Methods and Ethics of Research
The research for this project was conducted over a period of 10 months from January to October in 2012, with a short 10 day follow-up in late November/early December. During this period I was based in Madurai – the temple city in south central Tamil Nadu. The aim of the research was to chart the changes in Dalit politics that had occurred since the late 1990s when the largest Dalit movement in the state decided to contest elections. In some ways, the research was designed as a follow up to my earlier fieldwork in 1998-9 which hinged on an ethnography of the Viduthalai Chiruthaigal Katchi (Liberation Panther Party – VCK) during its transition from movement to political party.
As with the earlier fieldwork, research in 2012 was ethnographic and included interviews, participant observation, group discussions, and conversations in the urban enclaves of Madurai and its satellite villages. The intention here was to capture the diversity of individual and group experiences and chart the localised impact of political change. Having grown up and subsequently conducted research in Madurai, I was able to build on multiple contacts both within and outwith the party who helped to negotiate access to VCK cadre, non-affiliated Dalits, non-Dalit commentators and activists and respondents from alternative organisations and parties. Participant observation helped to contextualise interview findings and provided data on social interaction, caste discrimination and everyday practices. I attended party events and events staged by other Dalit movements in the state, observations from which were written up as fieldnotes. I also made use of the archives housed in the Dalit Resource Centre and Social Analysis Centre in Madurai to collect media and non-governmental organisation reports.
The key methods for research, however, were formal and informal interviews. Formal interviews followed a rough pattern of questions and tended to be recorded. In all, I conducted 62 such interviews. 56 of these interviews were with men and only 6 with women. In large part this reflects the male dominance of the VCK. 27 of the interviews were conducted in rural areas and the rest in urban settings. A further 58 lengthy discussions have been characterised as ‘informal interviews’. In these cases respondents were happy to chat at length but either did not want to be recorded or be formally interviewed, or where the interaction took the form of a group interview. In cases where naturally occurring groups were present, I was happy to engage them in discussion and found that their internal interactions could be extremely informative (see also Gorringe 205). In these instances notes were taken as soon as possible thereafter. Of these, 15 were with women though many more women than this were involved since in seven of these instances I was engaging with a group of village women who were happier to talk to me in a group than one on one. Additionally, I spent time interacting with and chatting to people about what they thought of the party, their experiences of caste discrimination and what (if anything) had changed in the intervening years and jotted down significant data.
All interviews were conducted in Tamil by the researcher and were subsequently partially transcribed/written up and translated by the researcher. Key interviews were transcribed in full, but the time involved in simultaneously translating and transcribing the recorded interviews meant that word for word transcriptions were not completed. In those cases where interviews/discussions were not recorded, notes were taken as soon after the discussion as possible. Whilst I am fluent in conversational Tamil, I resorted to native speakers to help translate idioms appropriately and also to translate the ‘high Tamil’ used by politicians.
Access
To access participants I used a mixture of purposive and snowball sampling. I began by approaching people I had interviewed in the late 1990s or those I knew from previous times in India both because I had already built up a relationship of trust with them and also because I wanted to get a sense of how their views and practices had changed over time. These participants then put me in touch with others and so on. In additional I sought out and interviewed some of the state level leaders – many of whom I had interviewed in the past - of the party in order to get a sense of the party line on key issues and debates. For roughly a third of the interviews I was accompanied by Tamizh Murasu - a party member who was able to make introductions and break the ice. In these cases, Tamizh Murasu often intervened with questions or comments that elicited greater detail than would have been the case had I been alone. Whilst there is a danger the people may have modified their responses due to his presence, responses gained in interviews where he was absent chimed with the others suggesting that the effect – if any – was minimal.
Finally, the research also had a quantitative dimension, but this involved purchasing data-sets from the Centre for the Study of Developing Societies in Delhi. This data pertained to election surveys carried out both in Tamil Nadu and in Uttar Pradesh.
Informed Consent
Securing the informed consent of participants is key to ethical research and is particularly important when the respondents are vulnerable and marginalised, as many of my interviewees were in this research. Like Liamputtong (2008: 12), however, I found that securing written consent was problematic. People were happy to discuss ideas and engage in debate, but were much more wary about putting signatures on a bit of paper. Seeking such consent, I was told, made me seem like a government official or census taker – especially given that the contentious socio-economic caste survey was underway during my fieldwork in 2012. Under these circumstances I decided to secure consent verbally rather than in writing. In each case I explained the purpose of my research and what I would be doing with the data I collected. In this endeavour I was greatly helped by my past research in this area. I was able to point to publications emerging from that project to explain what I was doing and what I was hoping to achieve. Several of my respondents had read some of my work and were able to engage with me on that basis. Despite this, some respondents still hoped for more from me as an outsider. I was generally successful at differentiating myself from authority figures, but often received calls to ‘tell the leader what we are thinking’. Trust and rapport were key here and so, towards the end of my time in India, I compiled a short report of ‘key findings’ that was sent to Thirumavalavan (party leader) and other leading figures.
People were told that they would be anonymised and that they could stop the interview at any point. Whilst anonymity was important to many of the interviewees, some of those interviewed wanted to be named. As I found in 1998/9 (Gorringe 2005), participants who are also activists are keen to leave a trace in history and wanted to have their name in a book at least. There may have been ulterior motives here for some, since one middle ranking leader at least credited an interview I gave to a magazine (which mentioned them) with their promotion within the party. The situation was complicated since several respondents were happy to be quoted on some things but wished to remain anonymous for others. The VCK remains a relatively small party and those voicing criticisms were less happy to put their name to confidences shared during the interview. In one instance, when I interviewed the propaganda secretary of the party, he insisted on vetting my questions beforehand, reviewing the transcript and having the interview published. Given the lengths participants went to accommodate me I felt that this was the least I could do. I have also shared interim findings with participants and commentators through lectures in India, by addressing a meeting of the VCK, and by sending work to party leaders.
I would, in this sense, echo Liamputtong (2008: 15) who argues that in cross-cultural research more emphasis needs to be placed on ‘trust building, reciprocity and rapport than the mechanistic process of securing informed consent’. In adopting this approach I was often subject to cross questioning by participants before they consented to be interviewed and was frequently quizzed about my background, aims and opinions. By engaging with people and entering into open discussions about key issues and debates I was able to gain the trust and insights of participants. This, in turn, places a responsibility on me in terms of ensuring that their views are represented fairly and that all names and details are anonymised where there is any potential for harm (material or social) to the participants as a result of their participation in the research.
Gorringe, H. 2005. Untouchable Citizens. New Delhi: Sage
Liamputtong, P. 2008. ‘Doing Research in a Cross-cultural context: methodological and ethical challenges’, in P. Liamputtong (ed) Doing Cross-Cultural Research: Social Indicators Research Series Volume 34: pp3-20. Springer Netherlands eBook.