The Ethical, Epistemological, and Conceptual Need to Resume Fieldwork

The ongoing Covid-19 pandemic has provided an additional justification—the protection of researchers and their interlocutors—for already existing but ethically, epistemologically, and politically problematic research practices. In fact, as Aymar Nyenyezi Bisoka and Kanisha D. Bond, Milli Lake and Sarah E. Parkinson rightly point out, the novel coronavirus only partially transforms the context for researchers (and individuals) already living in crisis contexts rife with other immediate risks. Moreover, many countries around the world, in the Global South and North, continue to struggle to enforce social distancing measures, which greatly limits the relevance of many arguments for protecting individuals. Building on the “Social Research and Insecurity” theme’s essays, which highlight the challenges posed by Covid-19 in conducting fieldwork, our hypothesis is that the pandemic, or more precisely the practices it entails, risk exacerbating the neoliberal bureaucratization of research that has played a role for several decades in what is—possibly—a decline of the social sciences.

Subcontracting empirical work

For several decades, the dominant trend in research has increasingly been one of uncoupling the “collection” (i.e., production) of data, from its analysis and theorization. This pronounced phenomenon is more a function of institutional power relations than of epistemological debates. Researchers at Western universities are facing increasing difficulties in pursuing field studies. Contrary to the often-stated justifications, this state of affairs does not stem from a greater danger of field work for researchers or their interlocutors compared to previous decades. Instead, it’s an outgrowth of the proliferation of regulatory measures and enforcement bodies (e.g., Institutional Review Boards, Research Ethics Boards), often taking a page from the field of medicine, which grew out of a neoliberal bureaucratization of academic institutions whose principal elements include: an entrepreneurial model of research through project funding, limitations of legal risks for the institution, increasing dominance of administrators over academics, precarious statuses, and a pressure to publish that favors databases over fieldwork. In practice, this evolution has increased restrictions on researchers’ travel and diminished opportunities for carrying out long-term field studies, conducting interviews, and recording observations. This has resulted, at least in the subfield of civil war studies,1Adam Baczko, Gilles Dorronsoro, and Arthur Quesnay, “Prolegomena: For a Sociological Approach to Civil Wars” in Civil War in Syria: Mobilization and Competing Social Orders (Cambridge University Press, 2018), 1–24. in the proliferation of mediocre studies and artifacts related to misused databases. In this regard, our stance is that oversight bodies must be composed of scholars from the same discipline, who themselves have a practical knowledge of this type of research.

This state of affairs has led to data production being entrusted to nationals of the countries studied with increasing frequency, in a significantly different way from the contribution by privileged informants in ethnographic research, for example. The data is produced by companies (sometimes legally chartered NGOs), local academics, students, and journalists, treated as “Mechanical Turks” and therefore not credited for their work. In similar fashion to research on the Democratic Republic of Congo (where local researchers are mobilizing around this issue), this is the case for numerous articles published on the Afghan and Syrian fields, as well as the conferring of doctorates by prestigious universities with no or merely symbolic fieldwork having been carried out. Several companies have been started in Kabul or Gaziantep (in Syria) that mix consulting, humanitarian, and outsourced data collection work for international institutions, including universities.

However, despite putting out impeccably formatted products, the data these contractors gather are not always reliable. No wonder, since the veracity of the questionnaires are most of the time unverifiable, especially when they pertain to difficult-to-access areas (e.g., Afghanistan’s Helmand province) and data furnished by large institutions (including the military) are partly censored or irremediably biased.2We refer here to Anne-Marine Vanier’s ongoing research on the constitution of an expert field on armed conflicts based on the Iraqi and Afghan cases, of which first results were presented at the 2019 congress of the Association Française de Science Politique. Anne-Marine Vannier, « Regards sur l’atelier: La sous-traitance de la collecte des données sur les conflits » (unpublished paper, 2019).

In addition, organizing research in this manner induces a transfer of risk-taking from Global North researcher to these contractors who are treated as petites-mains (“little hands”). Various ethical watchdogs that prohibit their researchers from going into areas perceived as dangerous authorize these contractors without batting an eye. Despite some established researchers criticizing these methods—for instance, as Scott Desposato did in 20143“In many developing countries, experimental research resembles a Wild West where almost anything is possible.” “Ethical Challenges and Some Solutions for Field Experiments∗” (unpublished manuscript, November 3, 2014), 2.—these well-known practices are still treated as par for the course. This implies a fundamentally inconsistent argument that encompasses other considerations: a form of implicit hierarchy between individuals according to their location (mixed with their academic affiliation), a hardening of North-South power relations within the world of research, and a return to a division of scientific work that has not prevailed since the 1950s. On this last point, it is notable that the practice of fieldwork tends to empower researchers, lets them produce their own hypotheses, and puts distance between themselves and the sorts of Taylorism making a strong comeback.

Less noticed, however, is that outsourcing is not limited to data collection but extends to data analysis. In particular, quantitative surveys, which generate large volumes of data, require hiring cheap labor to do the coding; in the universities of the Global North, this can mean students paid minimum wage. This practice treats data relevance and its categorization as a mere bureaucratic task detached from making a theoretical decision. Besides the problematic nature of exploiting precarious workers, rationalizing the research process in this way negates the theoretical nature of these essential stages in the work of conceptualization. By offering a new pretext for decoupling theorization from data collection, the current pandemic risks aggravating a disturbing decline in the social sciences under the particular influence of the dominant neopositivist current in political science.4Baczko, Dorronsoro, and Quesnay, “Prolegomena: For a Sociological Approach to Civil Wars.” Indeed, the risk of catching the virus now justifies the preference for remote research methods to compensate for the lack of access or, more often, the reluctance of researchers or their institutions to venture into the field.

The (intellectual) risks of remote research

While remote research does reduce certain psychological or legal physical risks, it also multiplies the sources of flaws in empirical demonstrations and tends to neuter theoretical debate. The first risk factor stems from defining relevant research objects based on the accessibility of sources, which in practice means the internet, social networks, as well as remote interviews. However, as we have seen in the Afghan and Syrian conflicts, these methodologies allow researchers to work on certain issues (e.g., visual representation, texts, accounts of individual or family trajectories), but do not make up for the lack of investigation in context.5We develop this issue on the Syrian case in our forthcoming article, Adam Baczko and Gilles Dorronsoro, “The Necessities of Fieldwork. An Investigation in War-Torn Syria,” Bulletin of Sociological Methodology, 2021. Remotely collected data reveal certain facets of the social world and produce discursive corpuses that are amenable to being treated with controlled research protocols. However, these discourses are of little use without a description of (and therefore a familiarity with) the contexts that requires inquiries on the ground. Moreover, access to the internet is very unevenly distributed, and this type of methodology tends to exclude the marginalized, i.e., those whose discourses and practices are the most difficult to access. Furthermore, remote research often produces fragile data, and the reconstruction of causal chains cannot be based exclusively on decontextualized discourse, without empirical verification by the researchers themselves.6For more on this topic, see Christian Davenport and Patrick Ball, “Views to a Kill: Exploring the Implications of Source Selection in the Case of Guatemalan State Terror, 1977–1995,” Journal of Conflict Resolution 46, no. 3 (2002): 427–450; Yoshiko M. Herrera and Devesh Kapur, “Improving Data Quality: Actors, Incentives, and Capabilities,” Political Analysis 15, no. 4 (2007): 365–386.

Second, the idea that fieldwork can be interrupted for a few years requires buying into the notion that deferring research until the crisis (pandemic or political) has passed does not bring major consequences. It ignores that after-the-fact narratives are reconstructions (although they have undeniable intrinsic interests); it is an illusion to think that that they can fill the void left by the absence of direct observation and in situ interviews. In this regard, Covid-19 raises anew the question of the limits of discourse analysis in reconstructing situations, as illustrated by several examples of armed conflict research. For example, even though researchers were barred from travel to conflict-affected areas in Peru, this has not prevented a good deal of research on those areas from being published. Later, it became clear that some researchers reconstructed events from discourses, while others, who insist on constraints on postwar accounts, reveal the tensions at work in those discourses in the postconflict period. The same problem has been raised in the case of the Lebanese conflict and is doing so in Syria, because the war is already transforming the discourses of actors who reconstruct events and their meaning to conform with the changing context.7On Lebanon, see Isabelle Rivoal, « Ecritures suspendues, vies engagées. Traverser la guerre civile au Liban », Ethnologie Française 44 (2014): 503¬¬–512. On Syria, see Anna Poujeau, « Drama chrétien ou comment parler politique en Syrie », Ethnologie Française 44 (2014): 449–460. Experience with the decades-long Afghan conflict shows the daunting work of reconstructing actors, for example, for a provincial history. More broadly, relying on later words to describe the past poses the risk of empowering the discourse of actors, while speech remains a social practice that only makes sense in a contextualized way and in relation to other practices.

Too often reduced to a “collection” of data, and therefore possible to be outsourced without too many safeguards to companies, or even to students, fieldwork is actually a decisive moment in theoretical questioning. In particular, the sequence—hypothesis/conceptualization, data collection, analysis, publication—describes very poor social science research. In fact, the investigation itself is an essential moment in conceptualization and obligates, especially in crisis contexts, a more reactive modality of theoretical elaboration, with a continual interaction between hypotheses and data production.

On the one hand, hypotheses are partly produced in confronting unexpected, singular situations. The transition from correlation to causality seems impossible without an investigation that provides direct knowledge of contexts of discourses, facts, and inventions. This invention of reality in crisis situations means that fieldwork can never be reduced to mechanically verifying assumptions developed out of context from concepts formed for routine situations. Novel circumstances conversely imply the necessity of adapting hypotheses in light of surprises encountered in the field, fueling the much-needed serendipity of the researcher.

On the other hand, fieldwork protects against the risk of major sociohistorical narratives overwhelming the complexity of causal chains, while investigation allows reconstructing the contexts of action and of unactualized potentials. Teleological risk obtrudes all the more when the researcher relies on a monocausal explanations, such as natural resources, climate change, or identity conflicts. The risk of apprehending these contexts by inscribing them in a long trajectory or by arbitrarily focusing on archival traces leads to smoothing out the uncertainties, the indeterminacies that notwithstanding characterize situations of conflict.

Avoiding the privatization of knowledge production

For researchers from the Global North, the choice is not between the interruption of research and its resumption, but between the accelerated development of highly problematic research modalities on an epistemological and ethical level and the resumption of research in the field. In particular, it is misleading to present the current restrictions as provisional given that the coronavirus is a long-term problem and that at a minimum the prevalence of its effects may last two to three years or more. The multiplication of administrative constraints and the shift in the balance of power—over who decides on the fieldwork of researchers—are likely to be long-lasting, with implications in terms of limiting research, and particularly serious consequences for a generation of PhD students who will complete their dissertations without doing fieldwork. Here, the health crisis also is a pretext for increasing the power of administrators over researchers, which in turn makes the collective mobilization of academics imperative.

In our case, the research we are currently undertaking on the Malian State cannot be conducted at a distance, which enables us to argue with our institutions (CNRS and Sorbonne University) the indispensable nature of our stay in Bamako. Indeed, the executives who run these Malian institutions, and thus constitute the core of our study, can only be met in Mali. Moreover, our research involves observations in the offices of these institutions and this professional setting is essential to contextualize interviews. Finally, on-site presence is essential to access the archives of state institutions. In Mali, as in many crisis situations, existing archives are not digitized or remotely accessible. Moreover, they are in a precarious situation due to the lack of resources to protect them, but also due to the voluntary destruction of administrative documents, which makes their consultation urgent.

The exceptional circumstances imply a protocol to control the risks related to the pandemic that takes into account the specific features of Malian society. Although the figures should be taken with caution, everything indicates that the pandemic is circulating in a limited way in Mali. However, Mali poses its own difficulties, as the wearing of masks and social distancing is not followed rigorously. Nevertheless, several measures can partially alleviate these difficulties in accordance with the recommendations of our academic institutions and international health authorities: avoiding public transport in favor of a private vehicle, conducting interviews in the open air or in open spaces, and the use of hand sanitizer before and after any interview and travel.

The resumption of fieldwork, while keeping in mind the specific situation of each field context, is indispensable. The growing tendency toward remote research increases the risk of accelerating the privatization of research underway since the 1980s, as the disengagement of researchers is paralleled by the increased involvement of private expertise that is problematic both epistemologically and politically. Research is often contracted out to companies, some of which have set up outposts in the most difficult-to-reach areas and make a practice of both collecting data and analyzing it for researchers.

But the problem goes beyond the issue of outsourcing. These private companies control a significant part of knowledge production, which the sponsors of the study then own. It is often forgotten that employees of these consulting firms, as well as journalists, have resumed their travels, while academics generally have not. The interruption of academic fieldwork does not signify a halt to knowledge production, but only the academic side of it, that is by legally noncommercial institutions, which are meant to make their results available to the public, including in the countries studied.

In the context of multiple crises, whether health or political, academics, thanks to their time and relative financial autonomy—at least compared to journalists and private consultants—have an essential role to play in deconstructing the dominant narratives and opening up a little the possibilities for decision-makers. In this sense, there exists an ethical imperative for resuming fieldwork as quickly as possible, while adapting it to the constraints imposed by Covid-19.