The ‘problem’ of undesigned relationality: Ethnographic fieldwork,dual roles and research ethics

Introduction

Whatever else ethnography is, it is about building, negotiating, losing, and celebrating relationships with fellow human beings. (James S. Bielo, 2015)

In this approach [ethnography], there is a profound acknowledgement of the relationality of the human subject. Furthermore, to talk of the ‘field’ is to talk of an entity which is itself relational and not merely spatial. (Bob Simpson, 2011)

Ethnographic knowledge is always relational, the product of multiple cross-cutting conversations across diverse contexts. (Wardle and Gay y Blasco, 2011)

If we disregard the dimension added by social media, the broader dilemma posed by the student is far from new: all ethnographers must grapple with the socially ‘contaminated’ nature of fieldwork, and the crosscutting demands and possibilities it entails (Lederman, 2007). However, the primary reference point for the student and my colleague was the Tri-Council Policy Statement: Ethical Conduct for Research Involving Humans – a set of national guidelines that inform institutional ethics review processes at universities throughout Canada (CIHR, NSERC & SSHRC, 2014). Although this document is frequently valorized as overcoming a biomedical model of research ethics (e.g. Tolich and Smith, 2014; Nichols, 2015), where discussed, relationality beyond the strict confines of the researcher/participant role is treated as a problem to be solved. ¹

In this essay, I aim to consider the ‘undesigned relationality’ (Lederman, 2013, 2016) that characterizes ethnographic fieldwork, along with the ways it is understood – both in the conceptual frameworks that underpin procedural ethics review and by the different disciplines that have come to claim a stake in it. As Rena Lederman (2017) observes, if anthropologists and other ethnographers are to better grapple with the implications of institutional ethics oversight, we need to devote further attention to how our ethical conundrums appear from the vantage point of research ethics administrators and our colleagues in neighbouring disciplines. This is clearly a topic of some political import: while the clash between ethnographic research and the prevailing bioethical framework is largely taken as given by ethnographers who have experienced institutional ethics review processes, it has very real effects on the ground (see Wynn, 2011). As Bob Simpson (2011: 377), writing specifically about anthropology, observes, ‘The issues at stake not only are practical and procedural but also touch on deeper ontological and epistemological concerns about the reproduction of the discipline and the practice of ethnography as its core methodology’.

Any discussion of ethics and ethnography is complicated by the fact that the term covers an array of disciplinarily-inflected practices and refers to both a process and a product. The latter perspective is dominant in anthropology, where ethnography is not so much a research method but an intellectual enterprise informed by ‘fieldwork’, which includes, but is by no means limited to, ‘participant observation’. ² As Blommaert and Jie (2010) observe, this means that the basic architecture of ethnography as anthropologists understand it ‘already contains ontologies, methodologies and epistemologies … that do not necessarily fit the frameworks of other traditions’ (p. 6, emphasis in original). Where ethnography has crossed over into other fields, it is typically decoupled from these moorings. ³ For example, although sociology has a well-developed tradition of ethnographic research, unlike anthropologists, who tend to be concerned with the epistemological foundations of ethnography, they are more focused on formulating methodological guidelines for its conduct ⁴ (Tavory and Timmermans, 2009).

However, although it is possible to speak of ethnography as a broadly disciplinary practice, and I am interested in teasing out these differences, I want to make it clear at the outset that such characterizations must inevitably be idealized representations that do not map neatly onto the ways that ethnographers perform and understand it. Certainly, there are many scholars outside of anthropology who exemplify dimensions of the ‘ethnographic imperative’, just as there are many anthropologists who do not do fieldwork of the kind historically associated with the discipline (Kapferer, 2007). But despite the risks of overgeneralization, it is important that we attend to differences in the practice of ethnography in order to avoid privileging a particular disciplinary sensibility under the official guise of neutrality. Therefore, my goal in what follows is to highlight the multiple epistemologies the term disguises, and to emphasize the need for ethical pluralism in how we respond to the practices it entails.

Procedural ethics, multiple relationality and ‘dual roles’

Regardless of its particular disciplinary configuration, the multiple forms of relationality that characterize ethnographic fieldwork are often framed in terms of the tension between ‘participation’ and ‘observation’, which is standard fodder for ethnographic methods textbooks (e.g. Wolcott, 1999; O'Reilly, 2005). This issue has been long discussed by ethnographers – from Evans-Pritchard's (1973: 3) observation that through ethnographic fieldwork the anthropologist simultaneously ‘enters into a culture and withdraws from it at the same time’ to Byron Good's emphasis on the ‘duality of the anthropologist's role as critic and participant’ (1994: 27). Indeed, these contradictory qualities of ethnographic research are not infrequently glossed under the label of the ‘dual role’ itself. For example, in their textbook Ethics and Anthropology: Dilemmas in Field Work, Michael Rynkiewich and James Spradley refer to the ‘contradictions inherent in the dual roles of detached observer and participant-friend’ (1976: 126), and similar characterizations are to be found in the work of anthropologists (e.g. Bielo, 2015) and ethnographic sociologists alike (e.g. Prus, 1996).

In an unfortunate terminological coincidence (or perhaps not – more below), the concept of the ‘dual role’ is also widely used in procedural ethics guidelines to refer to a situation where a researcher has a relationship with participants in a pre-existing professional capacity. Given the oft-observed biomedical underpinnings of this framework, and its attendant ‘ethics of the body’ (B. Simpson, 2011), it should come as little surprise that the prototype for the dual role is the clinician-researcher conducting a study on his or her patients. This is the context that the concept was initially designed to address (e.g. Churchill, 1980), although it has subsequently been extended to other types of relationships—those between therapists and clients, between teachers and students, and so on.

Importantly, the dual role is not seen as intrinsically unethical, but as a phenomenon – like other conflicts of interest – that should be carefully managed. This is what the Tri Council Policy Statement has to say on the topic:

Dual roles of researchers and their associated obligations (for example, acting as both a researcher and a therapist, caregiver, teacher, advisor, consultant, supervisor, student or employer) may create conflicts, undue influences, power imbalances or coercion that could affect relationships with others and affect decision-making procedures (e.g., consent of participants). … To preserve and not abuse the trust on which many professional relationships rest, researchers should be fully cognizant of conflicts of interest that may arise from their dual or multiple roles, their rights and responsibilities, and how they can manage the conflict. (CIHR, NSERC & SSHRC, 2014: 99)

Lederman (2007, 2016) argues that such ethical stipulations reveal much about the kinds of ideal relationships between researchers and participants that the ‘Standard Model’ is premised upon. As she observes, especially since the Second World War, biomedicine's epistemological value has relied on the careful delineation of research from therapeutic practice, with the two defined in conceptual opposition. In Lederman's words:

… the Belmont Report locates and buttresses the conceptual wall between two kinds of relationships on ethical grounds: therapeutic relationships into which persons enter with a conventional expectation that their individual care is the object have different ethical entailments from research relationships into which persons enter with a conventional expectation that socially beneficial knowledge is the objective. (2016: 48)

According to Paul Roote Wolpe (1998), medical research in the 19th century was primarily done by physicians on their own patients. ‘There was no clear theoretical or practical differentiation made between standard practice and experimental therapy’ (p. 51). Obviously, the notion of a dual role only acquires significance when ‘treatment’ and ‘research’ become conceptually separated, as happened with the rise of clinical epidemiology and its accompanying methods – most notably the randomized controlled trial. This led physician-researchers to ‘start looking for subject pools in orphanages, the army, insane asylums, and state hospitals’ (Wolpe, 1998: 51) as opposed to their own clinics.

Importantly, under the logic of the randomized controlled trial, an existing relationship between the physician and patient was treated as potentially jeopardizing the objectivity of the research. This is partly why the gold standard for clinical research is the double-blind randomized controlled trial, where neither the researchers nor the patients know who has received the treatment and who hasn't until after the study has been completed, thereby minimizing sources of bias and managing the contaminating influence of the doctor as a therapeutic agent independent of the treatment itself (see Moerman, 2002).

Ethnography, disciplinarity and undesigned relationality

Although this emphasis on clearly delineated and carefully managed relationships between the researcher and researched reaches its apotheosis in medical experiments, it underpins many kinds of social research – or what Harold Garfinkel (2002) disparagingly called ‘the worldwide social science movement’. As Lederman (2013) observes:

… mainstream social and behavioral sciences study relations by means of relations; however, they do so by constructing specialized research relations with heavily managed, conventionalized expectations concerning contact with research participants. They construe undesigned relationality in the research as noise distorting reliable, objective results. (pp. 599–600)

This view of how social science knowledge is produced is one that has been challenged from a variety of quarters, with the most strident critiques coming out of socio-cultural anthropology and ethnomethodology ⁵ – not coincidentally, fields where undesigned relationality is most valued. Writing specifically about anthropological ethnography, Johannes Fabian (2007) observes that a distinct change occurred in the discipline when anthropologists realized that fieldwork was not simply about collecting information. As a product of interaction, ethnography was intrinsically performative; thus, what we take away from fieldwork as ‘data’ is made more than found. In Fabian's (2007) words:

what we learn often does not come as responses to our questions but is enacted in, and mediated by, events which we may trigger but cannot really control (every simple interview is such a performative event). Similarly, concern with interpretation and hermeneutics made us aware of positive naiveté regarding the relations between research and writing in the production of ethnographic knowledge. (p. 13)

This underlying positivism is evident in the overview of fieldwork relations presented in the chapter on qualitative research in the Tri Council Policy Statement. According to the document: ‘the emergent nature of many qualitative studies makes the achievement of rapport with participants and feelings of interpersonal trust crucial to the generation of questions considered important or interesting by both parties, and to the collection of dependable data’ (CIHR, NSERC & SSHRC, 2014: 142). Here, we see the ways in which relationality becomes tied to truth and trustworthiness: once again, rapport serves the interests of generating ‘dependable’ data. The instrumental dimensions of relationality become even more explicit in the assertion that, ‘In some cases, gathering of trustworthy data is best achieved by closeness and extended contact with participants’ (p. 141). In this framing, relationality is perceived primarily as a means of transforming perceptions of the researched about the researcher and the attendant quality of the data produced.

According to some accounts, this is a core feature of ethnography. For example, Peter Magolda (2000) discusses the importance of trust and rapport in generating ‘good’ ethnographic data and Lisa Weems (2006: 997) highlights the ways in which reciprocal relationships have been ‘linked to the potential of ethnographic research to generate “more accurate” or “truthful” understanding by closing the distance of interpretation and identification between the researcher and researched’. Echoing this view, Julia O'Connell Davidson (2008: 54) writes: ‘It is widely recognized that to achieve an interpretative understanding of a given social practice, subculture or culture, the researcher must get “close” to her or his research subjects’.

However, this view of relationality is truer of sociological versions of ethnography than anthropological ones, ⁶ although it stems not just from the disciplines' somewhat differing orientations to knowledge production but other divergences in approach. Exploring the two disciplines' contrasting stances on the ethical acceptability of deception in ethnography, Lederman (2017) observes that the sociological point of view is often perceived as in tension with that of the objects of study. Likewise, in Katz and Csordas' (2003) discussion of anthropologists' and ethnographic sociologists' approaches to phenomenology, they observe that anthropological writings ‘characteristically have illuminated native groundings for subjects' perspectives’, whereas sociological writings are more likely to break with members' perspectives, ‘deconstructing what subjects treat as naturally significant’ (pp. 275–6).

As a result of these differences, Lederman (2017) argues that ethnographic sociologists are more likely to write about the insider/outsider distinction ‘as if their sociological, observer, outsider positional identities are the “real”, essential facts’ (2017: 14), and to represent their fieldwork relations as inauthentic. This is not to say that they perceive such relations as insincere or fake, but more that they are viewed along the lines of strategically productive ‘roles’ that aid the ethnographer's primary identity as a researcher. Where anthropology differs, according to Lederman, is its embrace of ‘open-ended intimacy as a scholarly knowledge practice’ (2013: 598). In the classical fieldwork situation, anthropologists enter unfamiliar social terrain, learning its mores by developing practical competence in the local language and those aspects of daily life enabled by relationships cultivated over a lengthy period of time. Their field identities are multiple and equally ‘real’: ‘anthropologist’, ‘friend’, ‘student’, ‘apprentice’ and all manner of relational attributions offered by their hosts (Lederman, 2017).

In this context, relationality is not conceived primarily as a means of transforming perceptions of the study participants. If anything, the desired transformation is in the anthropologist's perception, as opposed to that of the community she works with (see, for example, Evans-Pritchard, 1973). As Bruce Kapferer (2007) observes: ‘It is through the reflexive deconstruction of the anthropologist's own preconceptions in the fieldwork encounter that knowledge is gained, not just of the fieldworker's own taken-for-granted realities but also of those with whom the fieldworker engages as a function of the encounter itself’ (p. 81). These different views of relationality therefore have consequences for how its ethical implications are grasped (B. Simpson, 2011; Lederman, 2016, 2017).

The ethics of relationality as normal vs. exceptional

For anthropologists, the moral tensions of fieldwork arguably differ in degree rather than in kind from the tensions raised by human interaction in all its forms, with their vacillations between ‘sincerity and insincerity, genuineness and hypocrisy, honesty and self-deception’ (Geertz, 1968: 155). Viewed in this light, the perils of ethnographic fieldwork, as with its promise, stem as much from the anthropologist's humanity as their professional identity as a researcher. As Bob Simpson (2011) observes:

Such an approach takes us beyond a rule-based notion of ethics and into a much more complex terrain in which the skill of the ethnographer lies in developing and managing relations founded on trust, respect, and an avoidance of delimiting the subject. In short, the endpoint of this kind of engagement is not the body but the complex and dynamic relational worlds which anthropologists enter and in which research subjects live and exercise their own ethical and emotional sensibilities, realize deliberative strategies, and do so with the benefit of accumulated experience and knowledge (p. 385).

This sense of fieldwork dilemmas as human dilemmas comes across clearly in the work of Murray Wax (1977, 1995), an early and vociferous critic of the extension of human subjects regulation to anthropological research. He writes:

As opportunities present themselves, fieldworkers doubtless are tempted, fall into error, and harm their associates or the persons whom they are studying. If fieldworkers did not suffer human frailties, they would be likely to be unable to comprehend the ongoing human lives of those they wish to study. Moreover, even with the best of intentions, a fieldworker may disrupt normal events, because he is a stranger and ignorant of local custom. Yet the facts are that living is full of risks and troubles, and in any human interaction, those involved may comport themselves more or less strictly and with more or less regard for others. (Wax, 1977: 325)

The seminal case here is that of Laud Humphreys. Although a biography by fellow sociologists (Galliher et al., 2004) suggests that Humphreys (1970) was as much a participant as an observer of the anonymous bathroom encounters he documented in Tearoom Trade, this was not something he ever acknowledged, even after he came out publicly as gay. Galliher, Brekhus and Keys (2004) ultimately explain this not merely by reference to the illegal nature of such activities, but the strength of the requirement to present oneself as a neutral observer. In their words: ‘Even though “the existence of saucy tales of lurid assignations” abound among ethnographers, researchers never admit to such liaisons’ (p. 30).

Compare such depictions of sex in the field with the treatment of the topic by Russell Bernard (2013), the anthropologist who has written most extensively about field methods. ‘It is unreasonable to assume’, starts a section titled ‘Sex and Fieldwork’ in one such tome, ‘that single, adult fieldworkers are all celibate’ (p. 332). As Jill Dubish (1995) observes, although anthropologists absorb the idea that sexual relationships in the field ‘would neither be appropriate nor wise’ ⁷ (p. 30), such relationships are relatively commonplace. Not only are they discussed informally amongst graduate students, but they manifest in instances of anthropologists marrying people from their fieldsites, and in sporadic allusions to such encounters in ethnographies themselves. ⁸ In essence, although a kind of ‘don't ask, don't tell’ policy prevails in the discipline, ⁹ where such relationships come to light they are typically chalked up to anthropologists ‘being human’, and seem to crystalize the underlying view that anthropologists' roles – as ‘researcher’, ‘guest’, ‘friend’ (and, indeed, ‘lover’) – cannot be readily parsed. ¹⁰

Implications for conceptualizing ethnographic research ethics

These differing understandings of relationality (and the epistemologies that underpin them) have important consequences for how the broader ethics of ethnographic research are conceptualized and responded to. Compare, for example, Murray Wax and the ethnographic sociologist Charles Bosk on the difficulty of communicating the nature of ethnographic research to one's interlocutors:

Given that our hosts and we usually inhabit two different conceptual universes, it has been notoriously difficult to communicate who we are and what are our goals … Even when people have the concept of ‘anthropologist,’ it is often misguided, and we are easily confused with welfare workers, missionaries, and other folk, and made scapegoats by local politicians. Some anthropologists have behaved poorly; others have behaved with exquisite courtesy, parity, and consideration. (Wax, 1995: 330)

[E]thnographers trade quite freely on an almost universal misunderstanding between our research subjects and ourselves … Because our subjects are flattered by our attention, we are allowed to obtain data that it is not necessarily in our subjects' best interest to reveal. Few of us have ever reported that we inform our subjects of this fact. Rather, because it so often serves our purposes so well, we encourage any misperceptions that yield rich data. (Bosk, 2001: 206)

Arguably, if fieldwork relations are conceptualized as a strategically productive role one is playing, those aspects of relationality that are otherwise normal take on added significance, becoming laden with potentially sinister undertones. Thus, the sociologist Judith Stacey (1988: 22–3) argues that: ‘Precisely because ethnographic research depends upon human relationship, engagement and attachment, it places research subjects at grave risk of manipulation and betrayal by the ethnographer’. In such accounts, the fault lies not so much with the individual's capacity to behave ethically but with the ‘moral failings built into the structure of ethnographic research’ (Bosk, 2001: 200). In effect, fieldwork becomes a tradeoff between data and ethics, with a perceived ‘conflict of interest’ between the ethnographer's ‘researcher/observer’ and ‘participant’ hats (Stacey, 1988). Viewed in this light, ethnographic research is not so far from the concept of the ‘dual role’ promulgated in research ethics guidelines, and appears broadly amenable to its prescriptions.

We see this clearly in the vignette discussed at the beginning of the paper. Regarded through the lens of the ‘dual role’, existing friendships within the fieldsite need to be separated from the research. ¹¹ Recall the trinity of responses specified in the Canadian research ethics guidelines for responding to the dual role: eliminate, manage, minimize. This is precisely what my colleague tries to do, asking: ‘Should the researcher be pretending that she doesn't see the Facebook account? Should the researcher terminate her Facebook friendship during the research?’ In this fashion, the ethical conflict is managed in much the same way as advised for the prototypical doctor-patient relationship.

For those ethnographers who perceive fieldwork as a tradeoff between data and ethics, with an attendant conflict of interest between the ethnographer's ‘researcher/observer’ and ‘participant’ hats, this represents a potentially viable response. Certainly, we see various examples of researchers who have responded in much this way to the quandaries raised by undesigned relationality. For example, the edited collection Ethical Quandaries in Social Research contains a chapter by the sociologist Zethu Matebeni (2014) focusing on her experiences as a black lesbian exploring black lesbian identities and sexualities in Johannesburg amongst women of her acquaintance. Focusing on the access she had to her informants' lives beyond their formal representations of themselves, she writes: ‘While difficult, it was often necessary to separate my two interrelated identities of researcher and friend/comrade’ (p. 112). Somewhat to Matebeni's surprise, the signed consent form became a core means of managing these ‘dual’ roles. In her words, ‘When people signed a consent form, this marked a boundary, and a particular relationship between us. In essence, consent forms were the tools that defined the research relationship’ (pp. 112–13).

This approach only becomes possible, however, when one's ‘roles’ are conceptually separable, which I have argued is antithetical to disciplinary conceptions of fieldwork within anthropology. An instructive comparison is the sociologist Julia O'Connell Davidson's (2008) account of the friendship she developed with ‘Desiree’, a sex worker and informant in O'Connell's ethnographic study of prostitution in the UK, and the anthropologist Paloma Gay y Blasco's writings on the friendship she developed with Liria de la Cruz Hernández, a Gitana street seller in Blasco's native Madrid (see Gay y Blasco, 2011, 2017; Gay y Blasco and Hernández, 2012). Although both researchers raise similar ethical concerns about ethnography, they locate the source of these problems somewhat differently.

For O'Connell Davidson, the core dilemma she confronted in her fieldwork related to her friendship with Desiree, which she perceived as both a natural outcome of the research and in fundamental tension with it (in other words, the dual role problem). O'Connell Davidson initially attempted to deal with this tension by placing boundaries around her relationship with Desiree – an approach that Desiree resisted and that became increasingly difficult as they became more entrenched in each other's social circles. Ultimately, O'Connell Davidson relied on drawing lines between what counted as research ‘data’ in her fieldwork and what didn't, based on her sense of what information Desiree might feel betrayed by if shared. Thus, the crux of the problem, as O'Connell Davidson perceives it, relates to the possibility of informed consent and the ways in which ethnographic research requires interlocutors to consent to their own objectification and appropriation.

Blasco, too, is concerned with the role of informants as ‘objects rather than creators of anthropological knowledge’ (Gay y Blasco and Hernández, 2012: 1). But where O'Connell Davidson perceived her friendship with Desiree as in tension with the research, this concern is notably lacking in Blasco's and Hernández's accounts, despite their growing entanglement in each other's lives – an enmeshment that culminated in Gay y Blasco helping Hernández to elope with a younger non-Gitano man. Instead, Gay y Blasco's (2017) main focus is the ways in which the relationality upon which ethnographic fieldwork is based disappears in the act of writing about it; thus, the primary tension she identifies is between the ‘conflicting demands of reciprocity and ethnographic writing’ (p. 95) as opposed to friendship and research per se.

The contrast between O'Connell Davidson's and Gay y Blasco's positions is subtle, but it speaks again to underlying disciplinary differences in conceptions of the ethics of relationality. Simply put, anthropologists typically cannot parse their identities as ‘researcher’ and ‘human’ in the field because ethnographic fieldwork is not merely perceived as a technique for gathering ‘data’. ¹² The result of such an attempt can only be methodological and conceptual fragmentation, whereby ‘things that ought to be all of a piece are split apart’ (B. Simpson, 2011: 382). Instead, as Gay y Blasco's account makes clear, it is in the process of writing that the ethical issues raised by relationality are experienced most acutely, which is why the vast weight of disciplinary musings have focused on the morality of representation as opposed to the ethics of fieldwork per se, which, as previously described, is understood as relative and situational (see Castañeda, 2007).

Indeed, it's worth noting that even when anthropologists invoke the ‘dual roles’ framing, they use it to speak primarily to issues of representation, not to bracket the fieldwork experience itself. This is clearly evident in Rynkiewich and Spradley's (1976) discussion of dual roles, where they go on to state that:

The ethical implications of publication thus stem from the dual roles I played as: (1) anthropological researcher with the goal of extending our understanding of human behavior and (2) participant in Danish daily life, who, as a human being, developed close personal relationships with many of the Danes I met. (p. 131, emphasis added)

The kind of ethical engagement being advocated, embracing as it does an ethic of relationality that draws together process and product, is inimical to institutional ethics review processes, which are almost exclusively concerned with the ‘doing to’ that happens during ‘data collection’ itself ¹³ (see Bell, 2014). The unfortunate consequence is not only that such processes fail to address the real ethical issues that ethnographers confront, but that ethnographic fieldwork – especially of the kind based on the epistemological premise (and promise) of undesigned relationality – is made to appear ‘downright deviant on practically every count’ when judged through the lens of such models (Lederman, 2006: 485).

Conclusion

From an epistemological and ethical standpoint, ethnographic fieldwork is distinctive insofar as it relies on undesigned relationality. Where other approaches ‘aim to construct a purified relation between knower and known, carefully bracketing other potential identities and expectations by controlling the contexts and media of their engagement’ (Lederman, 2007: 319), ethnographers embrace it as a way into grasping social worlds. Unfortunately, it is precisely this feature that is the source of ethnographers' difficulties with institutional ethics review – and the requirements around informed consent, risk-benefit analysis, etc. – that comprise the heart of the review process.

The ‘dual relationship’ seems to be the only paradigm available within the Standard Model to conceptualize research contexts where the researcher does not have a fixed identity as a researcher. However, this is then transformed into a ‘problem’ to be managed or solved. In many respects, this response speaks volumes about the underlying conception of relationality that dominates biomedical and mainstream social and behavioral research. The ideal being strived for is that of a pristine ‘researcher’/‘subject’ relationship free of other social entailments, or where such entailments are put in direct service of the research itself (e.g. establishing ‘rapport’ to improve the quality of the data furnished).

Obviously, the fact of a mismatch between the assumptions of the Standard Model and the particularities of ethnographic fieldwork is a far from novel insight (B. Simpson, 2011). However, less recognized is the fact that attempts to accommodate qualitative research don't necessarily resolve the underlying problem. I'm on the record as having serious doubts about whether such ‘accommodations’ can be anything more than superficial lip service in the face of a commitment to a universal code of ethics. However, the situation is further complicated by the fact that even named approaches such as ‘ethnography’ cannot be assessed in isolation from their disciplinary moorings. As I have argued elsewhere, ‘There is a real danger that certain styles of ethnography (or phenomenology, etc.), especially those more amenable to established procedural guidelines, will become the standard against which other forms of ethnographic fieldwork are measured’ (Bell, 2016).

As I have hopefully demonstrated, certain versions of ethnography lend themselves better to procedural ethics review processes because they are based on a conception of relationality that affirms the essential rightness of the Standard Model of the researcher-participant relationship. For some ethnographers, especially those trained outside of anthropology, intimacy as a scholarly knowledge practice appears intellectually sloppy and ethically suspect from the perspective of their own disciplines (Lederman, 2017), with important implications for how they conceive and respond to the dilemmas it raises. In particular, it often leads to efforts to make ethnographic research more commensurate with the mainstream model: namely, by creating boundaries around the ‘research’ role and predefining what constitutes ‘data’ and what does not. However, this approach becomes inconceivable when open-ended intimacy itself is the premise, and where process and product are enmeshed in both name and deed. Thus, if ethnography has been ‘indigenized’ to a certain extent in the different disciplinary contexts in which it has been taken up, then it behoves us to acknowledge the plurality of values and meanings it entails – and the challenge this poses to the possibility of a pan-disciplinary consensus on ethnographic research ethics. Otherwise we run the very real risk of legislating it out of existence (see E. Simpson, 2011).