On Power

Researchers have been taught that they hold power over their ‘subjects’ simply by introducing them into conversation and, therefore, must carefully wield those responsibilities. One basis for this belief is the legacy of medical and psychological experimental research and its concern with human subjects protection, whereby the ‘researcher-researched’ relationship could be reduced to “the knowledgeable and powerful experimenter” and “the ignorant and powerless subject.”^{Footnote 21} Another basis for this paradigm is the historic scaffolding upon which relations between the Occident and its ‘exotic’ objects of study were built. This historical dynamic has translated into assumptions about the investigator’s responsibility for and power over her subject.^{Footnote 22}

Those who collect the opinions and narratives of others do have certain powers that bear recognition. Intuitively, this is especially true for researchers working in countries where their home governments have ‘boots on the ground.’ In Afghanistan, for example, we had the capacity to offer tangible support to those we met.^{Footnote 23} In addition to our financial assets, we had cultivated forms of social capital in Kabul that afforded us knowledge and connections otherwise unavailable to our contacts in the countryside. Our networks extended back to our home countries, where we were in a position—real or perceived—to support Afghans seeking access to university admission, scholarships, asylum applications, employment, and the policymaking establishment. Arguably, this imbalance of power might be further exacerbated by researchers who “alter the setting, scenery, and sometimes life chances of individuals”^{Footnote 24} through more intrusive methods.

There is also an “imperialism of knowledge production” in which all academics take part.^{Footnote 25} Researchers have the power to transmit (and distort) the thoughts of others.^{Footnote 26} There is a risk implicit in an informant’s sharing of his story with a researcher since he has no control over her subsequent interpretation.^{Footnote 27} There are also risks that arise once a researcher publishes her work, whereby those who contributed to the research might suffer consequences.^{Footnote 28} The work might also be utilized by governments toward ends that may not have been intended by its author.^{Footnote 29}

A researcher can try to mitigate these risks; Jesse Driscoll, for example, chose not to include a list in his book of the Tajik commanders he studied to avoid offering up a “targeting list.”^{Footnote 30} The formal omission of this list could not, of course, prevent officials from extracting information from the publication or those still in Tajikistan who contributed to the project; in many ways, these risks reflect the powerlessness of a researcher once findings are released. Séverine Autesserre explained, for example, how her work had been used by politicians on both sides of the Rwanda-Congolese border to opposite ends: “I have no idea of what I could have done (or could do) to prevent this kind of manipulation.”^{Footnote 31}

As such, a number of scholars have called into question the ‘researcher as powerful, researched as powerless’ paradigm and underscored the reality that their kith are often the less powerful parties in this dyad. Fieldwork that involves participant observation and interviews can only be accurately characterized as a symbiotic relationship between conversants, each of whom has the capacity to shape the conversation.^{Footnote 32} This kind of research can be conceived as a collaborative “process of sense-making” in which all those involved are part of a larger “information economy” that none of them controls.^{Footnote 33}

To understand field research as anything other than a symbiotic back-and-forth is to misrepresent it and undermine our informants’ autonomy. The power dynamics between researchers and researched are especially complicated in conflict zones, where violence is a key currency in all relations. There is a persistent “potential for danger” that comes with research in the midst of violence.^{Footnote 34} Situations arise in which a fieldworker is unable to control her interactions with her informants. Even research with youth, when they are armed and streetwise, can lead to “a dramatic shift in power.” The notion that the researcher has control over this context is simply out of alignment with facts on the ground. A researcher can actually be much more dependent on her informant than the informant is on her.^{Footnote 35}

Power asymmetries can be particularly pronounced in research with elites, especially those with outsized access to coercion and capital. As young graduate students, we had little reason to believe that we would command power over well-established, political and military elites. Field research can sometimes mandate a ‘guest’ and ‘host’ relationship that upends the presumed ‘investigator-subject’ paradigm entirely. Laura Adams, writing about her work in Uzbekistan, explained, “the guest dynamic plays a trick in which the apparent power relations are inverted … When a field-worker accepts the guest contract, she trades a measure of freedom for a measure of protection.”^{Footnote 36} During a trip to Afghanistan, one of us found herself invited to dinner at the home of a former militia commander. She had not expected an interview to parlay into a gathering of former mujahideen. Although the conversation proved fruitful for her research, one could hardly characterize her as the more powerful party vis-à-vis her informants. On the contrary, she was a guest whose research aims, reputation, even safety were in the hands of her hosts.

Researchers are not exempt, in these contexts, from becoming political resources to their research subjects, especially when political elites are the subjects of concern and elite interviews serve as a central research instrument. A handshake one of us shared with an Afghan governor he had met with for less than five minutes was broadcast on local television. This appeared to be a concerted attempt to counteract the governor’s reputation as an uneducated countryman by exhibiting his connection to a foreign university professor. In another instance, the same author was mentioned in a public speech by a warlord’s lieutenant as evidence of the international community’s interest in his liege. A researcher can become, wittingly or unwittingly, “an important element in [a research subject’s] strategies to attain certain objectives.”^{Footnote 37} Eva Gerharz, for example, described how researchers operating in Sri Lanka were used as “potential promoters and sometimes even advocates by the [Liberation Tigers of Tamil Eelam].”^{Footnote 38} A power dynamic of this kind can, in its extreme form, result in the researcher becoming a “mascot,” utilized by locals to advance their interests.^{Footnote 39}

The reversal of the conventional power paradigm occurs in a variety of contexts. Experimentalists can, for example, be instrumentalized by individuals who leverage their knowledge of the local political landscape to influence selection processes for the benefit of their communities. Perhaps more problematic is the manipulating power that fixers, interlocutors, and informants can have over knowledge production by vengefully pointing fingers at people who have wronged them (and who are not covered by the human subjects protection guidelines).^{Footnote 40} This is particularly concerning because, as Driscoll put it, “set[ting] in motion those downstream consequences is often the reason that people are talking to us in the first place.”^{Footnote 41}

These ambiguities have striking resonance with other forms of intervention. Peace-builders, state-builders, and counterinsurgents often believe that they have the authority, responsibility, and capability to transform societies plagued by violent conflict, authoritarian politics, and socioeconomic underdevelopment.^{Footnote 42} The perceived transformative authority resides within a system of “organized hypocrisy” in which the most powerful states respect the sovereignty of weaker states only so long as it remains in their interest to do so;^{Footnote 43} the “responsibility to protect” suggests that the sovereignty of these states is contingent on good behavior.^{Footnote 44} The international community’s perceived transformative capability lies within a “conceptualisation of modernisation … as a project of social engineering.”^{Footnote 45} These beliefs about the power of the outsider are grounded in some undeniable realities, but also in perceptions that are grossly exaggerated.^{Footnote 46}

In fact, local actors have long instrumentalized the presence of foreigners in conflict zones.^{Footnote 47} Since 2001, Afghan notables have leveraged the presence of Coalition forces to their own ends. They have, for example, exploited the Western eagerness to hunt the Taliban to advance their own interests and settle personal feuds.^{Footnote 48} In war zones, more generally, local elites bargain with one another over power and loyalty, often through the use of force, while foreign players remain largely disadvantaged outsiders, “readily manipulated or bypassed by the better-endowed and cannier national players.”^{Footnote 49}

In sum, interveners, scholarly or otherwise, are engaged in power relationships that are far from unidirectional. Westerners may represent, even wield, great power within the international system and still be unable to define events at the micro-level. It is possible, in other words, to intervene and still be vulnerable to forces beyond one’s control.^{Footnote 50} Like any other outsider, a researcher can become an object of manipulation by parties to the conflict she studies. Foreign researchers have access to specific resources that sometimes place them in a position of power. Yet, informants also have access to resources of their own, including their superior understanding and knowledge of all that is ‘local.’ They are in a position to leverage the presence of foreigners to ends often beyond their knowledge, let alone their control.

On (Im)Partiality

The fact that the researcher’s dominance over her subject cannot be presumed has implications for how researchers comport themselves. There are steps they might take because of the limited power they have that do not align with another related and long-held fieldwork precept—the aspiration to be “the fair ethnographer.”^{Footnote 51} Impartiality, sometimes framed as objectivity, has been historically understood as a prerequisite for rigorous fieldwork.^{Footnote 52} As Robert Keohane urged, “in our particular investigations we need to seek objectivity because otherwise people with other preferences will have no reason to take our findings seriously.”^{Footnote 53} Relatedly, it was almost axiomatic that a scholar should remain neutral and independent.^{Footnote 54}

This commitment to a position of impartiality, neutrality, and independence is captured by the “fly-on-the-wall model,” namely “the idea of a value-neutral, unbiased, and minimally disruptive social science.” In this ideal type, the researcher “takes pains not to express opinions, take sides, or draw attention.”^{Footnote 55} We add our voices to the chorus of scholars who question the implications of this commitment by drawing attention to the unviability of such a posture in the midst of violent conflict. As Jonathan Goodhand warned, “research, like any other form of intervention, occurs within an intensely political environment and is unlikely to be viewed by local actors as neutral or altruistic.”^{Footnote 56} In highly charged environments, partisanship of some sort may become unavoidable;^{Footnote 57} it might even be a security imperative.^{Footnote 58}

We start by considering neutrality, because it is the one dimension of this posture in which the scholar has the greatest volition. Today, some in the academy “feel a special obligation to … speak up for those trapped…in repressive political environments.”^{Footnote 59} Others seek to leverage their knowledge to influence Western policymaking. Activist research involves an agenda to improve the conditions being studied.^{Footnote 60} Michael Mosser encouraged social scientists to foster the “‘revolving door’ between the academy and the policy world,” in particular the military and defense establishment.^{Footnote 61} Charli Carpenter and Michael McCann acknowledged their (sometimes unintended) roles as changemakers in the arenas they studied, international human rights and American labor rights respectively.^{Footnote 62} And Lara Deeb and Jessica Winegar pondered why anthropologists, as a community, did not strongly protest “the horrors of escalating Islamophobia … and the unleashing of the world’s most powerful military on Afghanistan and Iraq.”^{Footnote 63}

Scholars of an activist persuasion might advance an obligation to study charged subjects like Afghan warlords in a way that acknowledges, even works against, the harm inflicted on those living under their reign. In one awkward conversation, one of us was asked by an international organization official to share any information on militias that could help “give justice to the people.” This was a difficult position: either betray the trust of future informants or go back on a tacit understanding that the official had imposed. Our decisions to abstain from moralizing about our research subjects have earned us critique from onlookers on multiple occasions.

But, in addition to the intellectual reasons for avoiding normative claim-making, trying to ‘do the right thing’ can have far-reaching consequences. For example, William Reno was menaced by former warlord and President of Liberia Charles Taylor himself after testifying before the U.S. Congress. “It was the first wake-up call,” Reno remarked. “Things that you did in the past have downstream applications … [they] can come back and they can haunt you.”^{Footnote 64} In a case like this, a researcher’s efforts at advocacy can jeopardize professional and physical integrity. The emergence of international criminal law and the existence of a criminal court that claims universal jurisdiction have further consequences. Fieldworkers who gather rich empirical data on armed actors now run the risk of being called to testify as expert witnesses, or even of being subpoenaed to release their data.^{Footnote 65} This legal and normative shift may take away the researcher’s choice to remain neutral altogether.

Beyond the ethics (and risks) of actively taking sides, many have come to wonder whether a truly impartial—objective and unbiased—approach has ever been genuinely tenable. After all, undertaking research often has social, political, and economic implications; moreover, the researcher can never control how her acts are perceived.^{Footnote 66} A rich body of literature has arisen around the fallacy of impartiality as a cornerstone of social science.^{Footnote 67} Every researcher has a host of different identities that shape the dynamics of interaction with research subjects.^{Footnote 68} As Charli Carpenter explained: “My various identities—American, political scientist, professor, woman, mother—were interpreted differently in different research contexts and always affected the data I was able to gather.”^{Footnote 69}

With some methods, like participant observation, the process of immersion into a community may make working at any remove impossible.^{Footnote 70} Similarly, in some contexts, an immutable feature of a given researcher—a Jewish Israeli studying the Arab-Israeli conflict, for example—may make it impossible for a scholar to be perceived as impartial.^{Footnote 71} Some scholars may wish to cling to their identities as genuinely disinterested observers. They do not wish for their curiosity to be confused with admiration.^{Footnote 72} And, yet, one’s aspiration to acquire knowledge may be read as support by others. In Sri Lanka, Gerharz found herself perceived as “being more sympathetic to the ‘terrorists,’” because she suggested federalism as a sustainable solution to the conflict.^{Footnote 73} Similarly, onlookers might have assumed an affinity between each of us and a given Afghan warlord we had chosen to investigate. This was a challenging affiliation to manage and one that became hard to shake after years of research. Throwaway remarks like, “your friend, the governor” indicated that the study of a controversial figure could quickly be conflated with sympathy for, even complicity in, that individual’s political project.

Researchers can also be identified as confederates to the policies of their governments. During field trips to the Turkish-Syrian border, one of us was confronted by anger about the U.S. government’s response to the Syrian civil war. On occasion, an interview devolved into an interrogation by the researched about American foreign policy, President Assad, and Syria’s fate. Assumptions about the opinions of a given researcher may be coupled with a more general belief that Western academics have the power to influence ‘the establishment’ back home.

These assumptions can be harmful, as being associated with a government’s domestic or foreign policies a fortiori, can put a researcher at risk. Scholars deploying various methods have been forced to contend with this reality. The act of, for example, undertaking a survey can be perceived as a partisan one. Arab-American scholar Amaney Jamal, conducting survey research in the Palestinian territories, found herself suspected of affiliation with the Palestinian National Authority.^{Footnote 74} In a field experiment in Lebanon, Daniel Corstange found that respondents were far less likely to take a survey they were told was sponsored by a Western government than one in which no mention of sponsorship was made. Although sponsorship by an American university did not dampen participation, for qualitative researchers who engage communities in less formal terms, distinctions between ‘governmental’ and ‘academic’ are not always easily broadcast.^{Footnote 75}

In Afghanistan (and Iraq), the risk of association was aggravated by the now-defunct “Human Terrain System,” a program aimed at embedding social scientists within the U.S. military to provide “socio-cultural knowledge of the indigenous civilian population.”^{Footnote 76} This program further blurred the line between soldier, spy, and researcher and may have exacerbated perceived associations between them. The line between research motivated by policy versus scholarly interest may be difficult for observers to decipher. This is particularly true if, as Dawn Langan Teele suggests, researchers “are increasingly engaged in bringing policy ideas to governments and think tanks to be tried out in real-world settings.”^{Footnote 77} We, ourselves, have received government funding, published articles with think tanks, and worked as consultants for both. These engagements do not connote our endorsements of the policies of these patrons, but they preclude us from dissociating ourselves entirely from them. To be fair, objective, balanced, and perceived as such in this context is nearly impossible.^{Footnote 78}

Moreover, today’s Western scholar, like the journalist or aid worker, can be an “object of interest” in her own right.^{Footnote 79} In the latest generation of “new wars,” marked by acute Western involvement (particularly those in the “greater” Middle East), militant organizations seem to make little distinction between different Westerners. This development might be dated back to the mid-1990s when, as Fawaz Gerges explained it, “The jihadi caravan took a new sharp and dangerous turn that [brought] it into a total confrontation with the world community.”^{Footnote 80} The growing conflation of a larger struggle—the “global jihad against the West”^{Footnote 81}—and a number of smaller-scale conflicts made Westerners working in their midst vulnerable in unprecedented ways.^{Footnote 82}

The researcher’s inability to be (and be recognized as) impartial raises concerns about a third aspect of her position—her independence. It is impossible, in these circumstances, to be truly independent absent a far greater degree of power than most social scientists have. Westerners have become lucrative prey for kidnapping as well as attractive targets of lethal violence in unconventional military campaigns. As such, work in hot conflict zones might demand protection from powerful local or foreign actors with whom one would prefer not to affiliate. While working in Mogadishu, for example, one of us benefited from the informal protection of an African state’s diplomatic office as well as from a warlord once backed by the CIA. The other took advantage of U.S. government-sponsored flights in Afghanistan to avoid the dangers of travel by road, even spending a few days on two Western military bases. To need help is to acknowledge the limits of one’s power.

Here again, an analog exists: the desire of humanitarian aid agencies to maintain the sacrosanct values of impartiality, neutrality, and independence.^{Footnote 83} In the last two decades, organizations that took pride and solace in their ability to operate above the fray perceived a shift whereby that ability could no longer be assumed, partly as a result of the militarization of humanitarian assistance.^{Footnote 84} Since 2001, many terrifying episodes have revealed the degree to which Western organizations no longer feel protected by the cloak of impartiality, neutrality, and independence.^{Footnote 85}

While some have argued on behalf of “returning to traditional principles,”^{Footnote 86} others posit that the presumption that “humanitarian action is not a political project” is a dangerous one. It “rings,” wrote Paul O’Brien, from CARE Afghanistan, in 2004, “like an honorable lament for the past, but no longer provides adequate guidance to humanitarian actors in highly politicized settings.”^{Footnote 87} O’Brien’s warning is one, we would argue, to which field researchers should also pay heed. Scholars have indeed been forced to ask many of the same questions our humanitarian counterparts have been asking themselves:

Honest answers would reflect the reality that 21^st-century interveners, including scholars, are part of the social, political, and economic equations that comprise life in these warzones; and they do not have the power to float above the fray with the aloofness they may have heretofore.

A field researcher routinely engages in a delicate dance with her subjects, striving to signal her role as listener, not partisan or judge. Still, there is little reason to believe that scholars are immune from the taint of unwanted affiliation or that they can operate independently in today’s conflict zones. We would assert that those working in particularly challenging environments have little choice but to confront their limited power and vanishing impartiality, and, therefore, independence. Those who pursue qualitative fieldwork with elites in today’s warzones are in fact engaged in a kind of tribal politics of their own.

The Method of Building a Tribe

The optimism of those who believe fieldwork in war zones is possible hinges, in part, on the notion that “investing in a local network” is the antidote to many of the challenges at hand.^{Footnote 89} As Reno said: “You have to become part of the system to study the system.”^{Footnote 90} In the places we study, this means belonging to one or more tribes, micro-systems of networks and individuals who recognize themselves as linked to one another and can offer a researcher access, support, and protection.^{Footnote 91}

Innate or imagined,^{Footnote 92} tribal affiliations remain a defining feature of politics in what Jean-Francois Bayart called “rhizome state[s],” with their “infinitely variable multiplicity of networks whose underground branches join together the scattered points of society.”^{Footnote 93} The focus of our own scholarship involves an examination of tribal politics in the broadest sense of the term—we study state-society relations marked by informal, patrimonial, and patronage-based institutions. These variants of political life are highly diverse in form, but there are commonalities: Personalized relationships are ubiquitous, rules are often informal, and information is of very high value, often hidden or hard to find. Outsiders are challenged navigating these systems, elevating the importance of brokers and go-betweens.^{Footnote 94}

These features affect not only the politics they produce, but also the ways in which students of those politics pursue knowledge and construct meanings about them. The features that distinguish tribal politics from other brands are determinative of the kind of research that can be done in their midst. We see snowball sampling as the methodological correlate to engaging in tribal politics, a way of grafting oneself on to one or more of Bayart’s networks of “underground branches.” N. Patrick Peritore described this approach as “respondent pyramiding,” whereby one builds a network of informants, one on top of the other.^{Footnote 95} In this sense, the method we use mirrors the politics itself.

Snowball sampling, a method very popular amongst scholars of the Middle East,^{Footnote 96} offers opportunities to address the trust deficit inherent in war zones,^{Footnote 97} where informants tend to be suspicious of newcomers.^{Footnote 98} The building of embryonic “trust networks” is in fact an absolute necessity,^{Footnote 99} as informants are more likely to share information with those they have come to know and trust.^{Footnote 100} Snowball sampling can therefore be understood as a form of social infiltration into an impervious setting in which the proper introduction can get a researcher a long way.^{Footnote 101} In these environments, “the goodwill of…gatekeepers among power elites is often essential to be able to begin, pursue, and complete research.”^{Footnote 102}

There is also a more human reason for cultivating trust with one’s informants. Students of war find themselves in foreign, frightening places where connection is not only a vehicle for knowledge acquisition, but also for the alleviation of their own fear. Anxiety shapes the field experience, much as theoretical knowledge, access, and analytical acuity do.^{Footnote 103} Scholars are not immune to post-traumatic stress disorder (PTSD) and, more generally, to the psychological impact of this work, in the field and thereafter. There is a danger that they adopt the comforting but inaccurate belief that violence is “symptomatic of some social pathogen that is to be circumvented while maneuvering about the field” when, in fact, it is at the heart of their experiences.^{Footnote 104}

Amidst such insecurity, the researcher’s “friendship group” becomes her “safety net.”^{Footnote 105} In keeping with the “friendly” images social scientists wish to have of themselves,^{Footnote 106} they pursue relationships, in part for the genuine sense of trust they afford. Trust is the currency by which fear can be overcome, entry facilitated, and relationships built. At times, successful tribe-building on the researcher’s part may not only compensate for relative weakness, but may actually establish a new balance of power in her favor. After all, the development of dense connections can mean access to information, key stakeholders, protection, and understanding in ways that enhance a researcher’s capacity to operate within her surroundings.

The metaphor of tribe is especially apt, because the kin-based quality of tribal affiliations makes them stickier than more volitional relationships. As the saying goes, ‘you can’t choose your family,’ and the same goes for the tribe. The tribe is a collective, in which an individual can access a kind of strength and protection she otherwise could not; but that individual, especially a newcomer, will never be in full control of her position within the larger system nor can she pretend these bonds are without their own politics. There is a deterministic quality to tribal relations, because they are anchored in kinship: who you are will meaningfully dictate with whom you will connect. Field sites are not free, open markets for the trading of allegiances; instead, agency is limited. In the midst of contentious politics, a researcher may be unable to move as she pleases from one network to the next; and previous affiliations will shape her path going forward. In this sense, the concept of tribe reflects the limits on a researcher’s power and impartiality.

In Afghanistan, Somalia, and southern Turkey, we have adopted our own brands of tribal politics, incrementally joining, exiting, and returning to various, sometimes overlapping, networks to build our own tribes. While the range of networks to choose from may be less predetermined for us than for those we study, who we are and how we are perceived by others still greatly influence our tribe-building efforts. Nationality and ethnicity may have been two of our most obvious tribal markers. One of us, of Indian descent, consistently received a warm reception amongst Afghans, many of whom expressed affinity for India, its culture, and people; perhaps the widespread antipathy for the Pakistani government found in both Afghan and Indian societies was also a factor. The other likely penetrated northern Panjshiri networks in Afghanistan more easily by virtue of being French. The pre-existing relationships between members of the ‘French tribe’ and Panjshiri commanders—who received support from the French government and NGOs during and after the Soviet-Afghan war—greatly helped him in gaining entry and intimacy. In both cases, these associations facilitated understanding, access, and hospitality, making these identities an undeniable variable in the tribe-building equation.

A scholar's reputation and actions (“on” and “off-duty”) are also key to being admitted and remaining in different tribes.^{Footnote 107} Researchers can attract attention in ways that make them uncomfortable or that they fail to recognize altogether. As Stephen Brown noted, “the separation of the public and the private is a fiction that is hard to maintain in most research sites.” This, we would argue, is especially true in societies marked by tribal politics. One’s gender, sexuality, marital status, personality, and opinions all shape the experience of being an “object of interest.”^{Footnote 108} One of us learned early on, as a female researcher, to be particularly cautious about her relationships with male counterparts.^{Footnote 109} The “intersection” of her gender identity with her perceived nationality, “Indian,” may have further complicated expectations about how she should behave. Perhaps she was expected “to play by the rules” that applied to South Asian women?^{Footnote 110} Her ethnic background likely opened some doors, but a failure to comport with associated expectations may have threatened that access.

In any number of ways, then, our otherwise very similar scholarly experiences—as two junior scholars studying warlordism in post-2001 Afghanistan—were quite different from one another. Our different nationalities, genders, and personalities^{Footnote 111}shaped the trajectories of our research when it came to access, mobility, and protection. And our experiences thereafter, in Somalia and Turkey, were distinct from those we had forged in Afghanistan, as we worked to build new tribes in very different societies.

The notion of ‘building a tribe’ has, once again, a number of analogs in Western intervention. Imperial practices in the Middle East, for example, involved the cultivation of close relationships with tribal leaders. T.E. Lawrence explained how he was “sent to these Arabs as a stranger, unable to think their thoughts or subscribe their beliefs, but charged by duty to lead them forward and to develop to the highest any movement of theirs profitable to England in her war.”^{Footnote 112} Tribal engagement has been resurrected in the 21^st century as a sometimes celebrated, often controversial instrument of statecraft and unconventional warfare.^{Footnote 113} But, from the “Anbar Awakening” campaign in Iraq to the Shinwari tribal pact in Afghanistan, many argue that Western efforts at building tribes come with real risks.^{Footnote 114}

And, yet, we have come to realize that, operating in a hostile environment in which our positions are ones of limited power, we have had little choice but to undertake our own smaller versions of tribal politics to succeed and, sometimes, survive. Of course, our tribe-building does not involve coercion and capital of the kinds characteristic of military intervention, nor are our motives strategic or ideological. Nonetheless, it would be foolhardy to assume that they can be universally interpreted as apolitical.

The Risks that Come with Being ‘All in the Family’

While joining and building a tribe may at times be necessary to conduct research in hot conflict zones, it also involves dangers. What was true for Philippe Bourgois twenty-five years ago—“we are outsiders…we have a formidable capacity unwittingly by our mere presence to cause trouble or to seriously complicate matters”—remains true today.^{Footnote 115} To start, joining a tribe implies embracing, knowingly or otherwise, its existing friendships and enmities. The researcher can become a target of violence or get caught in the crossfire. The base in Afghanistan on which one of us spent a few days was attacked years later, while the Mogadishu hotel in which the other stayed also came under assault. The Somali warlord who had provided protection (the owner) was killed, as were a dozen hotel guests.

Tribes can also involve a tacit degree of reciprocity and a set of obligations to the researcher’s new ‘family.’ In warzones, this can become ethically dubious^{Footnote 116} and dangerous. “You have to fight for us now,” a Somali warlord hosting Reno jokingly told him after his men were targeted in a suicide bombing. “In the back of your mind,” said Reno, “you know your relationship is not like staying in a hotel. The guest belongs to the host.”^{Footnote 117} This problem is magnified by post-9/11 U.S. legislation prohibiting the provision of “material support” to “terrorist groups.” Tribe-building may now have grave legal implications for those studying armed groups. The notion that communication with a designated terrorist organization could be construed as illicit support introduces an “inevitable chill on much-needed scholarship.”^{Footnote 118}

There are also risks researchers impose on those within their networks. These are further aggravated by the relative lack of anonymity both within the tribe—where “interviewees either know or can trace each other”^{Footnote 119}—and outside the tribe, as foreigners tend to “[draw] attention to those in their orbit.”^{Footnote 120} In one very uncomfortable situation with Afghanistan’s second vice president, one of us was asked whom he had met prior to this visit. He could do nothing but witness his interpreter divulge the names of the people interviewed earlier that day. Although this episode had no further consequences, it points to the lack of anonymity and the “dilemmas of proximity”^{Footnote 121} that can result from being ‘all in the family.’ Researchers also have little to no control over the life and career paths of their fellow tribesmen once their times together draw to an end: Interpreters, fixers, friends, and informants may rise in prominence and become players in the systems they were hired to study. Their decisions about what information to share and with whom may have unforeseen consequences as well.

There is also the concern of betrayal by the scholar—real or perceived—and the possibility that, once work is published, informants will learn of and be angered by the interpretations put forth.^{Footnote 122} For participant-observers, this risk exists from the start, as researchers toggle between the identities of “a ‘good friend’ and a ‘snooping stranger.’”^{Footnote 123} But, to varying degrees, this is a concern for all field researchers^{Footnote 124} that lies at the heart of what is almost always an exchange based on “partial trust.”^{Footnote 125} As Fine mused, “the illusion of being more sympathetic than we are aids research but is deceptive…Our informants have given us a ‘gift,’ but how have we returned the favor?”^{Footnote 126} Also problematic is the question of what a scholar owes “repellent” subjects in terms of not harming their reputations.^{Footnote 127} The simple act of describing an Afghan leader as a ‘warlord’ could create irreversible tension with informants in his circle. Militia commanders, opposition leaders, business tycoons, and government officials are not invulnerable to exposure and critique. On the contrary, their power rests, at least in part, on their ability to project credible authority.

The possibility of piquing one’s research subjects can also have implications for one’s ability to return. Scholarly work can now be accessed in just about every corner of the world by anyone with an Internet connection. As a function of technological innovations, the residue of interactions between researchers and their informants, fixers, interpreters, and hosts is stickier than ever before. Mobile phones, email, virtual libraries and bookstores, online newspapers and magazines, think tank websites, social networking platforms, and so on make it difficult to ‘exit’ the field. Researchers now have “digital profiles.”^{Footnote 128} The days are long gone when research “was done in remote areas where there was little possibility of reports of findings coming to the attention of those studied.”^{Footnote 129}

There is, of course, the reverse risk: Bias inevitably creeps in as one comes to know and even like the people one is studying. As detailed above, true objectivity is an illusion, but researchers must be aware of the degree to which they are at risk of losing it altogether.^{Footnote 130} Scholars experience fieldwork, not only through rational observation and analysis, but also through sentimental reactions that can shape their conclusions.^{Footnote 131} Profound biases may be insurmountable if a researcher only interacts with those within “the specific network being accessed.”^{Footnote 132}

Researchers can, inadvertently, take sides in surrounding power struggles in ways that block them from establishing links with others or put them at risk.^{Footnote 133} Alice Goffman, who spent years in a tough West Philadelphia neighborhood, admitted her “desire for vengeance” after one of her friends (and research subjects), was murdered—she drove herself and a friend (and research subject) around in search of the killer.^{Footnote 134} Tessa Diphoorn witnessed a vicious attack perpetrated by members of a South African private security team within which she was embedded. She wondered if her role as “audience” “influenced their performance of violence … and triggered a particular type of macho behavior.”^{Footnote 135}

Scholars must acknowledge their own emotional reactions as part of the research experience, including moments when they find themselves paralyzed, excited, or even sympathetic in the face of violence or power.^{Footnote 136} Antonius Robben found himself taken in by alleged Argentine war criminals, whose reputations did not comport with their “great civility…considerable knowledge of literature, art, and classical music…[and] affability and chivalry.” He described this discomfiting phenomenon as “ethnographic seduction.”^{Footnote 137}

One of us, after meeting with a famed warlord, reported: “He has a very unique way of looking into someone’s eyes. His tiny half-closed eyes seem to see right through your soul.”^{Footnote 138} This ‘celebritization’ of the research subject is a reflection of the real and perceived power armed actors can impress upon others. This was, in fact, an important dimension of our own research about them. And, yet, as visitors to their lavish palaces, we felt intimidated and excited by proximity to such blatant manifestations of power. We were also disarmed by the adeptness with which these strongmen engaged us. Seducers (conscious or otherwise) come in the form, not only of the bogeymen of violent conflict, but also civil society leaders and even victims of violence.^{Footnote 139}

Another, related risk, for the researcher, is the delusion of belonging when, in fact, she can never become a true insider. She may borrow and learn from the tribe she joins, but will never belong in the same way her native counterparts do.^{Footnote 140} Social media networks can be useful modern tribe-building tools, but they further this delusion. On the one hand, they allow researchers to maintain relationships with informants and research subjects;^{Footnote 141} they also open the door to a new kind of data collection about violent political actors living in secluded areas (Islamic State militants in Iraq and Syria, for example).^{Footnote 142} On the other hand, building a virtual tribe through social media can impute a false sense of kinship—Facebook “friends” are not necessarily real friends; it also introduces “the danger of becoming too publicly or privately friendly with sources, at the risk of giving them a higher profile or being perceived as validating their views.”^{Footnote 143}

The sense of belonging also carries with it a misplaced sense of knowing that has real implications for the researcher’s thinking, analysis, even safety:

Many researchers have experienced this ‘eureka!’ moment. Only after the encounter does it become clear that their emotional experience has betrayed their humility as analysts.^{Footnote 145} Such encounters can represent one of the great thrills of fieldwork and need not necessarily signify a failure to understand or learn. But they do represent a risk implicit in the tribe-building endeavor.

Innovating from a Distance

Today, little field research is being done in particularly violent war-torn regions such as Syria and southern Somalia, where the security conditions severely limit access to relevant populations.^{Footnote 146} Still, many researchers continue to venture into the hinterlands of conflicted countries to establish their credibility as experts with ‘ground truth.’ But even they must revisit their thinking about the conduct of this kind of research. In the most dangerous settings, some restrict themselves to conducting interviews in the country capital, while others limit themselves even further, inviting informants to meet at a few designated ‘safe’ places. Others may hire local researchers to conduct interviews and surveys on their behalf. Interviews in country capitals can be extremely informative, while collaboration with local researchers can add great depth to inquiries previously pursued only by outsiders.

New technological tools also afford researchers insights into places from which they are otherwise excluded. Jacob Shapiro and Nils Weidmann employed data about Iraqi mobile phone network coverage to investigate the impact of “improved communication” on “the rise or fall of violence during insurgencies.”^{Footnote 147} David Mansfield combined rich fieldwork with geospatial imagery to study poppy cultivation in Afghanistan.^{Footnote 148} Alexander De Juan and Andre Bank utilized satellite imagery to measure the provision of electricity in regime-controlled territories within Syria alongside crowdsourced accounts of war-related fatalities to establish patterns of violence.^{Footnote 149}

These are innovative approaches that enrich data collection; however, if they are used exclusively, they can introduce a space between researcher and researched. If so, they become reminiscent of the “verandah model” of ethnography, with “the field-worker, sitting on the verandah of the government station, sending for a ‘native’ who will be subjected to several hours of systematic questioning about his … language and customs.” Joan Cassell characterized this approach as “historically obsolete,” but we would argue that versions of it will grow more prevalent as conflict zones become increasingly inhospitable to Westerners.^{Footnote 150}

Here there are clear parallels to intervening actors once again—international aid, development, and diplomatic organizations—whose efforts in the field have been critiqued because of the way they operate in war zones.^{Footnote 151} Seclusion, albeit often reluctant, has been a hallmark of the aid worker’s life in 21^st-century warzones. This is increasingly the case in cities like Kabul where violence against foreigners has become more norm than aberration. Foreign researchers are by no means immune from this trend. In 2014, two scholars teaching at the American University of Afghanistan were killed in an attack on a popular restaurant in Kabul; in 2015, a researcher was killed on a visit to Kabul in an attack on her hotel; and, this year, two more researchers associated with the American University were kidnapped in Kabul, and the university campus came under assault, resulting in over a dozen deaths.

The analog between research intervention and aid intervention is acutely evident in Syria and Somalia. One of us set out to study the Syrian conflict from within Syria in 2013 but, on account of the deteriorating security situation, had little choice but to work from within southern Turkey. Similarly, Western donor organizations were forced to understand politics, forge relationships, and deliver aid from over the border. Those studying southern Somalia must operate primarily from Kenya or Somaliland, much like their NGO counterparts. The view “from the verandah” can give scholars leverage on certain subjects—like displacement, diaspora relations, and regional geopolitics—but interviews outside the conflict ‘theater’ have their own problems (sample representativeness, to name one).

Moreover, researchers (and their intervening counterparts) cannot rest easy that ‘safe’ places, proximate to but removed from conflict, will always remain so. The security situation in cities along the Turkish-Syrian border can no longer be taken for granted; this year, a massive suicide-bombing shook the town (and research site) of Gaziantep. The larger political climate in Turkey is also cause for concern. Turkish President Erdogan has singled out academics—local and foreign—for their critique of his government. It bears remembering that the pursuit of field research has proved challenging, not only in countries in conflict, but also in those with “hybrid regimes,” such as Vladimir Putin's Russia or Saddam Hussein's Iraq.^{Footnote 152}

Keeping Our Feet in the Field

Since the U.S. wars in Afghanistan and Iraq, Westerners of all stripes have become targets of violence in largely unprecedented terms. In these conditions, scholars of 21^st-century warfare may need to reacquaint themselves with the verandah, but many are not yet prepared to forsake the field. We argue that this kind of fieldwork can still be pursued but only if researchers acknowledge and engage with their positionalities in the face of violent conflict. Researchers may have no choice but to build tribes of their own. There are a host of risks—physical, psychological, analytical, even legal—that accompany this endeavor, but they are not necessarily insurmountable. Here, we identify three ways of mitigating against them. We do not claim that these are the only ways to deal with deteriorating conditions or that they are foolproof; instead, we simply offer insights from our own experiences.