Methodological and ethical choices

on positioning

In feminist studies and develoment studies there is a wide recogni-tion of the importance of understanding power hierarchies, both in

the situations one studies and in the research setting. Since my study is essentially about hierarchies, it is perhaps even more important to recognise the power relations both between me as a researcher and those I study, and between the two groups in my focus.

Since Chandra Tapalde Mohantys’ (1984) ground-breaking article

“Under the Western Eyes” which criticised western researchers’ ten-dency to perceive Third World women as a monolithic group of poor victimised women, there has been a rich and sometimes exhausting discussion on how to write about those one studies and about the power relations in the research setting. While Mohanty was later criticised in her turn for portraying Western feminist researchers as a homogenous group, her important reminder of the heterogeneity of Third World women became one of the main elements of what came to be known as ‘post-colonial feminism’ (Mattila and Vuola 2007, 212–213).⁹³ Other main elements were a critique of the assumption that Western and Third World women have totally different prob-lems, and of Western women’s blindness towarsd their own role in the colonialisation of Third World women (ibid).

Simultaneously, another major paradigm within anthropology,

‘reflexive anthropology’ (Scholte 1969), emerged to address similar issues, emphasising the need to acknowledge the researchers’ subjec-tivity, positionality and representation. By the 1990s, the main ele-ments of post-colonial feminism and reflexive critique were incor-porated into mainstream anthropology and gender studies, and to an increasing extent, into development studies, if not so much into development practice. As Nencel (2001, 74) summarises it: “Mak-ing the (research) decisions explicit is one of the responsibilities of practising a reflexive anthropology”. This is what I aim to do in the following sections.

The post-colonial discussion of subjectivity was extended to

93 Another paradigm ‘Third World Feminism’ stems, at least partly, from post-colonial feminism and is sometimes seen as a form of it (Mattila and Vuola 2007, 211–212).

the question: “Who can speak for ‘the other’?” (hooks 1988; Spivak 1988). The post-modernist concern as to whether the researcher is complicit in neo-colonial knowledge production of silencing ‘subal-tern’ voices (Sumner and Tribe 2008, 43) had a strong impact.⁹⁴ So influential has this discussion been that the fear of how to represent

“others” correctly has on occasions paralysed research (Nagar 2002, 180). The immobilising effect of the representation question is not unfamiliar to me either. At the beginning of the process, I often won-dered whether it was legitimate for me to carry out this research when there are many local researchers in India who could do it (see Schey-vens and Storey 2003, 2; ScheySchey-vens et al. 2003a, 155).⁹⁵ But the cure for fear of colonial anthropology is not its replacement by indigenous anthropology (Madan 1982, 16). To overcome this debate on repre-sentation, Saraswati Raju (2002, 173–174) pleads: “We are different but can we talk?” For her, the privileged researcher, whether native or foreign, can still have commonalities with those she studies.

It has been argued that for those involved in gender and women studies paid domestic work was long a sensitive topic since it high-lighted inequalities amongst women (Moors 2003, 387). Romero (2002, 43–44) discusses how her colleagues in US academia reacted when she presented her findings on the Chicana workers perceptions over working conditions. Many of these colleagues employed domes-tic workers themselves, felt uneasy about the results and began to

94 Harding and Norberg (2005, 2010) argue that feminist researchers have, as part of a broader debate over knowledge production, contributed to the episte-mological crisis of the modern West by writing out the complex ethical dilemmas embedded in research, and by challenging conventional methodologies.

95 Johanna Latvala (2006, 64) describes how in the early phases of her PhD research she asked: “Who was I to go to Kenya and study women’s lives there in the first place?” For her, the way to move beyond this question was to carry out more equal and more sensitive research. From this perspective, she found it easier to study well-educated, middle class women, with whom she had more in common than, for example, poor, uneducated women. But not studying poor uneducated women is definitely not the solution to the problem of representation: since there are no uneducated researchers, who would study poor uneducated women if educated women or men would not do it?

defend the employer practices. Given the differences in structuring the state and in organising domestic work in Finland by comparison with the US, Great Britain or India, for example, my experiences are different from those of Romero. Having domestic workers in Finland is not common, and women carry out hours of household work daily, irrespective of their class and educational standing. At the same time, the gender gap in domestic work has diminished, albeit mostly be-cause women do less work than before, rather than men doing more (except for increased participation in child care) (Miettinen 2008).

Historically, however, the situation has been different in Finland:

it used to be common to employ workers and up to the early 20^th century being a maid was among women’s most common occupations (Rahikainen 2006). By the late 20^th century full-time domestic work-ers had gradually disappeared from Finnish homes with, for example, the expansion of women’s labour market participation and develop-ment of public welfare services such as municipal child care (Rahikai-nen 2006, 247–249). Since the 1990s, the employment of domestic workers has again increased, and there are indications of this becom-ing a more common pattern (Rahikainen 2006, 247–249). While it is rare to employ people to do house work on a daily basis, with the exception of child care, it is increasingly common to employ someone for weekly or fortnightly cleaning. I have employed workers for child care and for cleaning. But I have also taken care of children in two private homes in Sweden at the age of sixteen and in France at the age of nineteen, through an organised ‘au pair’ arrangement. During my school years, I did manual jobs such as dish washing in a cafete-ria. These experiences, albeit short-term and student jobs, may have made it a little easier to understand the workers’ perspectives. My ex-perience both as an employer and as a worker is quite different from the employment trajectories of most educated women in India. In addition, despite the shortcomings of being an outsider with limited language skills and cultural experience, coming from a country where domestic work is differently organised gave me a different perspective

than local researchers, which may have improved my understanding of the phenomenon. This legitimises the study.

The post-colonial feminist critique assumed that the subject of the Western researchers’ studies was a poor, unprivileged person hierar-chically much lower than the researcher. This may have been the case in the 1980s, but more recently there has been a significant change in focus so that studies on developing countries, both by foreign and lo-cal researchers, increasingly explore the middle classes and the elite.

Kirin Narayan (1993, 671) argued against an entrenched distinc-tion between ’native’ and ’non-native’ anthropologists and instead proposed that researchers should be seen “in terms of shifting iden-tifications amid a field or interpenetrating communities and power relations”. Others have argued that the notions of insider and out-sider are not a stable binary, and should rather be seen as continuum where positionalities change in time and space. (Mullings 1999, 340;

Scheyvens et al. 2003b, 185). While recognising the changing posi-tionalities I see no reason for a total rejection of the distinctions ‘na-tive and non-na‘na-tive’, or ‘insider and outsider’. In India, I was definitely always regarded first and foremost as an outsider by both employers and workers. What shifted, however, was my class and status relation vis-à-vis different respondents.

Although all recognised me as a white, European researcher, my class status was reminiscent of the employers, given our educa-tional backgrounds and other forms of symbolic and cultural capital (Bourdieu 1984), but it was quite different from that of all the work-ers. In short, with employers I was neither studying down nor study-ing up (Nader 1972; Scheyvens et al. 2003b), but rather sideways (Boyer and Hannerz 2006, 9; Ståhlberg 2006, 58–59). With some respondents, I felt I was studying up,since they were in a similar class position but much older than I was, an important hierarchical aspect in India. Both I and the Indian interpreter were in a closer ‘positional space’ (Mullings 1999, 340) to the employers than to the workers.

In fact, as an educated working mother, there were many

similari-ties between my life and that of some employer respondents.These commonalities, together with my research topic, provided us with a common frame of reference (Ståhlberg, 2006, 50). It was easy for me to relate to housework and child care issues, despite the notable dif-ference that I shared the work with my partner and did not employ anyone for domestic tasks at the time of the field work.⁹⁶ In fact, these different experiences and perspectives often enriched the discussions (see Scheyvens et al. 2003a, 140). Moreover, I suspect that my inter-est in the mundane matter of household chores made me seem less a foreign scholar with a high status and more a woman with similar re-sponsibilities. The employers also implicitly acknowledged that, like them, I would have been in position to employ domestic workers had I lived in India.

However, there were fewer similarities between me and the work-ers: they were uneducated, mostly illiterate, and dependent on their daily labour. Clearly, as an educated woman from a rich country, a person who could travel across the world, move around freely in Jaipur without a male member of the family, and work independ-ently, I was in multiple ways in a more powerful position. Drawing on Maria Mies’ (1993, 68) idea of ‘partial identification’, as a working mother I could partially identify with their joys and concerns, espe-cially those related to the challenges of combining wage work and child care. While recognising what bound me with my respondents and what separated us (ibid), as the lowest common denominator we shared an experience of what it is to be a woman and the experience of being treated as a woman in our respective cultures (see Ronkai-nen 1989, 69).

Although the respondents were interested in my family situation, none asked about my religious background, which they, as Hindus and Sikhs, clearly knew differed from theirs.

96 Ruth Vanita (2004, 69) talks about a syndrome which she calls “Our patriar-chy is better than yours”. I have tried to carefully analyse my own thinking and be aware of such tendency.

Epistemologically, it has been useful for me to think of the re-search through Donna Haraway’s (1991, 188) idea of ‘situated know-ledge’. By using the metaphor of a camera lens through which one sees a highly specific view, “a partial but wonderfully detailed view”, Haraway called for research with a partial perspective and locatable knowledge. She argued that only this kind of knowledge can provide an objective vision, and thus be morally sound. I would be hesitant to claim ‘situated knowledge’ to be more objective than any other type of knowledge, but I have nevertheless aimed to provide a detailed view on domestic work in Jaipur, one city of India. My knowledge derives almost entirely from Jaipur, and I cannot make generalised claims about domestic work in the whole of India. I can, however, reflect my findings against studies on domestic work in other cities of India, through which I can locate my findings in a broader context.⁹⁷

studying hierarchical relations

My study is about hierarchies in society, where power is essentially a relation (Young 1990, 31). When I initially decided to study do-mestic work in India, I had thought of focusing mainly on workers.

However, taking into account the notion that rational knowledge is power-sensitive (King 1987, 192 cited in Haraway 1991) and the need to understand the conflictual nature of the culture one studies (Kumar 2006a, 84), I decided to study both employers and workers, the two sides of the same coin (Scheyvens et al. 2003b, 183). Only by doing so would it be possible to begin to understand the complex realities embedded in the labour relations between two sides, even if this meant spreading the focus and limiting the time I could spend with each of them. Thus, like Dickey (2000a) I chose a ‘relational ap-proach’ and interviewed both employers and workers, since “in order

97 For a study on domestic labour relations in Madurai see Dickey 2000a; in Kanpur see Froystad 2004; in Delhi see Neetha 2003 and Raghuram 1999; in Kolkata see Ray 2000 and Ray and Qayum 2009; and in Chennai see Tolen 2000.

to understand class relations in full, one must give equal weight to the perspective of each side” (ibid, 32). Ray (2000b, 693) made a similar choice when aiming to explore the dialectic of employer and employ-ee gender ideologies.

The employing class in Jaipur perceives a three-tier basic class structure consisting of the poor, the middle class, and the rich, and the ones in my data viewed themselves as different not only from the poor but also from the rich. The workers, instead, divide people into two main classes, bare log and chote log, big and small people.

They perceived themselves as ‘small’ and poor and the employers as

‘big’ which for them entailed all wealthier people from lower middle class to the very rich.⁹⁸ The employers and the workers perceive each other clearly as belonging to an opposing class, and used reciprocal oppositional images of each other (see Dickey 2000a, 37). Workers described their employers as rich, as having big and beautiful hous-es, and as having respected, well-paid jobs. Employers generally de-scribed workers as poor and uneducated, to which some added the adjectives uncivilised and dirty. On a more maternalist, and perhaps moralist note, some female employers generalised female domestic workers as “victims of domestic violence by alcoholic husbands”.

I have aimed to give equal weight to the situations of both while acknowledging the power asymmetries between these groups and as Visweswaran (1994) has suggested, I have aimed to recognise and understand, and make such hierarchies visible in my study, as well as to analyse their internal controversies.

The research process always evokes positive and negative emo-tions (Kleiman & Copp, 1993). However, most researchers doing field work allow themselves to have certain feelings, such as closeness with participants, whilst trying to deny or get rid of emotions they think inappropriate. Field workers are expected to feel for the par-

98 A similar difference between employer and worker class perceptions was found in South India (see Dickey 2000a, 33; Tolen 2000, 67).

ticipants, and when theyfeel disgusted with participants they prob-ably try to transform their inappropriate feelings into “better” ones.

If the researchers fail to have “appropriate” emotions, they may accuse themselves of empathetic incompetence. (ibid, 28–29).

I first went into the field with some negative ideas about the em-ployers, based on the NGO reports and media articles on the exploi-tation of domestic workers in India. As a result it was a particular challenge at first for me to listen impartially to the often contradic-tory voices of the workers and the employers. I tried to be constantly cognisant of my preconceptions so that I would not be tempted to see exploitation where it did not exist, or not merely seek confirmation of my preconceptions, and thus fail to see the contradictory information (see Scheyvens et al. 2003b, 190). As the research progressed, it also became easier to empathise with the employers and their daily strug-gles even if they seemed light compared to those of the workers. On the other hand, whilst I could understand their situation, I did not empathise with those who seemed to be practising outright exploita-tion. It would have been a difficult ethical question had I seen out-right, on-the-spot abuse, such as violence, in the employers’ houses.

Although this did not occur, both employers and workers told me about such incidents.

solidarity through research?

Whatever expectations they might have for these discussions to some-how improve their lives, I will not be able to fulfil them.

This line is an excerpt from my fieldnotes in Jaipur in 2006. The in-terviews had started well, and I had got to know many workers. With the daily visits to the workers communities, thoughts like this often circled in my head on my way back to the comfortable flat we rented.

I agree with those who think good research should aim to bring social progress (Harding and Norberg 2005, 2011) even if Kumar (2006a,

81) rightly notes that since reform is problematic, scholars’ responsi-bility has often been seen to end conveniently before reform is called for. Like many others, I initially took to development studies guided by concern about social justice and inequality, and committed to the generation of useful knowledge that could make a difference. (Sum-mer & Tribe 2008, 31; Scheyvens et al. 2003b, 187).⁹⁹ But difference of what sort? And useful to whom, when and where?

While I see the question of whether I can contribute in any mean-ingful way to the lives of those I study in a more positive way from retrospect, the sentence I wrote in Jaipur still captures some of the painful emotions I felt during the research process.¹⁰⁰

It has been argued that the researchers also influence the phenom-enon they study (Kleinman & Copp, 1993). The research process definitely has left a significant mark on me, but I am extremely wary of over-emphasising my role in the lives of those I studied. The in-terviews were at times very emotional and sometimes potentially em-powering for us all. However, to claim to have influenced their lives would overemphasise the role of a short-term research project.

In the very early phases of this research, I had thought about par-ticipatory action research (Wheeler 2009) but I decided to carry out a rather conventional interview study for several overlapping reasons.

First, the kind of potentially empowering participatory research I had in mind would have to have been a long process for it to be meaningful (Scheyvens et al. 2003b, 187). Having become a mother, with all the delays in the research process this entailed, I had more limited time for the research. Secondly, and as importantly, I felt that I did not want to work directly with local organisations, which would have been necessary for the action research with children I had envis-

99 Sumner and Tribe (2008, 36) talk about a ‘mixed purpose’ of Development Studies on the continuum of purpose ranging from instrumental to theory/ab-straction.

100 See Ruth Behar (1996) for a wonderful analysis of the pains and joys of ‘the vulnerable researcher’ who develops deep emotional ties with the subject of the study and with the research participants.

aged. I did not want to become entangled in the power play of donor funding at this point. I also feared that I would get involved in activi-ties related to child workers which I could not support. Some of the potential ethical problems became evident in discussions with some NGO workers or officials, who were keen to find domestic child workers with my help and punish their employers.

I agree with Kalela (2000), who emphasises that it is the duty of the researcher to do justice to those she or he studies. For Nancy

I agree with Kalela (2000), who emphasises that it is the duty of the researcher to do justice to those she or he studies. For Nancy