Abstract
This article contributes to debates about ethical and methodological dilemmas experienced in international development studies. It departs from a research experience based on a longitudinal study, the Reality Check Approach, that puts intimacy, immersion and consensus at its core. These concepts signify an ethically motivated approach that aims to ‘give voice’ to people living in poverty. They also describe an ideal research relationship assumed as the basis for good quality data. The article examines the difficulties encountered when faced with ambiguous meanings in people’s responses, and shortcomings of the approach. These include the combining of ethical and instrumental motivations in the research framework and ambivalent roles and conflicting ethics, highlighted in the conflictual notions of ‘giving someone space to talk’ and ‘making someone talk’.
I Introduction
This article departs from debates in qualitative research that encourage a more public reflection on research experiences (Knowles, 2006; Mauthner et al., 2002). They argue that it is important to go beyond the methodological rhetoric and recognize the ethical and methodological dilemmas that qualitative research practice generates (Karnieli-Miller et al., 2009). These debates have clear relevance to international development research and practice today. Over the last couple of decades, a string of participatory-based approaches have appeared as a result of a conceptual shift that emphasizes agency as opposed to an assumption that people living in poverty are passive, powerless and lacking in initiative (Lister, 2004). Examples of such approaches include appreciative inquiry (Ludema et al., 2001), most significant change (Davies and Dart, 2005), Action Learning and Planning System (ALPS) by Action Aid International (2006) and a plethora of participatory-based tools (ranking, mapping, etc.) that have come to influence small, local as well as large, international organizations, and academic research alike. New methodological frameworks are continuously launched for use in both academic research and evaluation studies carried out by practitioners. Whereas this tells of a great appetite for innovation, there is also a penchant for moving on from one approach to the next. Although much has been written on the topic of ‘the tyranny of participation’ (for example, Cook and Kothari, 2001 and Hickey and Mohan, 2004), these accounts primarily concern the values underpinning and processes within development interventions, and offer less reflection on research methodology and the experience from a researcher’s point of view.
The conceptual shift in development studies towards seeing people living in poverty as active agents has resulted in methodologies that seek more subjective descriptions to complement objective ones defined by the researcher (Hulme, 2004). This shift emphasizes qualitative approaches and integration with communities as the basis for understanding, and with this follows challenges and dilemmas related to relations in the field. While aimed at being more inclusive, for example, it ‘simultaneously risk[s] excluding’ individuals and communities that are hard to reach and to integrate with (Simpson, 2007: 156). It also raises issues around the position of the researcher: in a development context, researchers often find themselves presented with opportunities and/or expectations to act as agents of change, which may conflict with intentions and skills of the researcher and put ethical guidelines of research to the test (Mellor, 2007). Some of these debates build on long-standing discussions in anthropology and ethnography that address the nature, content and analysis of ethnographic work, and its relation to other methodological approaches (see, for example, Atkinson and Hammersley, 1994). But while there is indeed a wealth of discussions around the value and nature of this type of qualitative work, there are also arguments that while ‘qualitative research is saturated with moral and ethical issues’, there is also a tendency ‘to portray qualitative inquiry as inherently ethical’ (Brinkmann and Kvale, 2005: 157, 162) and thereby overlook ethical and methodological dilemmas provoked by close interactions between the researcher and the researched.
The aim of this article is to contribute to a public debate about methodological and ethical dilemmas experienced in development studies. Based on the author’s experience from a longitudinal study in Bangladesh, called the Reality Check Approach (RCA), the article illustrates and examines ethical dilemmas and methodological insights and shortcomings of the RCA study in Bangladesh. It describes how dilemmas arise when ethical research guidelines come into conflict with realms of ethics that call for care and reciprocity between the researcher and the researched. By examining these dilemmas, we find that the assumptions about the researched underpinning RCA lead to methodological shortcomings. The analysis uses the concept of intimacy as point of departure for a critical examination of the RCA. The concept of intimacy suitably describes what the RCA aspires to create in the field and convey through reports. The concept is also useful for elaborating on three particular sources of tension experienced during the fieldwork: (a) intimacy as ethically motivated and intimacy as tool; (b) how the striving for intimacy makes research ethics clash with other ethical realms, such as friendship and community norms; and (c) intimacy and assumptions about how such research relationships will generate good quality data. The latter refers specifically to RCA being a listening study, and its assumptions about ‘giving voice’. These three areas will be discussed in turn, referring to examples from the field.
II Reality Check Bangladesh
In 2007, the Swedish International Development Cooperation Agency (Sida) initiated a qualitative longitudinal study in Bangladesh aimed at investigating the implementation and impact of interventions to improve services in primary health care and primary education (GRM International, 2011). Sida is part of a donor consortium supporting two nationwide government-lead programmes which were implemented initially over a five year period, but have since been extended. The RCA comes from an identified need for better information about the use and experiences of primary health and education services, and what was happening at community level as a result of these large-scale reforms. Questions that were guiding this initiative were whether user needs correspond with the policies underlying the two programmes and to what extent, knowledge and local interpretations of interventions correspond with their intended purposes.
The RCA is devised to function as a complementary source of information that goes beyond statistics and frameworks for monitoring outputs such as teacher training or the building of new community clinics. The aim is to provide people-centred information and thereby address a gap in knowledge and understanding that comes as a result of a preference for quantitatively based evaluations. The gap is exacerbated by donors and policymakers often being removed from the local realities and people, that is, those for whom policies and interventions are intended.
III The RCA principles
The RCA places user voice at its core to address what are seen as weaknesses in conventional evaluation methods in capturing how people experience and interpret change induced by development interventions, and a lack in understanding how large-scale interventions translate to practice on the ground. The intent is to grasp and convey agency as expressed in opinions and actions by people at community level. The RCA is based on a set of principles intended to guide the fieldwork to ascertain that this agency can be better understood. First, the approach is based on the idea of living with rather than visiting: each year, team members spend four days and four nights with selected host households, and each year, for five consecutive years, the teams return to the same households. Second, the approach is focused on conversations rather than interviews: this is to avoid one-sided extractive forms of engagement, based on the assumption that conversations would have advantages over formal interviews and some participatory-based approaches (focus groups, ranking exercises, etc.) since they are more two way, relaxed and informal. The conversations were guided, to some extent, by scripts based on the rationales of the two health and education programmes, but were not based on conventional interview guides. Third, the approach emphasizes learning rather than finding out: again, this addresses methods that see data as something to be ‘accessed’ or ‘extracted’ from the field. This principle is phrased as a rhetorical device for the researcher, a reminder that the researcher takes a step back, rather than as a method as such. This is closely linked to RCA being framed as a ‘listening study’, which requires the researcher to listen effectively to other people’s views and perspectives. They should refrain from using predefined questions and be wary of implying that certain knowledge and viewpoints are preferred over others (see, for example, Corner et al., 2006). The intent is to prevent polite but limited answers, and encourage people to share their views and interpretations of their worlds. ‘Listening’ is, of course, also linked to the notion of ‘giving voice’ to people living in poverty (Chambers, 2005; Collins et al., 2009), which is in line with the ‘Voices of the Poor’ – a project initiated by the World Bank in the late 1990s (Narayan et al., 1999; World Bank, 2008). Fourth, the approach highlights inclusion as an essential ingredient: this implies a behaviour that is sensitive to gender- and age-based differences that may lead to (self)exclusion, something that is found as a weakness in, for example, focus group discussions, public meetings and other forms of participatory rapid appraisals. Finally, the approach advises that, being set in the context of development intervention, the researchers interact with frontline staff; in this case, from primary health and educational services. Exploring this side of service provision serves as an important point of reference, or triangulation, that can support more probing conversations about programme activities.
The RCA is not an action-based study that aims at instigating change, or advising people about rights, opportunities and new services. Rather than facilitating the two programmes, our intention was to listen to people’s views about them and to observe their effects. We foresaw that by staying with families living in poverty, we may encounter situations that would prompt our direct intervention, for example, requirements to give advice or to give practical support in situations of conflict or perhaps medical emergencies. For these reasons, the approach prescribes a passive, low-key role for the researcher. This was motivated by ethical guidelines used in qualitative research with the intent to ‘protect all groups involved in research’ (Economic and Social Research Council [ESRC], 2012). Taking action in situations of emergencies, for example, could jeopardize the safety of both researcher and researched. Taking on a passive role was also based on an understanding that should we become ‘advisors’, that is, some kind of authoritative voice, we may enter into roles that would preclude people from sharing their own views and talking about their own actions.
The reflections in this article are based on my experience as one of three team leaders, and my interpretations and practice of RCA. The approach is flexible, and the three teams were encouraged to experiment with different information-gathering methods. Apart from informal discussions and participant observations, these methods included drama, drawing pictures with children, participatory-based mini-surveys and ranking and mapping exercises. As a sociologist with experience from ethnographic work, interviews and surveys, I chose to emphasize particular aspects of the RCA based on a wish to bring out what seemed to be distinctive about RCA. These distinctive aspects were to listen, to take time and be patient, which implied waiting for conversations rather than setting up scenarios with the intent to encourage debate. To me, the emphasis on conversations and listening in the approach meant leaving tools that are often used to facilitate what I call ‘staged conversations’ behind: I, therefore, did not use mapping, ranking or participatory surveys.
My team of three and I worked on being low key, a steady but non-disturbing presence. The idea was initially to emphasize our role as listeners, and to see the researched as the ones to take initiatives. The slow-paced, longitudinal study would eventually yield relations and information that we sought for. We envisioned that with a short but still shared history developed through our consecutive visits, our conversations would change in character based on, first, an aim to get to know each other to, later, an appetite to learn and probe into stories together. Within my team, we discussed our method of working and how we experienced our interactions in the field on a daily basis. Our discussions related to the dilemmas, emotions and frustrations that are examined in this article, although they reflect my interpretations alone.
IV Intimacy in qualitative research and in RCA
The concept of intimacy captures what the RCA is striving to achieve. The idea of placing intimacy at the core of qualitative research comes from a feminist tradition, signifying what can be described as an ideal research relationship. It is based on a critique of positivist research that is characterized by power distance, hierarchy and expertise (Duncombe and Jessop, 2002). It comes from a wish to move away from an emphasis on objectivity, gained partly through the researcher taking on a detached role in relation to the researched, which in turn (unintentionally) fosters power distance and hierarchies. A relationship based on intimacy includes qualities of ‘mutual care and friendship as well as revelation of, and respect for, personal vulnerabilities’ (Busier et al., 2010: 165). To achieve such a relationship, qualitative researchers direct attention to both initial rapport and the continuing maintenance of good relationships, and this involves ‘self-examination, sharing, and self-disclosure’ (Dickson-Swift et al., 2006: 856). In RCA, the researchers aim to build rapport and trust with households and communities in which they stay. The fieldwork emphasizes creating space for conversations to take place in order to understand and convey needs and experiences as expressed by people not heard or listened to. Here, ‘space’ is not seen as geographical but as a relationship, an ‘affective space’ (Jensen, 2012; see also Birch and Miller, 2000) where there is trust and empathy and the individuals can express mutual care.
1 Reflections from the field – Building rapport
The initial aim of rapport building in RCA was, as in any such field-based research, to gain trust and reduce distance. The team’s arrival was carefully prepared by visits from one of our team members. He tried to ascertain that the chosen households, neighbours and key individuals in the community had been informed about the intention of our visits, and had had the chance to ask questions and give consent. Our host households were chosen based on poverty level. The definition of ‘poor’ was based, as far as possible, on contextual indicators, including occupation, ownership of productive and household assets and type and structure of housing. 1 We tried to ensure that no special arrangements were made to accommodate us and that we did not disrupt normal activities. Initially, we ‘stripped off’ visual attributes that could get in the way of open and non-judgemental conversations, and ‘put on’ features appropriate for the position we aspired to take on, that is, as members of the community instead of outsiders, researchers. We changed urban and Western clothes for a more modest dress, adhering to Muslim dress codes. We left behind research tools such as pen and paper, recorder and list of questions (although we later used notebooks). Sharing time and space, day and night, paved the way for trust: we exposed ourselves for others to see us as normal people with shared needs, rather than outsiders with unknown habits. Time turned out to be an essential ingredient: time to be with families throughout the day and time to revisit families. The longitudinal aspect of the study played to our advantage, an experience that echoes other longitudinal work (Thompson and Holland, 2003). During our fifth year, for example, we were told by some that ‘I actually lied to you last time we spoke’. Our revisits incurred trust that we had a genuine interest in understanding practice on the ground and that talking to us would not result in negative repercussions (Lewis, 2012).
2 Trust and intimacy in the field
While trust was gained at a community level, closeness and intimacy was established with particular individuals and families. We immersed ourselves in family life, sharing space, meals and time with all family members. As the only non-Bangla-speaking team member, 2 I often took on a quiet role verbally, and used body language to engage in conversations and to invite people to share space. Often, we gathered in small rooms, with little other furniture than a bed where we would sit, adults and children together, talking and playing. The first signs of success in gaining intimacy were when the families called us ‘sister’ and ‘brother’ instead of ‘madam’ or ‘sir’. We also came to understand how intimacy grew not only, or even primarily, through talking, but through understanding mundane things related to, for example, eating. While discussions of likes and dislikes in relation to food were endless, we felt intimacy and closeness tacitly expressed through knowledge of how we take our tea, how much salt we use with our food or us saving particular food items at dinnertime for the children to enjoy. Silent nods of understanding and care were shared during mealtimes, in the morning and at bedtime, that is, at hours when researchers would not normally be present.
In descriptions of research that aim at intimacy, the process of building and maintaining rapport is explained as ethically, and perhaps even politically, motivated: reducing power distance is important in itself. Rapport building and intimacy are also described as a means to an end, although often this comes across in language rather than in explicit statements: researchers use self-disclosure to build trust that, in turn, will allow access to personal stories (Dickson-Swift et al., 2006: 855). The rationale behind the RCA follows a similar pattern of combining an ethical and instrumental motivation for an intimacy-based approach. The RCA addresses problems of an ethical character that have political and policy implications, such as exclusion caused by power distance, by focusing on immersion, learning and listening. At the same time, this is also seen as the basis for a ‘relaxed and trusted context for conversations that can lead to enhanced understanding’. 3
In qualitative research, intimacy is promoted as a unique way of learning, and closeness in research relationships is essential to gain insight into an understanding of our world and those around us (Busier et al., 2010). While on the one hand, intimate research relationships are described as egalitarian, authentic, characterized by honesty (Duncombe and Jessop, 2002; Hewitt, 2007), it is difficult to get away from the notion that power remains with the researcher. Glesne (1998: 45) argues that ‘rapport is a trust-building mechanism that primarily serves the interests of the researcher’ (emphasis added) as it is used as a means ‘to attain ends shaped by their own needs’ (Glesne, 1998: 46). Furthermore, our view on intimacy as an end goal or a means to an end may change during the course of the research. Relationships are not static and any researcher can recognize that the researched has ‘options for countercontrol’ (Brinkmann and Kvale, 2005: 165), which in turn encourages the researcher to review strategies of social skills to maintain relationships and preserve an ‘affective space’ in order to gain or maintain a level of control. So, while the ‘building of rapport’ and ‘intimacy’ are primarily interpreted as ethically informed approaches, they have become commodified and professionalized, and are now part of any sociologist’s toolkit (Duncombe and Jessop, 2002). This dual meaning/function of intimacy in research relationships opens up tensions: what is initially presented as an ethically motivated approach can suddenly appear as unethical practice, which is captured in a string of words used by Duncombe and Jessop (2002) when describing the ‘ethics of faking friendship’ as phoney, inauthentic and insincere. The ambiguous meaning and use of rapport and closeness can be experienced with great discomfort by the researcher (Pitts and Miller-Day, 2007). Although engaging in intimacy and closeness can be seen as a process that takes the researcher to a place where we are out of control and power, it can also be ‘interpreted as a mask for some type of manipulation or exploitation carried out to obtain data needed for the study’ (Karnieli-Miller et al., 2009: 283).
Earlier, I described the position of the researcher in RCA as passive and low key. This was said in a context of comparison with action research and facilitators of participatory rapid assessments. The notion of ‘passive and low key’ warrants some further definition: it does not imply a researcher without focus, intent or any less effort than other types of fieldwork. Intimacy is related to physical presence (immersion) and to a mindset (to learn and listen), but the meaning of this had to be translated to the people with whom we were trying to be close with. The initial visits were filled with rituals familiar to many other situations when trying to get to know each other: we explored family histories, recent events in Bangladesh, geographical origin and so on. As our research progressed, over time, we developed toolkits suitable for our purpose: we used social skills and our increased situational knowledge of the host families to both disclose who we are, as individuals rather than as researchers, and to fit in, by showing we know them. This did cause discomfort and increasing frustration as I experienced that the balance between the need to ‘make them talk’ with the idea of ‘giving them space to talk’ could not be easily resolved with the research tools I had allowed myself. This will be further explored later in the article, but first we will turn to the theme of conflicting ethical realms.
V Intimacy and competing realms of ethics
As described earlier, intimacy has a dual meaning in research. Furthermore, intimacy also brings the researcher and participants to positions where roles are not well defined. ‘The rapport building process can require a merging of boundaries between researcher and participant,’ writes Dickson-Swift et al. (2006: 856). This may be desirable and seen as ‘benign’ but can also expose and exploit the researched with unforeseen results (Dickson-Swift et al., 2006). The blurring of boundaries may not be intentional but is a result of longitudinal work and is the ‘unintended consequences of growing emotional intimacy’ (Hewitt, 2007: 1152) between the researcher and the researched over time (see also Kleinman and Copp, 1993; Watson 2012).
There are two issues to consider here: first, the role of emotions in qualitative research; and second, implications of the blurring of boundaries. The two issues are closely related. As discussed earlier, the aspiration to create intimacy in research means bringing care, empathy and vulnerability into the research relationship. The blurring of boundaries between a research relationship and friendship leads all involved to new emotional territories. 4 For the researcher, the idea may have been to create a controlled affective space where important research material can be gained. While emotions may be placed at the core of producing and constructing knowledge, there are also requirements that we should remain distanced from emotions (Watson, 2012). The period spent reflecting and analyzing material provides us with an opportunity to distance ourselves from the emotions in the field and this is important for credibility of our research in the eyes of an academic audience and policymakers (Harrington, 2002; Watson, 2012). Emotional distance can also, theoretically, be achieved in the field by adapting a ‘detached concern’ approach and applying ‘the right amount of empathy’: enough to bond and create trust, but not so much that we get carried away by empathy and lose sight of our research focus (Kleinman and Copp, 1993). However, while working with intimacy as a core idea in qualitative research, the field may present us with a reality where emotions cannot be controlled. Watson (2012) describes how while doing ethnographic research, she became overwhelmed by emotions which, in her mind, contradicted the way she intended and was expected to behave, that is, she ‘crashed the boat and wept’. It was unprofessional, she felt, but eventually made her consider the role of emotional attachment in how she came to understand the community she had immersed herself in as a researcher, and their environment. Knowles (2006) describes how ‘emotional baggage’ caused her to exclude certain research subjects. An emotional event made her confront her own prejudices, and they opened up surprising and important research results. Emotions can cause problems as well as be revealing and helpful in the striving to understand the research subjects.
The blurring of boundaries may come as a result of growing closeness and shared emotions, and it may also cause emotions and stress (Dickson-Swift et al., 2006). ‘Concerns over roles, maintaining professional distance and being detached but concerned all add additional stress to research work,’ writes Dickson-Swift et al. (2006) In a study they conducted on boundary crossing in qualitative research carried out by doctors, counsellors and the like, researchers reported how emotional exhaustion and concerns over their roles caused them stress both during and after the research was carried out. There are obvious similarities of ‘rapport-building behaviour to friendship-developing behaviour’ (Glesne, 1989: 45) which, to some extent, prevents the researcher from controlling emotions and relations. Glesne (1989) contends that we must avoid friendship in research situations, and others warn against being unprepared for what research-based friendship may bring in terms of dilemmas (Brinkmann and Kvale, 2005; Dickson-Swift et al., 2006).
Although equipped with the RCA principles and aims, and with a clear understanding of ethical research guidelines that prescribe informed consent, respectful relations and the protection of individuals and organizations from any threat to their well-being (ESRC, 2012), the evolving character of our research presented us with real dilemmas related to emotions, roles and relations in the field. The dilemmas were caused by sudden events and by emerging emotional attachment to individuals. In the two following scenarios, I describe how, based on emotional attachment and blurred boundaries between research process and friendship, the principles guiding our research came to conflict with expectations and roles as we entered positions as friends and members of the community.
1 Reflections from the field – Intimacy, emotions and ethics of care
During the first year, we made friends with a young mother, only 16 years old, who had lost her husband and had been expelled with her then two month old son from her in-laws’ house. Since then, she had been residing with relatives who shared house with our host family. We spent a lot of time together and she often followed us around, introducing us to people and places. On our third visit, her situation had changed: when her relatives were in dire straits, she was forced to move out and find a job. She explained how she had sold all her belongings, including clothes given to her as a gift during the recent Ramadan. Mostly, she shied away from us and chose to talk to us only when no one else was present. This was different from our previous interactions with her. Now she was in rags, looking tired and unwell, and explained how she was under great pressure to pay back a debt to a local snack-shop owner. Things were not looking good: she was vulnerable and exposed to some unpleasant forces in the community.
On our departure after our third year’s stay, she stayed in the background of well-wishers waving goodbye. I was suddenly overwhelmed by emotions: looking at the young woman with her son, in distress, and recognizing that I know her and care about her, should I not also care for her? Would I be able to come back the following year as an observer interested to find out about her destiny? Or should I take on my role as a friend, a role that I had instigated? Was it not my responsibility, as a caring friend, to support her? I was struggling to cope with strong emotions caused by a situation I had not foreseen. In the crowded space, it was difficult to smuggle a rolled-up stack of notes to her without attracting attention, but I also considered the consequences of doing so. As tears began to cover my face, the teenage son of our household took my hand and whisked me away from the crowd. I left without acting on my emotions.
I hesitated to extend my relationship from empathic to actively caring. Although it would require what for me would be a very small sum to release her from her dire situation, this would mean going against an important principle of research ethics: financially supporting a person who is also being ‘researched’ could mean the motivation for people’s interest in us would change radically and it would undermine our aim to overcome prejudices against us as potential donors. Also, our providing resources would highlight power distance and may put the receiver of money in a position of indebtedness. It would, in other words, be inappropriate ethically as well as instrumentally. This event took only 10 minutes but the emotional reaction that came from my concern for the young mother, and the unease over my lack of support to her, stayed with me. It was a relief when the following year, we saw our young mother in better shape, with support from relatives and a job to pay for her expenses. But should my decision not to support her be judged on the fact that, in the end, she did well all by herself?
The emphasis on ‘living with rather than visiting’ in the RCA principles takes the researcher to positions that bring new expectations and obligations. This resulted in ethical dilemmas where we, as researchers, felt ourselves caught between research ethics as expressed in the RCA guidelines, that prescribed a passive role in situations of need and emergency, and norms related to other spheres of life. The care expressed in general research ethics – to protect all those involved in research – takes on an elusive meaning and does not provide much by way of guidance when intimacy brings the reality of the researched close to the researcher. What would ‘protect’ imply in the situation just described, where norms of friendship would prescribe action that may contradict our principles of not getting actively involved? On several occasions, our team felt we were presented with opportunities to practice norms related to patronage. The norms of patronage build on power relations and they prescribe actions that confirm power and status. But following these norms would not, we reasoned, result in a power distance but rather in closeness based on recognition of how support and care are expressed in the community. The following vignette will illustrate this further.
2 Experience from the field – Norms and expectations of patronage
We arrived to our peri-urban site for our second year’s fieldwork just after Ramadan. Shortly after our first stay here, the area had been struck by a major cyclone. Several of our host families had suffered severely as a consequence. In one case, the house, already in very poor condition, was damaged, along with furniture, courtyard, fishing net and boat. Although 10 months had passed since the damage was done, the family had not been able to recover. My colleague was torn: as an accepted member of this Muslim community, and being a Muslim himself, he felt obliged to help the family to recover. This was also his inner feeling: he had the means to support this family of five that was now suffering on a daily basis due to poor housing and lack of means (boat and net) to earn an income. Through simple means, he could provide support to the family, helping them to take a definitive step out of a disastrous situation. Again, we were faced with different normative fields, prescribing different sets of behaviour, all with reference to what would be ethically right. As a friend of this family, my colleague wanted to help them; and as a community member of considerable status due to his knowledge and affluence, he was expected to do so. Expectations were not verbally expressed but my colleague felt that by ignoring them, community members saw him more as an outsider rather than an insider. The situation was emotionally stressful for my colleague. We discussed it on a daily basis, making plans for how, after the research had been completed, we could continue a more private relationship that would allow us to practice care and show empathy in a more concrete way.
Apart from the tearful episode in the first scenario described earlier, I managed to maintain a ‘detached concern’ position, adhering to our ethically based principles of not getting involved through monetary or any other such support. While I saw my ability to largely remain detached in action, if not emotionally, as a strength, I also felt it disturbing. I considered whether moving from a principle-based approach, that suggested I adhere to our principles, to a care-based approach (Hewitt, 2007), suggesting I followed my emotions, had solved the ethical dilemma? Supporting the young mother could have compromised our relationship: it could have emphasized the power distance and caused uncomfortable feelings of shame and indebtedness towards me, which presumably would not have benefited a friendship-type relationship. It may also have compromised our position in the community at large: there is no end to people in need and one of our priorities was to avoid an authoritative, need-based relationship. A care-based approach to practising ethics may resolve an urgent emotional need, but such an action does not take long-term relations and unintended consequences into account. The dilemma confronted in the second scenario relates to how research ethics and RCA principles come into conflict with community norms of support and reciprocity. This clash reveals cultural and political perspectives of the RCA principles that we had not considered. The principles are based on the idea that by reducing power distance, we would be able to establish closeness. This principle was also grounded in the aim to provide room for the often excluded and marginalized to make their voices heard. Although we did not act on the norms prescribed by patronage, we had to imagine and recognize how closeness, and thereby access to people’s voice, may be established through the practising of norms based on hierarchy, status and power.
Dilemmas per se do not offer satisfactory solutions but trade-offs, frustrations and mistakes. Qualitative research, writes de Laine (2002: 2), is ‘a compromising experience that must be lived through and lived with’. But while some of the dilemmas must simply be accepted as unresolved, this should not prevent us from exploring how and why they happen (see, for example, Donnelly, 2007). Also, examining the dilemmas in detail can reveal methodological assumptions and flaws that can be rectified. It is to this topic we turn next.
VI Intimacy, distance and voice
The third and last point to address in relation to ethics and intimacy in the RCA concerns assumptions regarding the quality and type of material this approach is expected to yield. I will particularly pay attention to the idea of having conversations, and RCA as a listening study here. This was formulated with an understanding that structural interviews undertaken in formal settings would not be suitable. Our intention was to address distance and lack of empathy that can come as result of such interviews (Brinkmann and Kvale, 2005; Mellor, 2007; Mizen and Ofosu-Kusi, 2007; Simpson, 2007) and instead create space for voices, and affective space – intimacy – in order to develop an understanding based on shared experiences and empathy. The reflections here concern qualitative research and assumptions about data quality, highlighted through the concepts of consensus/conflict, unobtrusive/obtrusive, intimacy/distance. I also consider how ‘giving voice’ requires a range of different audiences.
Fieldwork is an intervention that cannot go entirely unnoticed. However, the nature of relations in the field has often not warranted much discussion in analyses of qualitative material other than through a reference to informed consent. In research on sensitive topics, and longitudinal research, we may be offered somewhat more detailed descriptions of what occurred in the interaction between researcher and the researched (see, for example, Birch and Miller, 2000 and Knowles, 2006). Overall, the in-depth interview is assumed to be a consensual dialogue of co-construction, and this consensus-driven approach is seen as key to ‘yielding good quality data’ (Knowles, 2006: 394). It is through relationships signified by consensual conversations and intimacy that we can gain ‘richness of the research data’ (Karnieli-Miller et al., 2009: 280). We are furthermore presented with idealized images of research relations as consensual and part of collaborative projects of knowledge production through languages that emphasize ‘genuine rapport, honesty, and emotional closeness’ (Hewitt, 2007: 1155). Knowles (2006: 394) argues that the fixation on consensus and intimacies in the field has led to a naivety and neglect of the role of conflict in research: ‘sociologists have trapped themselves inside a “research imagination” that fails to present the actors concerned in all their shades of color and circumstances’ (see also Crow, this volume). Knowles draws on her own experience of unintentionally facing a conflictual situation during her fieldwork, and she argues that conflicts can reveal prejudices and assumptions held by the researcher and lead to important research findings.
Similar to the consensus/conflict dichotomy, Harrington (2002) raises issues around the ideal image of the researcher as unobtrusive as opposed to obtrusive. The emphasis on an unobtrusive position stems from an ethnographic approach that sees the researcher immersed with the field and where a subsequent account and analysis from the fieldwork will gain credibility from such an immersion. Harrington (2002: 50), however, provides examples of how credibility and insight can been gained at the point of ‘subverting methodological rules about avoiding obtrusive research techniques’. She describes how being provocative can serve important purposes: it can be used as a method to gain insights and can be seen as ‘constructive engagement’ or ‘critical imperative’ that creates opportunities for better understanding of people’s views. This can occur either in individual conversations or in group settings, where a provocative comment can be followed by debates between the group participants that would normally not have taken place. Harrington (2002: 61) concludes that ‘if researchers accept that most participant-observers are obtrusive to some degree, unavoidably altering the data by their very presence, it might be constructive to consider how to maximize the potential inherent in the situation’. In revising the relationship between consensus/conflict, unobtrusive/obtrusive, we can imagine a continuum of positions that the researcher can take during the course of a research project, ranging from ‘fly-on-the-wall’ participant observer to being actively engaged in change, from working towards conforming and blending in to resisting to accept certain aspects of or events in a research relationship. One position does not need to exclude the other. The following vignette illustrates how, in RCA, I came to reconsider my role as passive and conforming.
1 Reflections from the field – Reinventing ‘distance’
Alia, the wife of our urban host household, had presented us with a life trajectory filled with drama and events that stood out from most others. Her daughter eloped for a love marriage, and once back in the community, a bitter feud erupted between Alia’s family and her son-in-law’s caused by arguments about dowry and the mother-in-law physically assaulting her daughter-in-law. Later on, Alia’s son got heavily engaged with politics and drugs, resulting in him losing his business, causing financial loss and great distress as the family tried to find a remedy for his addiction (he later recovered). As each year passed, we saw how these events took their toll on Alia’s well-being. During our fourth and fifth year, our intimate relationship with Alia was mainly evident by the fact that she did not feel compelled to look after us as guests, but got on with her very busy life while we could sit in her home and exchange a few sentences with her now and then as she passed us between her tiny kitchen and the little tea-stall they were running at the front of the house. Although her very relaxed and inviting manners made our stay emotionally comfortable – we felt safe, secure and privileged to enjoy her company – it did not really provide us with much by way of conversations anymore. In fact, we were ready to listen, but she had not much to say, or so it seemed. In comparison to our encounters with service providers and users that we met at health clinics and schools in the area, our conversation-based approach took on a particular character with Alia. In other settings, we accepted that those who spoke where forthcoming and confident speakers. This made it more possible to have a slightly more confrontational approach to discussing issues around what and how services were provided. In our relationship with Alia, which was similar to other relations with our host households, the combination of us cherishing our friendship and knowing she was vulnerable, physically as well as emotionally, made it more difficult to use an approach that involved presenting her with contradicting and confrontational arguments in order to create an engaged conversation. Frustrated by a situation that provided little by way of insightful conversations, I decided to change tactics. I explored how we could use ‘distance’ through the establishment of new roles to enhance our conversation, which meant relaxing our aim to erase boundaries. Although a guest in her home, I arranged a formal meeting with her. My research colleague and I turned up for our meeting, with pen and paper in hand, and with a large sheet of paper with her recent life history illustrated and a grid spanning the five years we had been there, with seven predefined topics to discuss. We were well aware that this may present Alia with a challenging situation: we were to present her with her life, as we saw it. This tactic meant that we were somewhat going against the principles of the RCA.
We entered a new territory with our relationship, one we had not explored before. Alia smiled hesitantly, sat down and looked slightly embarrassed, and so were we. A detailed conversation (or was it interview?) followed around the seven topics. A clear distance was established between us through the large sheet of paper. We asked questions and interrupted her in her storytelling. Our questions showed a level of ignorance, but through our probing, we were hoping to show knowledge and awareness that would provide encouragement for Alia to give us detailed accounts of the ins and outs of her life over the past five years. As we talked, a less eventful trajectory began to unravel than the one we had pictured initially, a story that ran parallel to the drama we had been witnessing and that included Alia’s very own perspectives on well-being and what makes for a good life. This meeting resulted in new and very important insights in terms of how we understood Alia. It also led to a new kind of closeness in our relationship, which illustrated to me that closeness and intimacy is gained through a variety of routes. Creating distance through repositioning myself as a researcher, with pen and paper in hand, questions ready, did not only present me with a role in which I felt more confident but also Alia with a position of clear purpose, a worthy audience. She engaged fully with this new setting.
The naive assumptions made in RCA are similar to those described in research settings aspiring to instigate change through action research (David, 2002; Greene, 1988; Simpson, 2007): there is no guarantee that the people an action research project has been tailored for are interested in participating. In RCA, I found that people were not always interested in talking, and this presented us with a problem, particularly if it was the same individuals who continuously rejected the opportunity to give voice. Reflecting on the meaning of ‘speech’ and assumptions about ‘voice’, Jackson (2012) argues that, in development studies, empowerment and participation initiatives have been preoccupied with the practice of ‘giving voice’ but have not paid much attention to the meaning of voice and ‘the “how” of speech’ or the role of listeners (ibid.: 1). Jackson (2012: 14) argues that ‘talk always needs triangulation with observation and other evidence’. In our work, ‘listening’ meant hearing what was said, but we extended the meaning to include experiences and observations. In this sense, we practised an understanding of ‘listening’ as based on triangulation.
Furthermore, Jackson (2012: 1) argues that ‘social inequalities are widely assumed to be based on speech deficits’ and consequently, ‘the social inclusion agenda is implemented through building voice through talk and testimony’. Here, a number of aspects of the RCA come into play. ‘Listening’ was primarily linked to providing space for voices, without considering the role of an audience. The audience, that is, the researcher, came, in my case, to take on a passive and low-key role. This was based on striving to achieve a relaxed and non-threatening atmosphere, and that we steered away from taking on roles as authoritative or advisors. The implications of such a position are described by Glesne (1989: 52) who warns that in research focused on attaining a friendship-like relationship, the researcher may end up ‘censoring questions in order to avoid alienation’. Being scared of interrupting a smooth relationship may lead the researcher to ‘assume a nonreactive presence’ (Glesne, 1989: 52). In this context, taking on an unobtrusive position, and practising ‘friendship’ in fieldwork over a period of time, presented us with frustrations. This reveals some naivety in the RCA: the focus on listening, coupled with a researcher taking an unobtrusive position and being a ‘nonreactive presence’, meant we failed to recognize that as listeners, we must be prepared to take on active roles and that the framework for the speaker relies on an audience that is interactive. The given story reveals that it is through interchangeably using intimacy and distance that revealing conversations can be encouraged. This is perhaps particularly important as we aimed at going beyond the outstanding and dramatic events that may be easy to outline, to understanding their implications and relations to longer trajectories.
VII Concluding remarks
The RCA is now being replicated around the world, and great interest has been shown in RCA by actors on the scene of international development. It is hence both timely and important to reflect on the approach and in so doing, contribute to ongoing debates about ethical and methodological dilemmas in the field. It is important to make a distinction between dilemmas and shortcomings: while the former highlights a need for reflection, the latter refer to critique of the approach. The dilemmas examined here refer to how the principles are ethically and instrumentally motivated, and can briefly be illustrated by the difficulties in finding a balance between the idea of ‘giving people space to talk’ with the need to ‘make people talk’. The dilemmas also concern ethics related to different types of relationships, where a balance between principle-based action and action based on emotions and care can come into conflict. There are no recipes for resolving such dilemmas: they are inherent in qualitative field research. But by examining these dilemmas in detail, the analysis shows that they are intimately linked to assumptions underpinning methodology, tacit views of what constitutes good research relationships and how this is linked with good quality data. The analysis suggests that we reconsider concepts such as intimacy/distance, consensus/conflict and unobtrusive/obtrusive as dynamic complementary pairs rather than dichotomies. This is well illustrated through the case of RCA being framed as a listening study: in order to facilitate ‘giving voice’, we must pay attention not only to space (a relaxed and non-threatening atmosphere) but also to audience. The role of the researcher should be allowed many characters: intimate as well as distant, unobtrusive as well as obtrusive.
The future success of RCA and similar approaches cannot rely on rhetorical presentations that hold them as ethically right, framed as responses to what is seen as previously flawed evaluation techniques, policymakers with a preference for quantifiable data or research that favours objective reporting as opposed to analysis based on empathy and immersion. This implicit comparison inadvertently leads to dilemmas and methodological ambiguities being brushed aside and as a consequence, the meaning and value of the findings cannot be fully understood. Here, it is also important to consider the impact of general ethical guidelines that perhaps unintentionally show a preference for consensus and conflict avoidance. The analysis of RCA shows that this kind of approach calls for an understanding of the complex nature of research relationships, which requires that we allow for a variety of positions, including distance, conflict based and the researcher as obtrusive.