Indiana Jones & the Institutional Review Board: Disciplinary Incentives, Researcher Archetypes & the Pathologies of Knowledge Production

Academia is characterized by perverse incentives. Universities increasingly employ contingent labor and eliminate tenure-track faculty jobs.1 The mounting push for faculty to obtain external grants that generate overhead for institutions encourages scholars to pitch projects by adhering to specific methodological approaches, to adopt readily accessible, of-the-moment research questions, and to grant agency to narratives surrounding topics such as violent extremism or forced migration.2 Scholars at underresourced institutions, especially those in the Global South, experience this pressure in particularly acute ways, as a lack of institutional resources pushes many faculty to tailor their scholarship to the requirements of external funding sources and grants disbursed in foreign currencies. Faculty in many North American and European doctoral political science departments train students to compete in contracting academic job markets by claiming they can do it all: surveys, experiments, machine-learning, even “two weeks of ethnography.” At least some fieldwork has become a marker of topical and regional expertise, even as fewer and fewer fellowships and departments provide graduate students with the resources, training, and advising support necessary to spend extended periods of time away from their home institution conducting fieldwork.3

There exists an extensive literature on research ethics in political science and the social sciences more broadly, with specific attention being paid to appropriate practices for, for example, conflict- and violence-affected settings, vulnerable populations, authoritarian states, and archival, remote/digital, and desk research.4 Researchers have critically considered the limits of institutional review boards (IRBs), underscoring their role in protecting universities rather than acting as an arbiter of what is and is not ethical research.5 Scholars have increasingly emphasized the on-the-ground power differentials between researchers from the Global North and Global South as the complex and oft-obscured relationships between researchers and research facilitators.6 Journals have increasingly requested evidence of ethical consideration and/or review before accepting submissions, though many articles that do mention ethics do so in what  international relations scholar Johanna Rodehau-Noack and political scientists Stephanie Schwartz and M. P. Broache have termed a “procedural” way, focusing on IRB demands and skirting a broader consideration of ethical engagement with the field.7

Ethical and responsible frameworks for robust data generation, honest analysis, fair recognition of labor, and responsible publication must be promoted and rewarded by academic institutions, starting with mentorship.8 Mentorship, both by formal advisors and by more-senior figures and collective actors within a discipline (such as editorial boards and organized conference sections), transmits the norms, practices, and attitudes that shape research via teaching, informal discussions, feedback, and promotion. Yet Schwartz and political scientist Kate Cronin-Furman have shown, for example, that graduate students in political science receive very little if any education in fieldwork methods and ethics, even in top-ranked programs.9

The results are disturbing. In this essay, I identify three archetypes of researcher that have emerged in the current social science and particularly the political science disciplinary environment. Each is a product of various incentives that individual senior scholars, journals, book editors, professional organizations, grant makers, and the policy world contribute to and reinforce. The essay centers on these three archetypes due to their relationship to knowledge production, the negative effects such research approaches have on research participants and facilitators, and the ripple effects their behavior has on academic communities, including by “ruining the field.”

Why focus on scholar archetypes and how they interact with mentorship practices and disciplinary structures, rather than on individual scholars’ ethics, behavior, methodological choices, and professional responsibilities? Archetypes provide a lens through which to recognize and evaluate individual-level behaviors while acknowledging their relationality and collective disciplinary responsibility for substandard norms and practices. Scholars do not exist in a vacuum; rather, disciplinary structures shape the opportunities they receive and the career pathways they pursue. Mentorship incorporates training, provides ethical guidance, conveys incentives, transmits priorities, and acts as a socializing, gatekeeping, and disciplining mechanism. Mentors, particularly graduate mentors, must actively sign off on research documents such as a prospectus, ethical clearance, grant documents, and, eventually, a dissertation. The choice to validate behavior, or not, to push for justification, or not, and to say “no” when necessary are all part of mentorship, in addition to reviewing, editing, and hiring.

Based on over a decade of experiences as part of the Advancing Research on Conflict Consortium, the SAFEResearch Initiative, the Qualitative Transparency Deliberations, and the Research Ethics in the Middle East and North Africa project, as well as several roundtables and panels at conferences such as the American Political Science Association Annual Meeting and the International Studies Association Annual Meeting, this essay identifies three archetypal scholars and provides defining traits of each: the Indiana Jones model, the legal-securitization model, and the Procrustean design model.10 For each archetype, I also highlight an associated practice—such as dual-hatting, ethical box-checking, and amputating history and context—and note how they further complicate research practice and knowledge production. Finally, I identify key practices and guidelines for academic mentors on committees, in departments, and in broader research communities to informally incentivize safer, more respectful, and more methodologically robust research practice.

The Indiana Jones archetype has become a ubiquitous reference in field-oriented social science conversations.11 In anthropology, references to the fictional Professor Jones are often used pejoratively, which is perhaps unsurprising given anthropology’s ongoing disciplinary introspection following debates surrounding the 1960s Camelot program, the 1980s reflexive turn, and conversations regarding scholars’ participation in the U.S. government’s now-discontinued Human Terrain Systems (HTS) program.12

Anecdotally, political science seems to have a more complicated relationship with Indiana Jones. In nearly twenty years of working in violence-affected and fragile spaces, I have heard uses of the moniker that span from complimentary to critical. In conversations surrounding fieldwork and ethics, it might be deployed to convey admiration for brave or daring work; reserve regarding seemingly rogue or reckless behavior; and critique of dated, dangerous (neo)Orientalist and (neo)colonialist practices. The comparison might be used in all three ways to refer to the same project or scholar. In general, I argue that structures that promote the Indiana Jones model (IJM) convey the following tenets:

1. Dangerous fieldwork is inherently valuable.13

2. Rules are meant to be broken.

3. Data are there for the taking.

The IJM is fundamentally competitive rather than collaborative.

Though the model may be adopted by people of all genders and backgrounds, the IJM is facilitated in its risk-taking, rule-breaking, and extractive practice by intersections of white, Global North, class, and frequently male privilege. In other words, the IJM is based on the advantages provided by the privileges of gender, whiteness, professional affiliation, access to financial resources, and citizenship. These factors, in turn, exacerbate power disparities between researchers, on one hand, and facilitators and participants on the ground, on the other. IJM graduate students often actively resist responsible mentorship. For example, one colleague noted that within their department, students who were told that their work was too dangerous or unethical by specialized conflict scholars often then went to a specific, nonexpert faculty member who would approve it.

The IJM often focuses on style over substance. It is more concerned with what the researcher does and where, rather than the utility of that positionality or the validity of the data it generates. Researchers may talk about military embeds without discussing why said position was necessary to knowledge generation, the ethics of undertaking it, or the risks that position posed to the researcher and those around them, despite the fact that all these dynamics affect the quality of the data gathered.14 Researchers may conceal their funding sources from participants, fundamentally compromising those participants’ ability to consent to involvement in research.15 Researchers may even skip ethics review processes, labeling them as “too cumbersome” or “just there to protect the university.”16 This occurs despite the fact that home university IRB approval or formal subvention to another university’s IRB is a legal requirement for most nonexempt international human participants research conducted by scholars at U.S. institutions. However, IJM researchers may deliberately dodge these requirements, wagering on a lack of oversight from their home institution and/or light accountability demands from journal editors during the publication process.17

The IJM’s inherent valuation of dangerous research for its own sake, and the concurrent labeling of the researcher as “brave” for having pursued it, places research participants, facilitators, and other researchers at risk. Political scientist Milli Lake and I have referred to the resulting competition over which scholar can do the riskiest fieldwork as “outdangering.”18 Disciplinary structures reward such behavior, rather than questioning its intellectual payoff. Among respondents to a survey conducted by Schwartz and Cronin-Furman, 76 percent of faculty “agreed or strongly agreed that fieldwork in ‘dangerous’ contexts earns credibility.”19 This orientation, as well as the broader incentive structure, encourages scholars to put themselves, research brokers such as fixers and enumerators, and research participants in unnecessarily risky situations for the sake of being able to claim “dangerous” research.20 These issues are not hypothetical; during interviews with journalists in Iraq in 2019, one informed me that a researcher well known for risk-taking had recently “nearly got her fixer killed.”21 Such behaviors also devalue the role of developing solid research questions, creative research designs, and robust data. Research projects in this model can often be addressed via different approaches, but are not due to the disciplinary value placed on risk-taking and the replication of sensationalist narratives.

Like the fictional Indiana Jones, researchers in the IJM might be encouraged to work as consultants with governments or think tanks, as full-time employees for international nongovernmental organizations, or as journalists. There may not be anything inherently problematic about these relationships, depending on the project at hand and the transparency with which the researcher presents themselves and their partnerships.22 Indeed, some projects necessarily rely on research collaborations with civil society organizations or humanitarian actors, with mutual benefit and advantages for knowledge production.23 Yet extant disciplinary structures and incentives may encourage graduate students and young scholars to uncritically build these secondary affiliations without considering the potential ethical pitfalls, safety concerns, conflicts of interest, and effects on the robustness of their data.

Dual-hatting, or taking on two or more roles, can thus bring researchers’ motivations, ethical commitments, priorities, and protected status into tension. For example, it may produce what political scientists Michael G. Findley, Faten Ghosn, and Sara Lowe call “nested incompatibilities,” which comprise distinctions “between a researcher’s ethical obligations and that of a partnered research entity, whether that is an NGO, IGO, or government.”24 These commitments can surface due to previous work experience, or become available during the researcher’s graduate career. Political scientists Kristine Eck and Dara Kay Cohen note, for example:

While some students may have relevant work experience that enables them to be sensitive, skilled researchers, students with professional or activist backgrounds may struggle to distinguish between their past work values and their current role as student researchers. Having worked for institutions like governments, non-governmental organisations (NGOs), advocacy groups, or news media, students may still subscribe to the professional imperatives of those institutions. This can bias their research and undermine the scientific process, and unwittingly makes student research less ethical.25

Here, Eck and Cohen are most productively understood as emphasizing the inevitable and necessary distinctions between the goals, processes, and ethical commitments involved with robust academic knowledge production and those involved with, say, journalistic reporting, rapid provision of life-saving aid, community advocacy, or the pursuit of a political platform (even if some methodologies might partially combine them, like participant action research). Additionally, dual-hatting surfaces questions about when the scholar and any assistants, translators, or fixers are bound by academic ethics, such as those surrounding consent and confidentiality, and when they are subject to rules governing, for example, U.S. government employees’ conduct (like in the case of active duty military) or journalistic norms. Moreover, dual-hatting can place scholars and those assisting them at increased suspicion of working for intelligence agencies.26

While plenty of scholars have dual-hatted in transparent, ethical, and intellectually productive ways, dual-hatting has had tragic consequences. For example, Johns Hopkins University political science student Nicole Suveges was killed in an insurgent attack while working as a civilian U.S. Army contractor for BAE (British Aerospace) systems under the HTS program, a U.S. government program that hired scholars with regional and cultural expertise to work on the ground alongside U.S. troops. She was simultaneously conducting dissertation research in Iraq, including by collecting public opinion data.27

It is worth asking whether Suveges would have been in the setting where she was killed without the access that dual-hatting provided and the professional incentives for “policy-relevant” work on U.S.-occupied Iraq at the time. A second, extremely relevant question is how her job with BAE systems and the U.S. military affected her positionality with relation to the field and the Iraqi population, as well as her ethical commitments. Third, it is worth contemplating how Suveges’s approach, should she have lived, might have incentivized other scholars to undertake similar behaviors.

This example does not mean that dual-hatting should be universally prohibited or disincentivized. Rather, it encourages faculty mentors, departments, and others in the discipline as well as the policy world to pose critical questions about these relationships and their products. How does a dual-hat role shift the researcher’s positionality and vulnerability, as well as those of their participants, facilitators, and others around them (for instance, if a government targets a researcher for applied human rights work or critical journalism)? How might research participants react to and modulate their interactions with a scholar who is simultaneously consulting with a European or North American government agency, the United Nations, or an aid-providing NGO? Which institutional entity is responsible for a researcher’s safety if they are working as both a graduate student and a think tank consultant? What does it mean for nonuniversity entities to piggyback scholars’ access on research visas and support from grant funding or, conversely, for graduate programs to rely on students supporting themselves financially via dual-hatting due to a lack of sufficient resources?

The legal-securitization model (LSM) centers on research bureaucracy: risk assessments, university permissions, research visas, and formal ethics approvals. Some of this bureaucracy (such as IRB approval in the United States) is legally mandated, though its permissiveness and enforcement can vary from university to university. Therefore, some researchers may gain permission to engage in data generation activities such as interviewing Rwandan prisoners while others are denied.28 One of the key tenets of the LSM is that risks and ethical considerations can be known and mitigated in advance. A second is that as long as scholars check the appropriate boxes, the research is deemed appropriate and ethical. A third is that scholars are equal and interchangeable, rather than embodying distinct positionalities and possessing individual skill sets. The rules and risks apply to all researchers and all projects, regardless of, for example, gender or citizenship. I argue that the LSM is problematic because it encourages box-checking, the exotification and securitization of “the field,” and a view of “the field” as a space that can be contained and managed. Mentors who operate through this model teach mentees that as long as they can gain ethical approval and demonstrate to a university risk assessment committee (and, sometimes, a contracted insurance provider) that a location is sufficiently “safe,” it is ripe for research. At the level of university administration, the LSM focuses on both intensive preplanning and devolving accountability solely to the researcher or mentor when things go wrong.

This is not to say that safety is not a serious issue for researchers. Among graduate students who conducted international fieldwork, 58 percent of women and 47 percent of men reported experiencing safety issues while conducting their research.29 Yet approaches to negotiating researcher safety tend to be procedural rather than holistic. At the university level, there is at least some recognition of potential risks to scholars, though this is often liability-oriented rather than cognizant of the broader politics of academic freedom and repression.30 However, Schwartz and Cronin-Furman underscore that, at the department level, “nearly three-quarters (72.6%) of our graduate student respondents who had conducted fieldwork reported that their advisor had not asked them what kind of health and safety measures they had put in place.”31

Of top-ten programs in political science, anthropology, and public health—disciplines that are among those most likely to have international fieldwork embedded in doctoral tracks—most have at least some restrictions on fieldwork travel.32 These are usually institutional approvals for locations deemed “risky” or the sign-off of a regional expert. Princeton University, the Massachusetts Institute of Technology, Vanderbilt University, Indiana University, and the University of Notre Dame ban student research in specific countries. By contrast, Stanford University has almost no restrictions on graduate student or faculty travel, though the university, in its words,

strongly recommends against, but does not prohibit, travel to countries where the International SOS risk rating (either the medical risk or travel risk) is “High” or “Extreme,” or where there is other reliable information of significant health or safety risks for either the country destination, or for the particular region or location of intended travel within the country destination, or where there is other reliable information of significant health or safety risks. . . . The University reserves the right to require the execution of an appropriate release or waiver before permitting such travel.33

This recommendation begs the question: What is the source of such risk ratings? Universities generally obtain them from government agencies or private companies (such as International SOS). Harvard University, Princeton University, the University of California, Berkeley, and the University of Chicago, for example, all cite the U.S. Department of State as a source for risk assessments. Several questions arise: For whom are these risk ratings generated? What is their goal?

Safety for both researchers and participants is contingent, relational, and intersectional, despite institutions’ attempts to quantify and rank risk.34 A country might be deemed “safe” for research, with no attention to what research is planned and who is conducting it. Because some institutions rely upon U.S. Department of State or UK Foreign Office country risk assessments, situations have consequently arisen in which a student of Middle Eastern heritage who normally resides in a Middle Eastern or North African country is informed that said country is “too dangerous” to visit based on Department of State threat coding, rather than an evaluation of relevant on-the-ground dynamics and the proposed research.35 It bears saying that there is likely a substantial difference in risk between a project that examines youth participation in recreational activities versus one focused on the clandestine dynamics of illicit markets. There is also a difference between a scholar who speaks a local language, is familiar with the spaces in which they will be working, and has considerable social resources on the ground, versus one who has never lived in that country. And as political scientist Jillian Schwedler’s work argues, assumptions about gender-based vulnerabilities often do not hold on the ground, where researchers are interpolated through local understandings of their presentation and role.36

The LSM privileges rule-following and adherence to the letter of the law regardless of context specificity. Perversely, this commitment can bring it into tension with ethical considerations. For example, when conducting team research in the Great Lakes region of Africa, my home ethics board required our team to obtain local ethical approval for interviews with journalists. The country in question is known for persecuting journalists. However, the flagship public university would not permit the research team to use oral consent (despite the home university approving it), instead requiring signed, identifiable consent forms from all journalists who were to be interviewed. The research thus could not ethically be conducted in country, given the small interview sample, the university’s ability to review the documentation, and political interest in the research population. For other studies, however, signed consent forms might have been entirely appropriate given, for example, histories of exploitative medical research in the region.37

In contrast, in a particularly poignant set of examples of the pitfalls of the LSM, colleagues, research assistants, and translators have written any number of pieces regarding the ethical violations their colleagues or employers from the Global North have committed, often while inhabiting a direct mentorship role.38 For example, Moe Ali Nayel, a Lebanese researcher and journalist, published an account of a Harvard researcher who was conducting research on Syrian refugees in the Shatila refugee camp in Lebanon with two students in tow. Nayel writes about the research team arriving to interview a woman who had recently fled Syria, relating how the local research assistant emphasized to the team that the woman’s son had been wounded and added “I hope you’ll be able to help the poor woman.”39 Nayel relays how the faculty member, upon reaching the refugee’s house, emphasized to her students: “We are not here to talk about her son.” When people in the house lit cigarettes, “the Harvard team coughed and complained [and] the cigarettes were politely put out.”

Nayel presents the continuing set of interactions as a case study of researcher callousness and lack of contextual awareness. He writes of one of the students:

Her human rights kit was out: a long list of questions laid out, voice recorder turned on and set on the coffee table, different color markers deployed, a bundle of papers next to us on the couch. The student organized her tools, gave a nod to the professor and the round of human rights questioning started. Her quick-fire questions started with the basics: name, age, marital status, number of children and place of residency in Syria. Human rights documentation training was now in action.

Nayel continues, emphasizing first how the student demanded the interviewee put out her cigarette:

The persistent human rights student, here only to conduct her by-the-book interview in the presence of her evaluating professor, continued with her tiring and condescending questioning.

Despite the dehumanizing and insensitive way in which this interview was conducted, it may not have “technically” violated ethical protocols. There is evidence of coercive dynamics, as several people involved in the interaction seem to expect that the research team would be able to help the woman and her son, a  long-recognized challenge for researchers operating in crisis-affected spaces.40 But despite the lack of respect for persons embodied in the interaction, it is entirely possible that the professor and her students did nothing that an IRB would consider to be grave violations of the tenets of their ethical review unless the interviewees reported it as such. Indeed, the direct, no-tangents-allowed questioning style Nayel describes closely hews toward common approaches to interviewing/deposing in the legal world.

IRBs and other ethics approval processes often do not address the framing of the study, whether the proposed site or population experiences overresearch, or the ethical nature of the project’s questions; they are predominantly concerned with whether the research team itself might do harm via the application of the research protocol. As Findley, Ghosn, and Lowe note, IRBs regularly approve research on questions regarding the efficacy of regimes deploying fear or lethal violence, without passing judgment on whether those are questions that ought to be asked in the first place.41 In a different vein, political scientists Daniel Masterson and Lama Mourad note that some populations’ vulnerability make them more easily accessible than others, meaning that their very “research-ability” is derived from characteristics that should engender particular care as well as commitments to beneficence that go beyond “do no harm.”42 However, the LSM and scholars in this practice may present this extra care or attention to context as optional (or “too political”) rather than as an inherent responsibility.

The Procrustean design model (PDM) deliberately excises history and context as an expedient way to make populations, groups, or cases “fit” a specific research design.43 PDM projects do the following:

1. Ignore, omit, exclude, or gloss basic aspects of history and context in design, coding, and/or analysis;

2. Treat history and context as inconveniences to be neutralized by research design or written off as random error, rather than as opportunities to be negotiated with innovative research design and acknowledged in the write-up;

3. Approach research sites as sterile, enclosed laboratories where each research intervention is discrete, independent, and unaffected by others and isolated from ongoing politics, rather than as complex, living communities and political spaces.

To be clear, there are vibrant and necessary disputes about infinite aspects of historical record and differences in how people experience and understand on-the-ground context. There are also different ways to incorporate the historical record, context, and researcher positionality into diverse epistemological and methodological traditions.44 PDM scholarship is not, however, engaged with or in them. By contrast, the PDM is detached and even dismissive of history and context. At best, it suffers from myriad issues including bias, data truncation, misinformation, and invalid findings. For example, PDM scholarship might look like a field experiment that researchers claim to have carried out with an incentive that the recipient population would not want or that would obviously endanger said population (for example, via easy de-anonymization or overwhelming risk). Or it might take the form of a survey that uses languages, categories, or questions that do not translate smoothly or accurately into local context or language, inducing nonresponses or preference falsification.45

PDM scholarship in the form of historical natural experiments as well as other forms of design-based inference has generated increasing critique, especially because it often relies on “thin or sometimes inaccurate historical evidence.”46 Following a proliferation of articles claiming that social/historical processes generated an “as-if-random” treatment appropriate for estimation, responses emphasized that said processes were demonstrably not random.47 Indeed, some work in this vein is ahistorical to the point of absurdity. Political scientist Ajay Verghese argues that the results of such studies often produced inaccurate history, violated design assumptions, omitted variable bias, and demonstrated other measurement issues.48 These works might generate productive discussion regarding the nature of causal inference.49 However, conclusions based on ahistorical and biased data contribute to low-quality future hypotheses, theorization, and conceptualization, which ripple through future work, lowering its overall quality and producing flawed conclusions as well as reflecting poor historiography.50

PDM projects’ problematic nature has particularly acute effects on the validity of research on identity and ideology. A PDM project might involve grouping populations by ethnic, racial, religious, or sectarian labels that they would not themselves choose (such as “Christian” and “Muslim”) in ways that paper over historically relevant and analytically salient identities (like “Assyrian” or “Shabak”), that misrepresent doctrinal traditions (for instance, incorrectly labeling Yazidis as “Muslim” or confusing Zaydis with Twelver Shiʿas), or that conflate members’ individual religious identities with an organization’s stated ideology. Scholarship on topics such as religion and sectarianism in the Middle East and North Africa routinely ignore longstanding political traditions of leftist, secular, and antisectarian identification and organizing (for example, by not incorporating a nonreligious ideological treatment or survey answer alongside doctrinally inflected prompts or identity categories).51

These practices can result in, for example, questionable sampling protocols, omitted variable bias, conceptual stretching, and misaggregation of data, all of which affect data quality and analysis while potentially perpetrating epistemic violence. The aggregate effect in some research is an ecological inference problem, where a glossing or omission of the historical record brings the validity of individual observations and aggregate trend analysis into question. The Minorities at Risk Organizational Behavior (MAROB) Middle East 1980–2004 dataset, for instance, codes the Iraqi Communist Party (ICP) as representing “Shi’is” since its founding in 1934.52 However, historian Hanna Batatu, a foundational source on Iraqi politics and society, extensively documents the party’s cross-sectarian and multiethnic leadership, which included noteworthy percentages of Christians, Jews, and Kurds in addition to Sunni and Shiʿa Arabs.53 Political scientist Shamiran Mako describes the ICP as “the most ethnically cross-cutting oppositional coalition force with the highest mobilization capacity among Shiʿa-Arabs, Assyrians, Kurds, and Jews,” and emphasizes that “the egalitarian principles of the ICP enabled working class Iraqis to join anti-government opposition and uprisings in response to increasing cost of living in the 1940s and 1950s.”54 Ahistorical coding choices that erase the diversity of membership and ideological foundations of parties like the ICP promulgate the false notion that Iraqi political organizations are inherently and exclusively sectarian, consequently masking key empirical questions about, for example, the sectarianization of Iraqi politics.

Elsewhere, assumptions of the primacy of ethnicity, religiosity, and sectarianism in group behavior, as well as a practice of confusing organizational forms in counting armed groups, are problematic for empirical analyses of armed group behavior. Two immediate issues are manifest: 1) ontological confusion is cemented in coding strategies and practices (and even further through publication and dissemination) and 2) a disproportionate number of leftist organizations, even with comparable attributes, are omitted from analysis (a missing data issue). Though some work on militant group alliances has made clear the need to distinguish between single organizations and the fruits of intergroup cooperation, other work continues to conflate the two.55 The Foundations of Rebel Group Emergence (FORGE) dataset provides one example: the authors describe the project as coding attributes of “rebel group organizations,” yet include in their final accounting highly institutionalized proto—state apparatuses such as the Palestine Liberation Organization (PLO) along with its constituent member organizations (like Fatah or the Popular Front for the Liberation of Palestine, which are nonstate armed groups), intergroup alliances such as the Palestinian Rejectionist Front, as well as the Palestinian National Authority (PNA) in the same dataset.56 This “counting” issue is compounded by the well-documented problem of systematically missing data on smaller organizations. Excluding smaller leftist and Ba`thist parties from datasets, especially when Islamist organizations such as Hamas and Palestinian Islamic Jihad are included, forwards (neo)Orientalist assumptions regarding the salience of religious identities and ideologies in the Middle East, erases the historical role of leftist and nonsectarian organizations in regional politics, and contributes to myriad inference issues.57

The existence of these three archetypes and the recognizability of their constituent practices imply both immediate and long-term actions that scholars can take in response. In line with Rodehau-Noack, Schwartz, and Broache’s findings, reviewers should regularly push for more detailed, holistic descriptions of projects’ ethical orientation and practice rather than fly-by mentions of an ethical certification.58 This approach would also provide space for scholars working in university systems without such reviews to demonstrate their own commitment to ethical procedures. Scholars’ treatment of history and context should be evaluated as a culmination of expressly methodological choices that affect the validity of data and analysis. For any historically oriented manuscript or dataset, reviewers, editorial assistants, and editors should check for relevant historical citations and closely evaluate coding decisions, with attention to whether relevant non-English languages and sources have been employed. For such projects, editors should assign at least one reviewer with substantial background in relevant regional or national history. Journals’ editorial boards should directly ask submitting authors about their employment of nonauthor research assistants, translators, and fixers, and require disclosure of their roles. Reviewers should also critically note the existence of sensationalist and other problematic language in articles.

There are also more substantial changes that scholars can make to their own practice and to encourage positive evolution of others’. Scholars, particularly those in mentorship or hiring positions, should actively seek to disincentivize dangerous fieldwork for its own sake. When presented with scholarship (at any stage) that incorporates fieldwork in fragile and violence-affected contexts, scholars should pose constructive questions about its motivation, ethics, and expected payoff (especially to affected populations). Doctoral programs should invest in more comprehensive fieldwork training.59 Mentors should deliberately talk through early-career researchers’ positionality, safety and health plans, and ethical commitments; they can benefit from extant resources such as the mentoring checklist produced by the Research Ethics in the Middle East and North Africa Project, as well as collections such as the Advancing Research on Conflict bibliography. Scholars who plan to engage in dual-hatting should complete comprehensive ethics evaluations that account for their dual-hat positionality and how their nonacademic roles might interact with their academic ones, especially when it comes to effects on research participants and facilitators. When dual-hatting could affect participant consent or well-being, researchers should either disclose or rethink their approach. This is an extension of practices already encouraged by programs such as Bridging the Gap, which include scholars developing peer-consulting networks for ethical practice.

These recommendations could conceivably slow research processes and publication timelines, which is one of the primary critiques their advocates have previously faced. Yet much of the research discussed here would not have been published if the expectations and support for researchers, editors, and reviewers were of an appropriate caliber. If the options are high-quality, valid, ethical, safe research versus fast, invalid, and unethical research, the choice should be obvious: incentives must change. Pursuing such a tack would require an express embrace of shifting priorities and an active reorientation of standards within departments, within journals, and across disciplinary institutions. This challenge presents a clear collective action problem within our professional community, but one that needs to be confronted if social science research is to be taken seriously both within academia and beyond it.