ukraine) but have to be restored, transferred, or custom-built within designated spaces. The distinction between “public” and “for-profit” industrial research is in my view an important analytical focus, notwithstanding the obvious fact that public science and academic institutions have themselves been transformed in recent decades and have been reshaped by economic and cultural models of audit and market.

  • 3

    I choose to “unknow” my main field site's particulars both because my descriptions are informed by comparative experience from several similar large-scale scientific installations in Africa and because it is not my intention to critique a particular institution or its staff or to disturb important scientific investigations but to discuss general patterns.

  • 4

    The attractiveness of trial participation is witnessed not only in long, early-morning queues when new trials recruit and in former participants’ persistent requests for more trials but also in phenomena like double enrollment in multiple trials and manipulating pretrial “screening” (e.g., giving false information about sexual behavior, age, place of residence, health status, or marital relationships, which determine eligibility for a trial), tactics that necessitate trial coordinators’ vigilance and technological innovations like fingerprint databases.

  • 5

    GCP is a regulatory protocol, issued by the International Conference on Harmonization (ICH), that details practical procedures, including those pertaining to ethical matters (consent, incentives, etc.) around clinical trials. Endorsed by the FDA, responsible for licensing pharmaceuticals to the world's largest pharmaceutical markets, it is of crucial importance to the conduct of clinical trials all over the world.

  • 6

    This is the common interpretation of “undue inducement,” following the classic logic of the “gift to strangers” (Titmuss 1970; see Oakley and Barker 2004); recently, some more-market-orientated ethicists have defended the rightfulness and necessity of inducement, including straightforward payments, but their views remain, as yet, a minority position (see Emanuel 2005).

  • 7

    Recurring confrontations with versions of difference—unemployment and lack of food or dysfunctional government and insufficient health care—continue to produce, despite the closure that GCP implies, bioethical controversies (see Angell 2000; de Cernival 2008).

  • 8

    After the end of our fieldwork, senior staff initiated a corporate responsibility program funded by voluntary contributions (alongside an initiative to provide benefits to surrounding communities by letting them recycle waste paper and used furniture). Although this project raised the profile of charitable actions and drew some attention to material need, it remained separate from the sphere of actual scientific production by avoiding the formal order of SOPs and relying on private funding.

  • 9

    In the available public facilities, medicines are scarce and diagnostics largely unavailable, and even basic treatments, which should be free of charge, may imply payments.

  • 10

    These contributions to well-being are to no little extent an achievement of the 1990s political struggles about “standards of care” in transnational HIV research (see Angell 1997; Lurie and Wolff 1997). While these debates focused on the question of whether “local” or “best” (i.e., U.S.) standards of HIV treatment should be used as comparison in African controlled trials, the underlying issue was the differences in health care technology and survival options associated with average African and Euro-American medical settings. Academic debate brought these differences temporarily to the fore. Widespread consensus that, in principle, best standards of care should apply closed this debate as far as controlled trials were concerned. Yet poor public health care continues to raise questions (see, e.g., Kent et al. 2004) and has not led to definite regulations or been commonly discussed in scholarly publications or research protocols (for a provocative exception, see de Cenival 2008); instead, responses to low standards of care are relegated to the individual morality and authority of scientists implementing clinical trials in practice.

  • 11

    This example underlines the continuities of unknowing between the research setting described here and the realms of public administration and policy, where certain material conditions—lack of drugs or hospital malfunctioning—tend to be systematically unknown.

  • 12

    In two recent hospital-based studies in the area, scientists observed that inpatient mortality had been more than halved in the study group, not because of an intervention but by implementing the hospitals’ own algorithms correctly. This very positive effect of the trial will be reported, as a step toward greater openness and further constructive critique of the health system.

  • 13

    This paradox was made clear by a senior North American researcher in Uganda, cited by Lotte Meinert (in press) as quipping, “Which health system?” when asked why his organization did not integrate more into the government system; yet the same project referred participants posttrial to local health facilities.

  • 14

    To my knowledge, none of the ethical review boards I have engaged with demands that research protocols that suggest referring participants “back” to public facilities display knowledge of the quality of these facilities or document the actual public availability of appropriate care, such as second- or third-line antiretroviral medication.

  • 15

    In some Northern contexts, research participants can be paid to induce them to participate in trials. By contrast, I found in various African ethics review boards adamant objections to any form of “incentive”; this might reflect stronger awareness among African review board members of participants’ often abject poverty; according to African scientists I informally spoke to in different sites, it also is motivated by their fear that if monetary inducements were permitted and thus expected by potential participants, it would become impossible for African institutions (given their relative poverty) to conduct human subject research without overseas partners.

  • 16

    There might be more material similarities between the latter than among the former. In large-scale collaborative sites, income differences between research workers and expatriate scientists, taking account of benefits, can amount to a factor of 30, and even incomes and allowances of similarly qualified scientists employed, respectively, on “local” and “international” terms are very different.

  • 17

    These quasi-racial connotations are transported through ambiguous terms such as Kiswahili mzungu; while this term is used for people with a “white” complexion, it can also designate the expatriate social category, irrespective of skin color or particular social traits and behaviors. Discourses about class can thus take a racial tinge and vice versa.

  • 18

    While the use of the term ghetto probably refers to North American “black consciousness,” the areas referred to are socially diverse and include informal settlements as well as planned flats for civil servants and mansions with all amenities. This internal class structure and historical diversity are here glossed over by a simple distinction of “sides.”

  • 19

    This use of political resonates with the inflationary use of the term in academic institutions and organizations, and it carries older values of science, and civil servants, as outside politics (see, e.g., Geissler 2011).

  • 20

    Per diems have become a dominant feature of collaboration in science and development across Africa since the late 1970s. Their rise to prominence goes hand in hand with the economic deprivation of African governments (and the attendant decline of scientists’ and doctors’ salaries) and the emergence of the collaborative model of scientific coproduction. The institutional history of per diems—as compared to other forms of transfers like the “top-ups” of the 1970s, the government salaries and allowances of earlier periods, and contemporary short-term contracts and volunteer compensation—deserves historical study.

  • 21

    This is the case when experimental interventions or research-based policies function insufficiently or have iatrogenic effects (see Parker and Allen 2012). If, for example, a new HIV treatment policy is introduced that relies on the availability of specific groups of drugs that, according to official reports, should be available in a given health system but are known not to be or to be in inadequate supply, then the discrepancy between institutional truths and practitioners’ experiences must be made known to protect lives.

  • 22

    Claire Wendland shows how, for Malawian junior doctors, confrontations with patient's suffering lead to “diagnosis of political etiologies, and the proposition of political activity as therapeutic intervention” (2012:42). Such politicization of social suffering would be less likely to arise if, as in the case of doctors in research collaborations, the confrontation with lack is cushioned by research-related resources and by unknowing the social origins of suffering.

  • 23

    The problem of articulating critical dissent within public health today is instantiated by the medical anthropologist and doctor Paul Farmer, the closest social and cultural anthropology has come to producing a global moral icon. Apart from his admirable double act as clinician, moral voice, and scholar, his well-considered oppositional stance vis-à-vis the global health establishment and its “bioethics” has informed debates for over a decade (Farmer 2005). Yet even his medical-cum-anthropological enterprise Partners for Health (PIH) cannot escape the contemporary idiom of transnational “partnership,” at the same time that its fund-raising rhetoric relies on images of extreme poverty, behind which the trope of “established star-academic at world's richest university cares for world's poorest country patients”—Partner to the Poor (Farmer 2010)—invites cynical critique. Without sharing in these criticisms, I note that PIH's self-representation does underline the difficulty of articulating transformative social critique within a charitable endeavor under given political-economic conditions.