The methods adopted were also made possible by being based within an NGO. The integration of the COPE research with the research needs of existing ACORD programme activities provided a degree of security in that it allowed us to work with very diverse groups of people with whom the organisation already had a long relationship, to whom it had demonstrated commitment, and through whom it had created a constituency. The active involvement of long-standing ACORD staff in these activities added legitimacy, and gave access which would not have been possible if we had been perceived as doing ‘research for research's sake’. Undoubtedly this way of working also influenced the methods adopted such that they diverged from a more purely academic model of research. In particular the need to identify and implement research activities in response to the agency's emerging programme meant that research instruments, rather than being pre-determined and piloted in advance, were developed on the spot.
On the other hand the experience highlighted some of the shortcomings of the agency when it came to dealing with a complex research process. While they provided support to the implementation of the research as outlined, they were less able to respond to the processes which the research triggered. In particular, the momentum and political support developed by the September 1999 conference was not capitalised on. More generally there was no institutional capacity to deal with key issues which such research can uncover; there was no specific budget for legal support, no staff to provide psycho-social support, and no capacity to provide the kind of national and international witness which both respondents and our findings demanded.
Evidently there were also certain limitations to the methods adopted. One of these has already been mentioned, namely the lack of women fieldworkers in the protected villages. The already considerable diversity within the types of information given by the male fieldworkers would undoubtedly have increased had women been represented as well. Certainly, were I to carry out another such project, I would insist on having a gender-balanced team. For while women's voices were heard in many of the reports, and women and men were were equally represented in all the focus group discussions, women's subjectivity was not brought to bear on field-work in the protected villages. Their involvement would have added an important further dimension to the findings.
A further limitation was that, while I developed an ear for quite a lot of Acholi, I did not learn to speak it properly and this meant that some of my discussions were mediated by the need for translation. Similarly, while all the fieldworkers wrote their reports in English, this was considerably more difficult for some than for others, and undoubtedly some details and nuances were lost. I felt, however, that because we were using several different methods of data collection, the short-comings in language in one area would be made up for in another. When it came to translating the transcript of the 1994 peace-talks video, for example, the final version was the result of several days’ discussion involving the translator, a number of colleagues and myself. As we compared the translated transcript with the video almost sentence by sentence I became very aware of some of the ambiguities of language used, and also of the particular force with which language was used in that meeting.
From a personal point of view the research process had a high psychological impact; although at the time of collecting the data I felt well able to process it, I subsequently found, for more than a year after finishing the main period of field-work, that it was difficult if not impossible for me to work on the data. Even now, some years later, some of the data still has the power to disturb me. I was somewhat reassured to find that I was not alone in this, even though I had been given no warning it might happen. With the benefit of hindsight I feel that the standard model of PhD research, in which a year of field-work is followed immediately by a year of write-up, does not allow sufficient time for the individual to process the experiences and information that conflict zones provide.
Subjectivity and Objectivity
The question of subjectivity and objectivity is a vexed one. It could, for example, be argued that, because the nature of the information collected changed over the course of the field-work, it does not allow an objective analysis as the data collected at the end was not directly comparable to that collected at the beginning. There are several responses to this. First, certain questions were repeated throughout the fieldwork in the monthly questionnaire filled out by the fieldworkers. As such, some basic forms of data were collected in a consistent fashion throughout the research. Secondly, what would an objective account entail? It is clear when conducting field-work that nobody, whether ‘insider’ or ‘outsider’, has the only description, the whole picture or the only answer to a situation. The extent to which each of us saw the situation through very different eyes was perhaps most evident in the fieldworkers’ photographs; the majority of photographs of one fieldworker, for example, were of group activities and official events, with most pictures featuring dozens of people against a wide backdrop. Those of another fieldworker were far more intimate in character, placing one or two individuals or objects right in the centre of the frame. This very diversity of viewpoints was proof – if proof were needed – of the dangers of assuming any homogeneity of perspective on many issues. The contradictions between different positions (insider, outsider, local, national, regional, international, diaspora, refugee, youth, women etc.) become a basis for dialogue within the team and with those beyond it. It is necessary on the one hand to be able to pin down some of ‘what happened when’, while still keeping multiple interpretations of these events in consideration.
A further question is whether quantity of data on a narrow question is better than quality of data on a diversity of issues. My own view is that, leaving co-researchers to determine the level of risk they were prepared to take was methodologically powerful in that it lead to findings on issues which could not have been pre-determined, such as suicide.
As fundamentally, the changing nature of the qualitative data over time was to me an indicator of the success rather than the failure of the methods, for I believed that recognition of the subjective and its influence was essential to an objective understanding. When, as time went by, the fieldworkers allowed their subjective voices to become more visible, it was for me a case of objectivity through subjectivity.8
In particular it alerted me to the complex interplay between trust, time, memory and disclosure; fundamentally research on sensitive issues requires relationships of trust between respondent and researcher over time. The fact that issues could emerge which would never have done so using a more rigid data collection strategy, must raise a considerable question-mark over the objectivity of any data collected without taking the time to build relations of trust and without the engagement of the researchers.
Notes
1. The Monitor, 16 May 1998, ‘Let's Vote on Kony War’.
2. Justice & Peace News, August 2002, Vol. 2 – No. 5: p4.
3. IRIN, 28 January 2004, The 18-Year Old War that Refuses To Go Away.
4. Our starting point was for research staff to contact people known to them to have some family members living abroad.
5. ACORD, ACF, Amnesty International, AVSI, CRS, Christian Aid, Conciliation Resources, CPAR, DENIVA, IRC, Interpares, Life and Peace Institute, Mennonite Central Committee, NRC, Oxfam, Redd Barnet, SNV, Stromme Foundation, TPSO, ARLPI, ISIS, JYAK, Legal Aid Project, EPRC, Northern Uganda Media Forum, Peoples Voices for Peace, Gulu Development Association, Gulu Youth Peace Forum, GUSCO, Hunger Alert, Justice and Peace