Journal of Bioethical Inquiry

, Volume 9, Issue 3, pp 327–338

Framing Responsibility

HIV, Biomedical Prevention, and the Performativity of the Law


    • Department of Gender & Cultural Studies, School of Philosophical and Historical InquiryUniversity of Sydney
Original Research

DOI: 10.1007/s11673-012-9375-x

Cite this article as:
Race, K. Bioethical Inquiry (2012) 9: 327. doi:10.1007/s11673-012-9375-x


How can we register the participation of a range of elements, extending beyond the human subject, in the production of HIV events? In the context of proposals around biomedical prevention, there is a growing awareness of the need to find ways of responding to complexity, as everywhere new combinations of treatment, behavior, drugs, norms, meanings and devices are coming into encounter with one another, or are set to come into encounter with one another, with a range of unpredictable effects. In this paper I consider the operation of various framing devices that attribute responsibility and causation with regard to HIV events. I propose that we need to sharpen our analytic focus on what these devices do, their performativity—that is, their full range of worldly implications and effects. My primary examples are the criminal law and the randomized control trial. I argue that these institutions operate as framing devices: They attribute responsibility for HIV events and externalize other elements and effects in the process. Drawing on recent work in science and technology studies as well as queer theory, I set out an analytic frame that marks out a new role for HIV social research. Attentiveness to the performative effects of these devices is crucial, I suggest, if we want better to address the global HIV epidemic.


Criminalization of HIVBiomedical preventionRandomized control trialsResponsibilityStigmaPerformativity

In this piece I develop an argument and an analytic frame for tracking the ways in which responsibility for HIV infection is attributed and the effects of these attributions. My hypothesis is that responsibility for HIV transmission is always framed in some concrete way and that these framings have effects that can be attended to. My hope is that the analytic frame I outline will provide a means of attending, empirically, to the production of HIV effects—effects such as HIV transmission, HIV prevention, and well-being—where these effects can be conceived as events: that is, as the contingent outcomes of the collective activity of a diverse range of actors, both human and nonhuman, including technologies, devices, discourses, scientific practices, health care settings, environments, juridical decisions, and desires. Or, what Rosengarten and Michael (2010) have termed an “HIV prevention assemblage.” I believe that developing such a frame is crucial in the context of the complex convergences and tensions between biomedical, technological, and regulatory apparatuses that are currently producing HIV effects.

To do this, I make use of one of the most widely deployed concepts in the contemporary social sciences and humanities—the concept of performativity. Well elaborated in the somewhat disparate fields of gender studies and economic sociology, performativity has in recent years become a way of thinking about effects, “and in particular, to supply an alternative to causal frameworks for thinking about effects” (Butler 2010, 147). When Judith Butler says, for example, that gender is performatively constituted, she is calling into question “whether there is a stable gender in place and intact prior to the expressions and activities that we understand as gendered expressions and activities” (2010, 147). Here, performativity serves as a means of countering a certain sort of positivism and essentialism that structures gendered relations and is used to explain gendered effects. Similarly, when Michel Callon (2007) talks about the performativity of economics, he is drawing attention to the way in which certain practices of description and measurement, characteristic of the discipline of economics, participate in the making of this thing we call the economy. On this approach, economics does not simply stand outside the economy, describing and analyzing it, but is an important element in the practical making up of the economy. Though Callon writes from a different disciplinary perspective than Butler, both scholars generate reflection on the effects of certain descriptive, calculative, and gestural practices and invite us to think more pragmatically—and in a less deterministic fashion—about the production of certain ontological effects.

Performativity is associated for some with a certain linguistic emphasis, but this need not be the case. There is now a substantial body of work within the social sciences on the performativity of objects (Cochoy 2010), devices (Callon, Millo, and Muniesa 2007), research practices (Law 2004), even scientific instruments (Barad 2003). The concept was used inventively by the late Eve Kosofsky Sedgwick (1993) to think through the social and political workings of affect. Introduced into the sociology of HIV/AIDS, the concept may provide a means of reflecting on—and becoming more cognizant of—the excess and constitutive effects of interventions, whether these interventions are legal, biomedical, or methodological. To illustrate this potential, in this essay I consider recent developments in two of the major institutions that organize our responses to HIV/AIDS: criminal law and clinical research (in particular, the randomized control trial). Both of these institutions have achieved a certain ideological predominance in the organization of HIV knowledge and intervention globally in recent years, and both can be linked to a certain biomedical individualism, insofar as they institute biological and individual-level factors as overriding concerns at the expense of more systemic and relational analyses. This biomedical individualism in turn can be understood to be premised upon what Latour (1993) has called the “modernist settlement,” which separates out human subjects from the devices and technologies that might otherwise be understood to be constitutive of actions, subjects, and ethical capacities, as I will demonstrate here. Bioethical disciplines can only adequately address this scene of biopolitical production, I will argue, by undergoing a certain critical reframing that effectively and carefully counters such biomedical individualism—in particular its narrow attribution of agency.

The paper has three parts. In the first section I discuss how the criminal law frames responsibility for HIV transmission. Drawing on an emerging body of scholarship and advocacy work, I argue that the criminal law does not transcend the social field, but is immanently involved in its production, including the production of HIV effects. Thus, while the use of criminal provisions to prosecute HIV transmission may seem like a neutral application of moral principles to individual action, it also serves to frame that action as individual, to the exclusion of other participants, relations, and processes that might otherwise be implicated in HIV prevention. I argue that the best way to grasp the ethics of criminal approaches is to conceive of the law as performative and to ask what its exercise in specific contexts does. Through these means, we are better able to see how the law participates in wider social dynamics, such as the exacerbation of stigma, avoidance of HIV testing, the singling out of infected individuals, and the creation of affective climates that paradoxically make it more difficult for HIV-positive people to be open about their status.

In the second part of the paper I consider whether legal discourse has the capacity to distribute responsibility for HIV effects beyond the actions of HIV-positive individuals and to take into account the operation of extended networks of actors, both human and non-human, in the materialization of HIV events. I discuss the case of Primary Health Care (2011)—in which a specific configuration and enactment of medical services played a part in a case of sexual transmission. I argue that questions of distributed agency and responsibility are pertinent to contemporary scenes of sexual negotiation, where a range of technical knowledges, devices, procedures, and assumptions may come into play, ranging from the results of HIV diagnostic and viral load testing to antiretroviral drugs and their actual or presumed preventative effects. A focus on performativity extends the scope of ethical discourse by asking that we take into account the social performances of a whole range of actors, both human and non-human, when determining responsibility for HIV events.

In the third section of the paper I bring this reframing of responsibility—such that it encompasses more than just the human actor—into articulation with questions of biomedical prevention, which see a range of biomedical devices invested with HIV prevention potential. I focus in particular on the procedures used to evaluate these devices, such as the randomized control trial. What is often missing from these procedures is a consideration of what these devices may do in the world beyond the frame of evaluation: in other words, their constitutive effects. To some extent, these effects are impossible fully to predict, and herein lies the challenge. This raises the responsibility of those issuing these devices and procedures to engage those affected by them with regard to how they are enacted: an ongoing training of the attention that I outline as responsive attentiveness.

At one level, this essay attempts to intervene in the processes through which medical consumers are made exclusively responsible for the unaccounted effects of biomedical interventions. By drawing on notions of hybrid action from science studies, we can build an alternative understanding of public health responsibilities that avoids these disciplinary implications. The argument also diverges from the pre-emptive role that has tended to characterize social scientific responses to biomedical developments. Instead, I favor a training of the attention to processes of emergence and the cultivation of extended relations of ongoing responsiveness. I learned some of these lessons about experimentation and experimental responsiveness as a drug-using, HIV-positive gay man. Experimentation may be deemed necessary for changing a situation, though it has all sorts of knock-on effects, some of them dangerous. This is no reason for despondency or attempted purification. Addressing these circumstances actually requires us to enter into new, more attentive techniques, relations, and experiments in reframing.

Performativity of the Law

What I would like to open up initially in this essay is a consideration of the performativity of the law. This is by no means a new consideration for those concerned with public health in the context of stigmatized and criminalized activities such as homosexuality and drug use, but it does provide a way of situating the law in relation to key health practices—such as HIV testing—as well as a body of social theory that may help to articulate the issues in new ways. Criminal prosecutions for various activities associated with HIV risk—including nondisclosure of status, exposure to HIV, and onward transmission1—have increased globally in recent years in tandem with biomedical advances in HIV monitoring and treatment. This has included the implementation of criminal legislation in several jurisdictions that creates new offenses related to the onward transmission, or potential onward transmission, of HIV.2 It is not clear why this trend has emerged. It seems that the more identifiable and clinically calculable HIV disease has become at the level of the individual body, the more individual bodies have been rendered responsible for the risk of onward transmission (Race 2001). What is clear is that a complex intermediation exists between biomedical and legal domains, which, while differently configured in different contexts, invariably creates new subjects of HIV prevention: that is, new ways in which persons become subject to the moral principle of HIV prevention (whether through force of law, through commitment to community, as one whose body is positioned as posing a threat to the public health, as one who might presume to be protected by the law, etc.). (On modes of subjectification, see Foucault 1985). These subject positions may or may not be viable, and they may or may not be well adapted to the practical challenges of HIV prevention.

Those concerned with public health in this context have found it necessary to consider the law’s effects in relation to a range of social outcomes, such as HIV transmission, access to services, HIV testing practices, drug harm, and so on. And they have found it necessary to question criminal approaches to HIV prevention on the basis of the effects, or likely counter-effects, on certain socially desirable outcomes. Thus, extensive efforts have been made in recent years—by HIV activists (Bernard 2011), researchers (Worth, Patton, and Goldstein 2005; Weait 2007; Burris and Cameron 2008; Mykhalovskiy, Betteridge, and McLay 2010), and international agencies such as UNAIDS (2008)—to identify and communicate internationally the negative public health effects of global trends to criminalize HIV transmission. These efforts seek to force an evaluation of the law on public health grounds—an evaluation to which the law appears partially, albeit insufficiently, receptive.

To ask about the performativity of the law is to ask a series of questions about whether and when the law does what it says and whether these doings are desirable. Certain statements or enactments of the law might be said to be successful, bringing about the social reality that they name, binding subjects in desired ways. And certain enactments or statements of the law might be seen to misfire, producing circumstances in which the law does not quite do what it says, or worse, can be seen to be participating in processes that undermine some of its express aims. J.L. Austin (1975) might approach these latter instances of performativity as perlocutionary instances: They do not play out in expected ways. Tracking what they do do is an empirical matter, prompting questions such as: What does the law do, in given conditions and circumstances? What might it do, in foreseeable conditions? How does the law participate in the making of worlds that depart significantly from certain of its express aims? In what conditions is the law, as it is enacted or practiced, effective or counterproductive?

To some extent, criminologists have a way of thinking about these things, which takes place under the rubric of deterrence. But deterrence is only one of the purposes of criminalization, legal scholars say. Other purposes include punishment, denunciation, rehabilitation, incapacitation, and restitution. Within criminology generally a growing body of work nevertheless attempts to calculate the deterrent effects of criminal sanctions, both generally and specifically (Dölling et al. 2009). These studies represent a desire on the part of legal specialists to render the law calculable in terms of its impacts and effects—a desire that might initially appear to provide some ground with HIV prevention specialists. The focus on deterrence in terms of individual behavior does, however, tend to presume a rational choice actor—the sort of actor that may be difficult to presume in the case of the sexual and/or drug practices most associated with HIV infection. A performative approach, by contrast, would seek to understand how a range of effects (including behaviors) are collectively and interactively produced through the performance and interaction of a diverse range of elements, extending beyond the individual subject, including (but not limited to) the exercise of the law.

Critics of criminalization have cited a lack of evidence that criminal sanctions effectively deter risk, and they have pointed to a range of potential effects that militate against the wisdom of criminalization. These include the propensity of criminal sanctions to discourage HIV testing; to generate mistrust between patients and health service providers; to intensify fear and stigma; to undermine the supportive environment that is needed to stop the spread of HIV; and to endanger infected individuals—especially women in contexts of maximum stigma where they are more likely to be diagnosed first by virtue of their closer engagement with health services such as antenatal testing (UNAIDS 2008). Critics have also pointed to the law’s potential to maintain a false sense of security among HIV-negative and untested individuals by framing a shared risk or activity as the sole responsibility of diagnosed (if not infected) individuals.

It is this latter possibility upon which I would like to expand, for it reveals how responsibility for HIV transmission is always framed in some concrete way. By virtue of this framing, certain actors (such as HIV-positive individuals) are rendered culpable for HIV transmission, while certain others (in this instance negative and untested individuals) are rendered as externalities—that is, considered external to the culpabilities at hand. The manner of this framing not only raises ethical issues with regard to the equitable distribution of responsibility for HIV prevention. It is also performative: It participates in a series of concrete and practical effects. Specifically, it produces HIV-negative and untested individuals as not responsible for HIV transmission, and it may fail to protect them by fostering expectations that the law is efficacious; for example, that HIV-positive individuals will disclose their status in circumstances where much of the evidence suggests that this is generally not likely to be the case.3

This framing of responsibility enacts the HIV-negative individual as an ill-equipped or unsuspecting sexual actor—or so research among homosexually active men in the United Kingdom would suggest (Dodds 2008).4 HIV-positive people are framed as the primary agents capable of transmitting HIV—or preventing infection—while HIV-negative and untested individuals are dissuaded from testing or sustained, somewhat impractically, in the “practical belief” that the law will protect them and/or that HIV-positive sexual partners will automatically disclose their status. From a certain perspective, this operation of the law might be understood as an instance of effective performativity, in the sense that the law does what it says: It renders HIV-positive individuals responsible for HIV transmission. But this rendering, or this framing, could be seen to be contrary to the goals of HIV prevention, insofar as it produces sexual risk as an individual rather than shared responsibility and thus lends itself to other culturally prevalent but characteristically avoidant defensive strategies, such as the fantasy of HIV-positive predation—“the evil perpetrator”—as a default explanation for HIV infection.5 In binding certain social actors, the law can be seen here to be unbinding certain other participants from responsibility for HIV prevention. It thus becomes necessary to subject this unbinding to interrogation, precisely on the grounds that constitute a broad aim of the legislation.6

It is worth reflecting here on how the performativity approach proceeds. It evaluates the law as one element in the materialization of a range of effects that include, but are not reducible to, the general question of deterrence or prevention. If deterrence is concerned mainly with the impact of criminal sanctions on rates of future commissions of the deed, the performative approach treats HIV transmission as an event, or an effect, for which a more complex and extended account can be supplied.

Of particular significance here are the affective repercussions of the law in relation to certain other significant considerations, such as presentations for HIV testing, the meanings of which change in the context of punitive legislation and other forms of identifying surveillance. The HIV test may come across as detrimental to well-being in a context where diagnosis is associated with further stigma and the social positioning effected by criminalization. Here the law can be understood to participate in the production of what I have described elsewhere as an “affective climate”: a shared context of fear, shame, secrecy, suspicion, rejection, and avoidance (or conversely, trust, hope, care, reciprocity, and openness), which historically and materially accumulates to make care practices more—or, in this instance, less—possible (Race 2010). There is a name for this concern in the field of health promotion; it is called “creating a health-enabling environment.” I am suggesting that we need to think more carefully about the production of affective climates of HIV prevention and in ways that move us beyond the tokenism that sometimes infuses the enunciation of important policy goals such as “reducing stigma.”

Extended Responsibility

This argument approaches the criminal law as a device that frames responsibility for HIV effects. It suggests that the criminal law performs in ways that may inadvertently undermine public health. But what alternatives to this manner of framing exist within the law? In particular, what capacities exist within juridical discourse to conceive the participation of a wider range of actors, both human and non-human, in undesirable effects such as HIV infection? Key to this suggestion is the argument from science and technology studies that agency cannot be exclusively contained in human being, nor even in the norms, values, and discourses which are assumed to animate human behavior. Action is hybrid, in the sense that it takes place in collectives comprising human beings as well as material and technical devices, texts, objects, and so on (Latour 1993, 2005; Callon 1998). Such devices should not be regarded simply as supplementary to preexisting bodies, minds, and their related capacities. They actively shape them and must in this sense be considered constitutive of subjects and actions (Braun and Whatmore 2010).

The criminal law tends to evoke a humanist understanding of action, with its focus on the intentions and behavior of human subjects presumed sovereign and autonomous. We nevertheless see glimpses of how biomedical devices and technologies have been understood to play a part in the relevant forms of action and cognition; for example, in recent disputes over whether the reduction of viral load associated with antiretroviral therapy might mediate criminal culpability for risk activities. In 2008, a Swiss HIV Advisory Committee formulated the “Swiss Statement,” a document proposing that, for treated patients for whom blood viral load was suppressed for six months, unprotected sex with an informed partner was acceptable (Vernazza et al. 2008). The statement was controversial and generated intense debate internationally. What is less well known is how the statement sought to intervene in trends around prosecution for HIV exposure under the Swiss penal code. The original report makes this point explicit, stating that unprotected sex between an HIV-positive individual on antiretroviral treatment and an HIV-negative individual ought not qualify as an attempt to “propagate a dangerous disease” or “engender grievous bodily harm” under the provisions of the code in question. The controversy escalated in 2010 when the Swiss parliament expressly rejected a call from the Swiss Federal Commission for AIDS-related issues that it take into account scientific findings on the infectiousness of HIV-positive persons using treatment when enacting criminal legislation. Even in these circumstances, the Swiss assembly undertook to maintain the offense of negligent exposure, which under the Swiss code remains an offense irrespective of HIV transmission.

This controversy can be approached as a “hot situation,” which Michel Callon describes as a situation in which “everything becomes controversial: the identification of intermediaries and overflows, the distribution of source and target agents, the way effects are measured” and one which usually involves “a wide variety of actors” putting forward “mutually incompatible descriptions” (Callon 1998, 260). (With regard to sexual relations, Callon is not referring to that sort of “hot situation” necessarily, but perhaps the description is just as apt for such purposes!) In this case, various parties—legislators, scientists, policy-makers, HIV activists—dispute whether the presence of certain technical practices (treatment and viral load testing) alters the risk and thereby the character of the relevant transactions and judgments. It is worth pausing at this point to draw out certain tensions in the general legal reasoning that animates such decisions. While some technologies, such as HIV antibody testing, are taken to go toward the production of a culpable subject required by the law, certain other devices, such as antiretroviral therapy and viral load testing, are discounted in their ability to mediate the character of the relevant transactions. The law would appear to be selective here in its determination of which technologies can be taken to constitute moral subjects and mediate the effects and intentions of bodies. Where the antibody test gives rise to criminal responsibility, the law reverts to its default position when it comes to viral load testing and dismisses the test’s capacity to reconstitute bodies.

And yet within legal discourse, there are recent decisions that suggest a greater capacity to account for extended networks of hybrid action in the occurrence of HIV transmission. In a recent Australian case, the question of responsibility for HIV transmission was found to be inseparable from how various technologies and procedures interacted and performed across a relay of clinical and everyday domains. The case is a civil law rather than criminal case, which of course involves different jurisdictional aims and principles. Unlike the criminal law, the law of torts is concerned with compensation for harm and the distribution of responsibilities, rather than the adjudication and punishment of offenses against the state. However my aim is not to offer legal commentary on discrete jurisdictions, but to explore some of the instituted rationalities that are currently used to adjudicate questions of responsibility with respect to HIV infection.

The case concerns a woman who unwittingly infected her partner with HIV (Hall and Wallace 2011). She was the patient of a medical center, of the sort that is common within contemporary health service provision today: a busy medical center that receives back-office support and pathology services from a large corporation, Primary Health Services. The woman attended the center and underwent an HIV antibody test, which was handled by Primary Health Services. The test was inconclusive, but the recall letter advising urgent retesting was sent to an old address, the reception staff having failed to follow the company’s procedure and check the patient’s current address. When the woman returned to the medical center three weeks later to pick up her test results, she was seen by another doctor, her original doctor being unavailable. This doctor pulled up her records on the computer and gave her the all-clear for HIV, but had missed a note on the computerized record advising of the patient’s need for a repeat test. A week later the woman had unprotected sex with her partner, who became infected with HIV. The medical center doctors admitted liability and the case was brought to determine the liability of Primary Health Care, the corporation that ran the back-office support, to establish its duty of care.

The case illustrates the extended chain of relations and coordinated action that characterizes the scene of health care provision today. It reveals the highly contingent mix of practices, procedures, devices, circumstances, and occasions of cognition that can make up an instance of HIV transmission. The case provokes anxiety around the extent to which we can be considered responsible for the elaborate forms of coordinated action that constitute everyday practice. Such practice routinely involves participation in complex assemblages of human and non-human actors, the components of which are fallible though their interaction generally goes unnoticed. It is impossible to account for the circumstances of this case without thinking about the performances of a diverse array of actors: doctors, patients, computer programs, clinical assays, sexual partners, reception procedures, misdirected envelopes, labor conditions in doctors’ offices, beliefs, practices and understandings. “The technicity [of these elements] is such that they can be combined and deployed in relation to countless other elements, gestures, practices and institutions” (Braun and Whatmore 2010, xxi). What is revealed is the contingency and unpredictability of these relations—their eventfulness, their ever-present margin of indeterminacy.

In deciding that the corporation had a duty of care to patients, the case is effectively determining which assemblages of human and non-human actors will be considered agents of HIV infection, with implications for all the other elements involved. Certain actors become implicated in the event of HIV transmission, while certain other actors are framed as externalities; considered external to the circumstances at hand. This framing has implications for future performances of agency and responsibility: It constitutes subjects of HIV prevention, even as these framings are always available to contestation and overflowing. But the implication of human action with technologies ought not to provoke humanist despair or nostalgia; we have always been technically constituted in particular ways. Rather it calls for responsive attentiveness to the practical dynamics in question: the specific arrangements and relations we find ourselves in, which also happen to effectuate our capacities and actions.

Framing Effects: The Limits of Efficacy

So far I have considered the performativity of the law in relation to HIV events. I argued that the law acts as a framing device that configures responsibility in particular ways that, while consequential, are not essential: Things can always be organized otherwise. I went on to illustrate the capacity of legal discourse to grasp the work of hybrid assemblages consisting of human and non-human actors in the production of inadvertent and harmful effects. From this example, it becomes possible to consider HIV events as the outcome of extended networks of activity involving human and non-human interdependencies and co-implication, the effects of which may be difficult to anticipate. I now want to bring this discussion into more explicit articulation with questions of biomedical prevention, which are currently reconfiguring the field in various ways. This is a pertinent concern, because the issue of who will bear responsibility for the unaccounted effects of these technologies is still active and undecided. “Biomedical prevention” refers to the use of certain biomedical technologies and procedures for the purposes of HIV prevention, the implementation of which is a key theme of international HIV policy speculation currently. Alongside procedures such as male circumcision (which has been found to be partially efficacious as a form of protection for heterosexual men), biomedical prevention consists mainly of various new applications of antiretroviral therapy: its use among HIV-positive individuals to achieve viral suppression some time before therapeutically indicated (termed Treatment for Prevention); among HIV-positive mothers to prevent mother-to-child transmission; among at-risk HIV-negative individuals as a form of pre-exposure prophylaxis (PrEP); and as part of a topical preparation to be used by HIV-negative individuals during intercourse (topical microbicides, most of which are still under evaluation). These interventions are evaluated in various populations around the world using the procedure of the randomized controlled trial, considered the gold standard of medical evidence. This procedure measures the efficacy of an intervention in a control group over a set time period through a process of random assignment and comparison.

I would like to suggest that the randomized control trial is also a framing device. It configures responsibility for HIV effects in certain ways. Typically this framing stops at the body of the individual trial participant, or sometimes the partnered couple, insofar as it is organized around questions of prevention efficacy. This operation is designed to render the pharmaceutical agent an independent variable, the causal activity of which can be evaluated in linear terms. So where the criminal law aims to isolate the human subject in its framing of responsibility for HIV events, the experimental setting attempts to isolate the activity of the non-human actant—i.e., the intervention—in order to draw causal inferences.

What is externalized from this frame is the possibility (and actual occurrence) of other effects that follow from the intervention, whether good or bad, predictable or unpredictable. Biomedical interventions are also cultural interventions, and they always participate in knock-on effects and overflowing. Such overflowing could take the form of beneficial effects, such as the increases in HIV awareness and self-efficacy observed among trial participants in some clinical settings. Or it could take the form of less beneficial effects, which concepts such as “risk compensation,” “disinhibition,” “typical versus ideal adherence rates,” “drug resistance,” and “transmission of drug-resistant virus” attempt to name in an emerging terminology of apprehension. These events are typically produced as externalities: beyond the ambit of clinical research’s concern and responsibility. Significantly, they consist not only of possible changes in behavior, cognition, expectations, and interaction, but also the creation of new organic structures such as drug-resistant viral strains that may be difficult to treat with available pharmaceutical agents.

In this respect, the randomized control trial is only a partial framing.7 While it stands as a powerful mechanism for isolating the short-term activity of an individual intervention, it can rarely predict what will happen in the real world, when the intervention encounters other actors and relations. This is partly because the trial is conducted over a finite period, and the independent variable, the experimental candidate, is often dependent on other factors for its efficacious enactment. But it is also because the effects of interest are pre-specified in the construct of “efficacy,” which represents the measure of the intervention in trial conditions (which are themselves tightly controlled in order to eliminate other variables).8 While the construct of efficacy is generally considered necessary for the experiment, it cannot possibly register all the different processes or effects the experiment puts in motion. Yet in policy discourse, efficacy tends to trump all other considerations, instituting some degree of neglect toward the local and material relations that would need to be engaged to achieve the necessary outcomes or that might otherwise be affected by the intervention. Some indication of the relevant relations can be taken from a recent report on a sample of patients treated with HAART (highly active anti-retroviral therapy)—itself a “proven intervention”—in Lomé, Togo (Dagnra et al. 2011). After just 1 year of treatment, 30 percent of these patients had experienced virologic failure, and drug-resistant mutations were found in 46 patients, corresponding to 25 percent of the patients enrolled in the study.

In health discourse, such events are typically attributed to individual behavior in the moralized language of noncompliance. (For an extended analysis of the discourse of compliance and its ideological effects, see Race 2009). But in this instance it is clear that these events are not attributable to patient fallibility alone. The authors point to other features of the hybrid collectives through which care is routinely delivered in this setting: inadequate laboratory facilities, lack of virological monitoring, shortage of qualified personnel, unreliable supply systems, inadequate storage. The effects of the drugs are contingent on the relations they enter into, which are always particular, as these circumstances make plain. How should we account for mutations such as these or their likely transmission? How would the experimental ethic differ if responsibility were more extensively configured?

My point is not to denounce the products of randomized control trials—many of us would not be here without them. Nor is it to deny the use of antiretroviral medications for prevention purposes—these drugs may offer something that some HIV-negative people, not to mention HIV-positive people, need for their lives and to prevent illness. Rather, my aim is to open up the question of how best to take account of this wider range of variable effects in the process of their making. Specifically, what research practices, what modes of attention, might be devised to attend to these events-in-process more adequately?

Responsive Attentiveness: A New Frame for Social and Cultural Research

In the wake of proposals around biomedical prevention, a host of research practices have come to the fore. One of these is epidemiological modeling, which sets out to depict what might happen when a range of variables come together or when interventions are implemented in this way or that. The recent prominence of modeling is interesting, because it represents a partial intuition that the findings of efficacy delivered by randomized control trials can provide only a limited window onto the real-world activity and implication of drugs. But while these models accrue great power to enact different futures, the models themselves are only as good as the assumptions that inform them, which of course are always social interpretations and political decisions. Alongside modeling, there is growing interest and investment in the question of how best to implement or operationalize the interventions that control trials have endorsed. “Implementation science” takes a more observational form, with a focus on the rigorous definition and replicability of all the components of an intervention. But alongside this focus on prediction and replicability is another question: how to account for indeterminacy? This question is beyond the reach of the population and medical sciences. Nancy Padian and her colleagues come close to acknowledging this point when they advise that “careful monitoring of what happens in the community when the intervention is scaled up will be essential” (Padian et al. 2008, 10). In the end, though, the research strategies her team proposes have no way of engaging the understandings, meanings, and affective responses of the groups most subject to the relevant interventions. Whose monitoring of what happens will be taken into account, and how will this be made to matter? How have the targets of the intervention been engaged in the construction of “the problem” and its possible solutions; how has their agency been valued? It is this “what happens” that I am concerned with, which can never quite be answered by the predictive sciences. It represents the space of collective participation and experimentation we are all responsible for. How can scientific action be reframed to acknowledge and even embrace this indeterminacy? (On indeterminacy as a space of democratic participation and shared futures, see Diprose et al. 2008). In more theoretical terms, how can we be faithful to the unknowns released into the field of possibility by prevailing regimes of knowledge?9

I want to conclude then with a proposal for social research that I think these reflections help clarify. Some have argued for greater involvement of social scientists in trial conduct and design, and while I think this is useful and important, it tends to position social science in a predictive and ultimately pre-emptive role and it is not precisely the practice I am advocating here. Rather, I think we need to articulate a new frame for social research. We need to devise methods and practices of responsive attentiveness to the unpredictable hybrids that will inevitably emerge from this scene of biomedical, social scientific, and corporeal production.10 This might sound rather abstract, but it will consist of concrete research practices that will be experimented with over time and modified. Responsive attentiveness involves taking an interest in how the practices of medicine and the practices of bodies alter each other in the process of their making and coming-together. Just as a body is a different entity in the presence of a pill—bearing different capacities and attributes—a pill is a different entity depending on how it is enacted. Neither of these actors—medicine nor persons—is fixed or static; they become with one another. We must be alive to the range of differences that each is making in and for the other. Responsive attentiveness is different from forms of action based on prediction and predictability, which presume to know everything there is to know about a situation in order to enter it responsibly. And it is different from the civil law notion of a “duty of care,” which attempts to determine what one should have known, but in the process gives rise to a general disposition of litigiousness. Responsive attentiveness involves a responsibility to what one cannot know: the risk of being altered by a situation. This represents also a space of possibility: It involves giving oneself over to a shared but indeterminate future; in this case, biomedicine and the law acknowledging (and perhaps even welcoming) their implication in the theater of social interaction and their necessary responsiveness to it.

Certainly, this means better attention to how interventions are given meaning by their participants and impact upon social relationships (Nguyen et al. 2007). But what is also at stake here is how to intervene in a situation in which end-users become responsibilized for the unaccounted effects of biomedical products. 11 In the top-heavy world of globalized biomedical prevention, users are not simply configured by technologies (Woolgar 1991) but will be responsibilized for their efficacious enactment in a manner that shifts responsibilities from biomedical providers to consumers.12 This is why devising approaches that can better conceive how medical devices and objects actively participate in the shaping of “human” practice is urgent and essential. To this effect, I have tried to draw attention to the calculative practices that constitute the field and that license such objects and devices. This is to suggest that the same forms of critical scrutiny and engagement with which HIV social researchers are led to subject sexual and risk practices be extended to the practices of medicine and science. The international field of HIV/AIDS is extensively administered (you could say it has reached a point of administrative saturation!). But this administration consists of diverse agencies, which render HIV effects calculable in certain ways. These calculative practices are analyzable, they are ontologically consequential, and they can always be organized and articulated otherwise: In short, they are performative. How can these different practices and regimes of calculation—controlled trials, observational studies, the practical calculations of affected groups and individuals—be better articulated with one another and more responsively enacted?


The dynamics of HIV transmission are always grasped in some concrete way—by biomedics, population scientists, jurists and legislators, sexual participants, sociologists, psychologists, economists, drug users, epidemiologists, poets.13 Each of these disciplines consists of concrete practices: ways of knowing that can be analyzed. These ways of knowing are hierarchically organized: Some are given more truth-value than others. The findings of each are consequential—though the practices involved in coming to these findings can always be organized or articulated otherwise. Each discipline inevitably has blind spots, affects, and overflowings—they are in this sense “excessively” concrete. What are the effects of these framings, both alone and in combination? How can we bring them into better articulation to bring an end to the epidemic and greater well-being among affected populations?

Here I have sketched some possibilities for an extended account of agency, and the attribution of agency, in the HIV field. These attributions can be studied, and they must be studied if we are to take collective responsibility for the ongoing dynamics of HIV/AIDS. Social scientists and affected individuals are part of this collective, as are our biomedical colleagues, and so are the devices and calculative practices in which we variously engage. The question that emerges from this reformulation of ethical discourse is: How might we account for HIV effects—practically, conceptually, empirically—so that we can better attend to the present epidemic? Keeping this question open is a framing responsibility.


The relevant legislation and case law are highly variable internationally and from jurisdiction to jurisdiction. Where some jurisdictions criminalize only intentional HIV transmission, others, such as Switzerland, criminalize exposure to HIV whether or not transmission has occurred, while others—such as Ontario, Canada, and New South Wales, Australia—maintain nondisclosure of HIV-positive status in sexual contexts as a criminal offense.


For example, following the creation of the “African Model Law” by the (USAID-funded) Action for West Africa Region group in 2004, 13 African countries proposed HIV-specific criminal laws and at least seven passed them. A number of countries in Eastern Europe and Central Asia enacted laws that criminalize HIV exposure or transmission over the same period (Bernard 2008).


Numerous social and behavioral studies among various populations at risk internationally indicate that HIV disclosure is rare in many of the contexts in which sex and/or injecting drug use takes place. Worth, Patton, and Goldstein offer the following succinct commentary: “The very idea of disclosure assumes that sex takes place face-to-face between partners of equal power and with some impetus to speak” (2005, 10).


In this survey of U.K. gay and homosexually active men, Dodds (2008) found that support for criminal prosecution of HIV transmission was strongly associated with a range of indicators of HIV prevention need, including never having had an HIV test and believing that HIV-positive sexual partners would disclose their status before sex.


In Western countries, the legal reproduction of this myth plays out in racialized terms, where a disproportionately high number of those charged with HIV transmission offenses are heterosexual men of African origins, typically characterized in terms of sexual excessiveness and hypersexuality in media commentary surrounding their cases. See Persson and Newman (2008).


Of course, this is not to suggest that HIV-positive people do not bear some responsibility for HIV prevention, but, rather, to question whether the criminal law is an appropriate mechanism for fostering this capacity. An alternative approach to the further responsibilization of HIV-positive individuals pursued by criminalization would be to educate HIV-negative and untested individuals about the different sexual scenarios they may find themselves in and the constraints affecting differently sero-positioned participants in these scenarios.


For groundbreaking work that grapples with the bioethical implications of this claim in relation to HIV pre-exposure prophylaxis, and which informs my thinking here, see Rosengarten and Michael (2009, 2010).


For this reason, HIV social scientists have increasingly come to insist on a distinction between efficacy and effectiveness. See Kippax (2003).


For Badiou (2006), enacting fidelity to the event involves performing a “generic procedure,” which in its undecidability is necessarily experimental and potentially recasts the situation in which being takes place.


I adapt this phrase from Donna Haraway (2011), who uses it to refer to the practice of taking care for the unpredicted terrain that emerges from technocultural production.


As Woolgar (1991) has shown, the process of technological design incorporates specific presumptions about human capacities and behavior, which then configure users of the technology in certain ways. Here I am highlighting the way in which this process shifts responsibilities from producer to consumer, in a manner that has political and moral implications. It is known, for example, that general medication adherence rates typically stand at about 50 percent, far less than the rate of adherence that antiretroviral medications require in order to achieve the desired effects. But this knowledge tends to be bracketed from trial design and/or publicized statements of efficacy, which renders adherence a uniquely “human” or user responsibility.


Barry Adam (2011) and Cindy Patton (2011) have each pointed to the disconnect between the technical or population-level analyses that currently underwrite major policy decisions and the practical concerns and insights of affected groups and individuals. In this paper I have suggested that this technical situation deserves a technical response that may be formulated most effectively in terms of framing. As Isabelle Stengers and Olivier Ralet put it: “It is always possible to maintain that [a given] solution is a solution to a problem that is technically badly formulated, that is, to a problem posed according to certain a priori imperatives that have resulted in handing over control to some experts and ignoring others” (Stengers and Ralet 1997, 218–219, emphasis original). In this context, the relevant expert-others are those who contend with the practical realities of HIV infection and HIV prevention in their local environments.


For an argument about poetry’s capacity to register the affective dimensions of the epidemic in a manner that gives insight into the ways in which HIV prevention and testing campaigns are apprehended in the context of political and historical inheritances, see Hoad (2010).



This paper was first delivered as part of a symposium of Biomedicine and Subjectivity organized by Peter Aggleton and Carlos Caceres for the 1stInternational HIV Social Science and Humanities conference held in Durban, South Africa, in June 2011. It was developed for publication over the course of special study leave granted by the University of Sydney, which I spent as a visiting scholar at the Center for the Study of Gender and Sexuality (CSGS) at New York University. I would like to thank my hosts at CSGS for their constructive and generous responses to this work as it emerged—in particular Ann Pellegrini and Robert Campbell. Thanks also to Judith Auerbach, Martin French, Susan Kippax, Gail Mason, Marsha Rosengarten, and audiences at the Center for AIDS Prevention Studies at the University of California, San Francisco; the Institute for Research on Women and Gender at the University of Michigan; the Center for Human Rights and Justice at the University of Texas; and two anonymous reviewers of this journal for their discussion and suggestions. All errors are my own.

Copyright information

© Springer Science+Business Media B.V. 2012