13
Area (1999) 3 1.4, 3 13-325 The final fifth: autobiography, reflexivity and interpretation in cross-cultural research Chasca Twyman, Jean Morrison and Deborah Sporton Department of Geography, University of Sheffield, Winter Street, Sheffield S10 2TN. Emails: [email protected] and [email protected] Revised manuscript received 22 January 1999. Summary This paper addresses concerns about the complexities of cross-cultural field- work, and the importance of the positionality of researchers and translators in the research process. These concerns build upon debates within the social sciences about auto- biography, reflexivity and the research process, as well as notions of validity, reliability and ‘truth‘. The paper re-examines these debates in the context of cross-cultural research by focusing on praxis-actual experiences in the field. Interpretation is four fifths of the truth, but not enough. (McFarlane 1996, 15) Introduction This article has evolved from some of the common concerns and shared experiences of two teams conducting society-environment research projects in Botswana between 1994 and 1997. Both projects aimed to provide a greater understanding of societal-environment interactions under the simul- taneous impacts of climatic variability and structural land-use changes. The main project focused on changes arising from the expansion of commercial pastoralism, while the doctoral research examined the recent changes in community development and wildlife management policies. The article draws, in particular, on the fieldwork experiences of two researchers working independently on the projects in Botswana. Their experiences are used to examine the issues of autobiography, reflexivity and interpretation in the cross-cultural research process. The discussion of methodological issues presented here is informed by wider debates within the social sciences, which critique the bases upon which knowledge is (re)produced and articulated in a cross- cultural context. Central to these debates is the postmodern rejection of grand narratives and total- izing theories predicated, in the context of this research, on Western modernist assumptions about the environment and local knowledge (Gandy 1996). Leach and Mearns (1996) argue that our under- standing of environmental issues such as ‘desertifi- cation’, overgrazing and wildlife management has been informed by orthodoxies founded upon ’received scientific wisdom’, which point to the destructiveness of local practices, casting local actors as both agents and victims of what are widely perceived as environmental catastrophes. Accord- ing to this view, the physical environment and society are represented in essentialist terms as sep- arate analytical categories. Orthodoxies are then re-enforced as they become embedded in insti- tutional structures and hence inform power relations and policy at all levels (Foucault 1981). These dis- courses of knowledge about the environment are tied in with dominant colonial/post-colonial repre- sentations of local actors in developing countries as the ’other’ whose subjugation leaves them without voice (Hobden 1995; Shiva 1988). As Leach and Mearns (1 996, 5) note, ‘rural people’s ecological knowledge is notable mostly by its absence, silenced before it is investigated’. This ’received wisdom’ has come under attack from two fronts: by scientists, who now recognize the inherent spatial and tem- poral variability of the natural environment and the ISSN 0004-0894 0 Royal Geographical Society (with The institute of British Geographers) 1999

The final fifth: autobiography, reflexivity and interpretation in cross-cultural research

Embed Size (px)

Citation preview

Area (1 999) 3 1.4, 3 13-325

The final fifth: autobiography, reflexivity and interpretation in cross-cultu ral research

Chasca Twyman, Jean Morrison and Deborah Sporton Department of Geography, University of Sheffield, Winter Street, Sheffield S10 2TN.

Emails: [email protected] and [email protected]

Revised manuscript received 22 January 1999.

Summary This paper addresses concerns about the complexities o f cross-cultural field- work, and the importance o f the positionality o f researchers and translators in the research process. These concerns build upon debates within the social sciences about auto- biography, reflexivity and the research process, as well as notions of validity, reliability and ‘truth‘. The paper re-examines these debates in the context o f cross-cultural research by focusing on praxis-actual experiences in the field.

Interpretation is four fifths of the truth, but not enough. (McFarlane 1996, 15)

Introduction This article has evolved from some of the common concerns and shared experiences of two teams conducting society-environment research projects in Botswana between 1994 and 1997. Both projects aimed to provide a greater understanding of societal-environment interactions under the simul- taneous impacts of climatic variability and structural land-use changes. The main project focused on changes arising from the expansion of commercial pastoralism, while the doctoral research examined the recent changes in community development and wildlife management policies. The article draws, in particular, on the fieldwork experiences of two researchers working independently on the projects in Botswana. Their experiences are used to examine the issues of autobiography, reflexivity and interpretation in the cross-cultural research process.

The discussion of methodological issues presented here is informed by wider debates within the social sciences, which critique the bases upon which knowledge is (re)produced and articulated in a cross- cultural context. Central to these debates is the

postmodern rejection of grand narratives and total- izing theories predicated, in the context of this research, on Western modernist assumptions about the environment and local knowledge (Gandy 1996). Leach and Mearns (1996) argue that our under- standing of environmental issues such as ‘desertifi- cation’, overgrazing and wildlife management has been informed by orthodoxies founded upon ’received scientific wisdom’, which point to the destructiveness of local practices, casting local actors as both agents and victims of what are widely perceived as environmental catastrophes. Accord- ing to this view, the physical environment and society are represented in essentialist terms as sep- arate analytical categories. Orthodoxies are then re-enforced as they become embedded in insti- tutional structures and hence inform power relations and policy at all levels (Foucault 1981). These dis- courses of knowledge about the environment are tied in with dominant colonial/post-colonial repre- sentations of local actors in developing countries as the ’other’ whose subjugation leaves them without voice (Hobden 1995; Shiva 1988). As Leach and Mearns (1 996, 5) note, ‘rural people’s ecological knowledge is notable mostly by its absence, silenced before it is investigated’. This ’received wisdom’ has come under attack from two fronts: by scientists, who now recognize the inherent spatial and tem- poral variability of the natural environment and the

ISSN 0004-0894 0 Royal Geographical Society (with The institute of British Geographers) 1999

3 14 Twyman et a1

problems of over-generalization (Behnke et a/ 1993; Thomas 1993; Thomas and Middleton 1994); and by post-structuralist critiques, which stress the impor- tance of local non-Western discourses of knowledge at the margins (Peet and Watts 1993; Said 1993).

Within anthropology, epistemological tradition has long favoured in-depth studies that attempt to under- stand local discourses of knowledge. The interpret- ation and representation of such knowledge has, however, been criticized (see, for example, Moore 1996; Okely and Callaway 1992). Okely (1996, 1 ) suggests that ‘anthropologists of “exotica” may have unwittingly relied upon taken-for-granted assump- tions of difference when confronted with other cul- tures’. The positionality and authority of the ’expert’ researcher vis-A-vis the researched in the field has been called into question (Radcliffe 1994; Rose 1997), and the textual representation of ’local’ knowledge has been criticized for re-enforcing the politics of difference (Miles and Crush 1993; Okely and Callaway 1992). Research accounts can there- fore only be understood with recourse to the biography of the researcher (Okely 1996) and the contextualization of the research process (Frenk 1995).

Issues of representation and unequal power rela- tions between the researcher and the researched are further compounded when language interpretation and translation art! involved (Smith 1996). Language plays a central role in the construction of meaning, which may be lost through translation (Derrida 1991 ), thus misrepresenting the views of the researched (Spivak 1993). This can lead to the compensation of issues of representation and unequal power relations between the researcher and the researched (Nast 1994; Smith 1996). The mul- tiple texts produced in such instances have the potential to distort languages and construct new meanings (Derrida 1991). Some of these issues can be circumvented, in part, by recognizing what Smith calls:

the hybridity of contact between self and other, between ’home’ language and ’foreign’ language [which] can give rise to ‘inbetween’ forms of under- standing. (1996, 165)

tiowever this requires a more critical awareness of the positionality of ‘self‘ and ‘other‘ (Mather 1996), and a focus on the political dimensions of translation (Sturge 1997). In essence, this requires a shift from researcher as observer to researcher as observed,

and with this an emergence of self-reflexive and context-laden fieldwork (Frenk 1995; Mather 1996; Okely 1996; Reay 1996; Rose 1997).

The first section of this paper provides more detail about the two research projects, the research teams and the methodological approaches adopted. The second section provides two self-reflexive accounts of fieldwork by team members. Through auto- biographical accounts, each draws on their different disciplinary backgrounds to examine fieldwork experiences, emphasizing the influence of their reflections on the interpretations of meaning in the research process. Translators played an important role in both projects and thus their positionality must also be considered. In the third section of the paper, these issues of positionality and reflexivity are used to examine the problems and possibilities of cross- cuhural translation and interpretation of texts. Here, Smith’s ( 1 996, 165) notion of ‘ “inbetween” forms of understanding’ is central to the unpacking and con- textualization of the ‘shades of meaning’ that emerge from the varied texts of this complex fieldwork process.

The research projects

fhe main project’ examined society-environment links arising from the expansion of commercial pastoralism into the dry-subhumid to arid Kalahari Desert of Botswana (Thomas and Sporton 1997). The context of structural and land-use changes in Botswana is provided by the 1975 Tribal Grazing Lands Policy (TGLP), which aimed to enhance national economic and social development by reforming Botswana’s livestock industry, improving the livelihoods of rural dwellers and ameliorating environmental degradation. To complement and extend the main research focus, an additional (doctoral) study2 was carried out on the Wildlife Management Areas (WMAs) that had been estab- lished in areas zoned as ’reserved’ under TGLP. These areas were developed to redress some of the social problems associated with the commercial expansion of the livestock industry. The research examined the extent to which the Kalahari WMAs, and the more recent implementation of community- based projects, were changing the access to, effective use of and management of the natural resources of the rural populations living within these areas.

The main project comprised a multidisciplinary team of four researchers: a physical geographer,

Autobiography, reflexivity and interpretation in cross-cultural research 3 1 5

a range ecologist, a population geographer and a social anthropologist. A range of environmental and social methodologies were used during the three- year project. These included environmental surveys and rangeland assessment exercises, questionnaire surveys, interviews and in-depth case-study work in remote ranches. Local government dices were vis- ited in the three study districts, as well as the central ministries in Gaborone. The doctoral research was conducted by a development geographer, with some field assistance from a biologist for an initial resource availability survey. Oral testimonies pro- vided the main sources of datainformal, semi- structured and repeat interviews, group discussions and informal conversations (compare Finnegan 1992). These were complemented by observations, participation on trips (such as gathering wild foods) and use of secondary sources. Fieldwork was located in two small remote settlements, as well as in the administrative centre of one district. Figure 1 shows the location of the field areas of the two projects, Study Areas 1-3 and Study Area A, respectivety.

Though both projects had a similar overall approach to the fieldwork, different disciplinary backgrounds and differently positioned researchers meant that the realities of fieldwork were often quite diverse. Both approaches shared a recognition of the validity of autobiography as part of the contextualiz- ation of the research process, and of its integrity to the unravelling of meaning in the crosscultural research context. However, both avoided the extreme position where reflexivity is followed to the extent that ’other cultures and other people’ become the ’exotic backdrops of authorial self-discovery’ (Lancaster 1996, 131). Furthermore, both field researchers ascribed significant importance to the networks of relationships and interactions surround- ing themselves and their translators. In much past ethnography, translators (interpreters, assistants, etc) have remained conspicuously absent in texts or are consigned to footnotes (see Sturge 1997). This paper therefore addresses both the autobiography of researcher and the biography of translator, and questions the power relations and subsequent production and processing of texts. Critical to such a linkage of theory and praxis is the acknowledgement of unequal power relations (vis-5-vis respondent, translator or researcher) in such a process, and the difficulty of attaining transparency in the autobiographical process (Rose 1997).

Researcher as observer or researcher as observed?

Post-doctoral research across the Kalahari: lean’s story I joined the team in Sheffield as a postdoctoral research associate from an academic background in social anthropology. My fieldwork in Botswana took place over a three-year period, located in three study sites across the Kalahari rainfall ecological gradient. The initial six-month trip in 1994 was spent administering a questionnaire survey with the aid of translators, at both the household and individual level. The second five-month period in 1995 involved a return to two of these districts in order to conduct case-study work on a much smaller scale. Owing to pregnancy, I was unable to visit the third district at that time, and consequently returned in 1996 for one month to complete the case-study work.

Presentation of self I was acutely aware of the political implications of the research (ie that we were researching potentially politically sensitive topics such as poverty, discrimination, marginalization, etc) and felt a need to be disassociated from the journalists and others doing ‘quickie’ surveys and dissertations based on one-off encounters, as these often present very deterministic portraits of societies and environments. Respondents’ previous experi- ences of direct questions were with government census officials and tax and wildlife officers. I felt there was a strong possibility that respondents would give what they felt were the appropriate answers in order to dispose of the intruder, a classic sign of interview fatigue and an illustration of the complex power relations involved in any such encounter. A number of respondents claimed they had pre- viously been questioned by researchers, but none knew who these researchers were, where they were from or the purpose of their visits. In light of this, I felt that it was extremely important that I made my intentions as clear as possible. I therefore presented myself as a Scottish researcher from a British univer- sity, working on a research project in conjunction with the University of Botswana. I explained that I wanted to learn about their lives, problems and needs in order to write a report. However, I also stressed that I was powerless to ensure that the report was read by the government or acted upon in any way. Confidentiality was, of course, ensured for all respondents.

3 16 Twyman et al

During the course of these three periods of field- work, respondents’ perceptions of me were continu- ally subject to modification. Attitudes toward me shifted as I moved from being seen as young and rather anomalous in terms of gender identity during the first period of fieldwork, to being seen as more clearly gendered during the second period (when I was obviously pregnant), and even more so during the third trip (when I was accompanied by my nine-month-old daughter).

The initial field trip Establishing contacts with respondents during anthropological research is usually a long process of patient negotiation and renegotiation, of introductions and mutual familiar- ization. Owing to the timing of the research project, this process was not possible, and my relationships with respondents were by necessity rather fleeting, a positioning with which I felt uncomfortable. Initially respondents were uncertain how to ‘place‘ me (compare Edwards 1990): for example, I wore a wedding ring, yet was alone (without a man) and childless. Ranch workers tended to interpret my presence as that of an outsider and a foreigner, as well as a woman-an unfamiliar combination in such areas. Being female caused fewer problems than I had anticipated in such a male-dominated society as Botswana, and overall I think I was generally viewed as a rather non-threatening, uninfluential young foreign woman. At first, people were suspicious of me for being blonde and having a vehicle with South African number plates: they often thought I was Afrikaner. In some of the study areas there is a controversial history of indigenous contact with Afrikaner settlers in Botswana, and this no doubt influenced my ‘positioning‘ (both initially and subsequently, following introductions and my disassociation from this category).

The ranch workers were all reasonably eager to participate in both the questionnaire and the case- study work, perhaps partly in anticipation of the ’gifts’ they hoped to receive. Understandably, many viewed me as a resource to be exploited, a store of desirable goods, and I was constantly subjected to a barrage of requests for tea, coffee, sugar, clothing

been set by previous anthropologists in the same location and the 1Kung expected them to follow suit. In the majority of cases, I agreed to the ranch workers’ requests for small items such as cigarettes and tea, but this often led to acrimony. For example, I was soundly scolded by one women as she felt I was favouring another over her. The implications of what I did and did not do undoubtedly had a marked influence on how I was treated within the ranches and thus had far-reaching effects upon the processes of interviewing and interpretation.

The return trip During the second and third field- work periods, those respondents who remembered me from the previous year were more ‘open’ with me, that is, they were more willing to disclose information about their personal situations and rela- tionships, and in return were more curious about my own circumstances. Often they would retell the story of when I had first arrived the previous year, where I had pitched my tent, how 1 had given someone a red hat, how I had taken a sick child to the clinic and so forth, re-establishing and reaffirming relationships between us (compare Nelson 1998).

Returning with a child, I was perceived to have attained adult status. Women in particular were more likely to question me, particularly about my pregnancy, and also about my daughter. In Botswana, children are viewed as a necessary com- ponent of a satisfactory life, therefore for someone to reach their late twenties and be childless is seen as unusual. The overwhelming majority of women I encountered during my third field trip were eager to pick up my baby and play with her, and she became a focus for discussions on issues of gender, fertility and childcare. The transient relationships that devel- oped between myself and female respondents were characterized by a persistent tension between feel- ings of commonality regarding our gender identities (as mothers), and of difference regarding our cultural identities. I therefore made particular efforts to use this hybrid experience to look deeper at the ’ “inbetween” forms of understanding’ that such an experience highlighted (see Smith 1996).

and cash wherever I went. Margorie Shostak (1 98lj encountered similar requests during her fieldwork with the !Kung in the northern fringes of the Kalahari. Unwilling to promote dependency, she and her husband initially resisted requests for tobacco and money but quickly came to the realization that they could not ignore these demands. The precedent had

Working relationships The time I spent conducting fieldwork in Botswana was filled with a curious mixture of elation, depression, pleasure and con- fusion. I found Setswana a difficult language to learn quickly and thus was forced to rely on hired trans- lators, none of whom was particularly keen to teach

Autobiography, reflexivity and interpretation in cross-cultural research 3 1 7

Figure 1 Map of Botswana showing h x a h of study areas

me their language. There is a tendency within anthro- pology for ‘native languages’ to be ‘easily mastered’ by researchers, but it i s questionable whether subtle nuances and multiple meanings can be inferred from such rudimentary understandings (Sturge 19971, given the timescales of research involved in much of this work. Moreover, this still leaves the complex issues of translation and interpretation to be addressed. 1 employed several translators during different stages of the research, as I found it difficult to secure anyone who was fluent in English and prepared to spend any length of time in ‘the bush’. 1 generally preferred to employ female translators for several reasons-we had to share a tent and spend 24 hours a day together for extended periods- moreover, one component of the research involved

asking women questions concerning fertility issues. 1 respected my translators and liked them, but there were also moments when I was irritated and even enraged by them. The translator-researcher relation- ships that were formed were ones of constant renegotiation and mutual exploitation.

Doctoral research in Western Botswana: Chasm‘s story 1 joined the project in Sheffield as a postgraduate researcher aiming to conduct research into commu- nity development and wildlife management in Botswana. My research was based in Ghanzi District in western Botswana and over half my time was spent in two small settlements in the Okwa WMA. Over a fifteen-month period, I made repeat visits to

3 18 Twyman et al

these settlements, ranging from three to four weeks to just a few days at a time. During this time, I employed the same translator, Lefenya, who was local to the district. The rest of my time was spent in the main town, meeting and interviewing govern- ment officials, completing the transcription of inter- views or travelling to Gaborone to use libraries and visit central ministries and departments.

Presentation of self My reception by government officials and people in the settlements varied. Over- all, in the local government I felt warmly received and people were very helpful. I presented myself as a student from a British university conducting inde- pendent research, with a link through the University of Botswana. Living in the district for fifteen months certainly helped, as I became ‘known’. Many researchers simply make one-off visits and provide little feedback to those interviewed, and so I was aware that discussion and feedback were necessary conditions for the successful continuation of the research. Part-way through the fieldwork I provided a preliminary report (Twyman and Hurst 1996), which was discussed in the interviews, and offidals were invited to comment on other written work throughout the fieldwork if they wished. This not only provided valuable feedback for my own research but enabled officials to ’see what I was doing’. As I began to share my work with them, they simultaneously allowed me insights into their work, giving professional and personal views on my research topic. I felt my status as an independent student (ie a learner), rather than a professional academic or consultant, helped this relationship to develop.

In the settlements, I again tried to stress my independence from government or aid agencies, consultancies or companies, for similar reasons to those given by Jean. The politically sensitive nature of some of the research meant that perceived allegiances to government or industry were to be avoided. I presented myself as a student, and this was often translated as ‘school child from a big school’. I tried to be honest and open about what I was researching and what the outcomes of the research would be (see Burgess 1984; Devereux and Hoddinott 1 992). Respondents recalled times when researchers had visited, and in all recent cases they had received no feedback, despite promises of reports and projects; some regretted having given up time to help the researchers. I stressed that while

I would write reports and distribute them in the government and settlements (which I subsequently did), I could not promise that they would be read or that any changes would occur.

All interviews with individuals and groups were taped. Permission to tape an interview was gained after the research project had been explainecCthis was done at the beginning of every meeting or interview, even towards the end of the fieldwork, when some respondents had been interviewed many times. It was not considered unusual to repeat and stress explanations and was seen as part of the structure of the interview, along with any introduc- tions and greetings. Respondents gradually became familiar with the nature of the interview and would often explain it to others who were less familiar before we began. Making the interview framework transparent enabled respondents to understand our process of gathering information. As respondents became familiar with this process they were able to ’package‘ information more consciously in response to questions, thus changing the power relations of the interviews considerably. Respondents reflected on past questions and returned to points or brought up new ones sparked from earlier discussions. This way they became more actively engaged in the interview, rather than simply answering questions.

Undoubtedly, being white and female had an effect on the way I was received both by govern- ment officials and in the settlements. My status as white, and having a vehicle, meant I was, as Jean also found, a potential source of ‘gifts’ in the settlements. Demands were continually made for tobacco, soap, tea and sugar. Though other researchers have found themselves forced into this demand-based sharing (compare Jean’s experience; Shostak 1981 ), Lefenya and I decided that this would perpetuate depen- dency and the attitude that ’whites will give’. We were due to keep returning regularly to the same people and places, hence this was not an attitude we wished to promote. Durham (1995) considers that many demands in Botswana are ‘playful’ and part of a social dialogue that does not actually require the demand to be fulfilled on most occasions; Lefenya had an understanding of when demands should or should not be met (see Peterson 1993) and thus our own social dialogue was established. There were no formal payments made to respondents for inter- views, but some mutually beneficial arrangements were made: for example, one respondent spent the day telling us about collecting grass, building materials and veld foods found in the area, in return

Autobiography, reflexivity and interpretation in cross-cultural research 3 1 9

for our transporting the collected grass back to the settlement in our vehicle.

Presentation of ourselves The ethics of ’third-world research’ have long received wide attention within the geographical literature (Farmer 1983; johnston 1984; Gilbert 1987; Potter and Unwin 1988; Sidaway 1992; Crush 1993; Kay 1993; Porter 1995). it is often assumed that the researcher is an ‘out- sider’, whereas if they have a research assistant or translator they are ‘insiders’. Researchers investigat- ing their own cultures are similarly seen as ‘insiders‘ and thus privileged in the research process (Mather 1996; Miles and Crush 1993). These assumptions are misleading and this outsider/insider duality is not clear cut; rather, such positions allow a ‘between- ness’ that implies that no one is an ‘outsider‘ or ‘insider‘ in any absolute sense (Nast 1994, 57). In the case of my research, I, as an outsider, felt that my presence in the settlements was strongly influenced by the translator-who she was and how she (as well as I) interacted with respondents and other people in the settlements. Lefenya, the translator, was also an outsider, but to a different degree.

Lefenya, the translator, was part-Scottish and part- Nama (she called herself ‘coloured’) and came from a well-known family in the district. Her elder brother was active in local politics and held an influential government position. Lefenya spoke five languages fluently: Setswana, Afrikaans, English, Sekgalagadi and Naro (and was literate in the first three). She could also understand some Herero and G/wi. It was unusual for someone of her ‘status‘ to speak Naro (one of the Basatwa-San or Bushman- languages), but as a young child she had spent some time living in Bere (a remote settlement in the Matlho-a-Phuduhudu WMA), where she first began to learn the language, and later she went to school in D’kar, a Naro-speaking area. Since then she has continued learning the language out of her own interest.

I was initially concerned that Lefenya‘s association with her politically active brother would be detrimen- tal to the research, that respondents might regard her and therefore me with suspicion (see Katz 1994, 69 for similar concerns about the political dynamics of her research in Sudan). However, it was Lefenya’s ethnicity that seemed the most influential factor in our relationship with the respondents. During the first few months of the fieldwork, we were told by respondents on numerous occasions that they

would not be talking to us if Lefenya was ’black’. Considerable tension exists in the settlements between the Basarwa and the ‘blacks’. ‘Black‘ is the term used by Basarwa to describe Batswana or Bakgalagadi people, most often government workers (such as extension workers or teachers) in the settle- ment, or people who had come to the settlement to live and set up small trading businesses. These tensions were of interest to me as part of the research, but initially I had not realized Lefenya’s ethnicity would be so important. That Lefenya was not ‘one of them’ to either side was of greater importance than her family associations.

All the issues discussed above had a significant influence upon the research. As an outsider, I had a history that was unfamiliar to people. Lefenya, how- ever, had a background that people could relate to and understand more easily, but she was clearly very different from the people with whom we were working. Therefore, when considering my positional- ity, I had also to consider her positionality in the field and in the interview, and this must be viewed as equally as important Given Lefenya’s presence throughout the settlement-based research, I often refer to decisions we made and things we did. This is because the fieldwork was not carried out alone and Lefenya and I would be in constant discussion about plans, decisions and meanings and interpretations of interviews.

Discussion: differences and commonalities The two accounts above highlight some common themes and differences between the researchers’ fieldwork experiences, which reflect current debates surrounding autobiography, and the links between positionality, interpretation and translation. Major differences are reflected in the different time spans of the fieldwork and the location of ‘the field‘ in the two projects. While Chasca was able to build upon interviews and relationships with respondents over a fifteen-month stretch, Jean was working in three different areas with three shorter field visits over a longer time period. Jean was also working with a highly mobile population, meaning that respondents had often moved on by the time she returned. Again, the temporal and spatial difference in the fieldwork meant that Jean employed numerous translators, while Chasca was able to employ one person for the duration of the work. A further difference was the gendered experience of the fieldwork, which in Jean’s case became significant as she conducted

320 Twyman et al

fieldwork throughout her pregnancy and later with a small child.

Whilst the different methods adopted by the two researchers did not stereotypically reflect their disciplinary backgrounds, the ways in which they positioned themselves within ‘the field’ were greatly influenced by their respective disciplines. For Jean, disassociating herself from ‘other researchers‘ and establishing her place as ‘an ethnographer’ was important to her positioning and how she felt this affected her relationships with both respondents and translators. Chasca, however, did not feel ’con- strained’ by anthropological tradition and felt she could draw upon multiple disciplines to support her fieldwork. By working so closely with one translator, problems and confusions could be discussed and returned to in later interviews.

Herod ( 1 993) stresses that geographers who use interviews as part of their panoply of information- gathering practices need to be critically aware of how particular social relations within these practices shape the research process. This is crucial for all interviewing, but perhaps more complex for those conducting interviews in cross-cultural situations, particularly through translators. The gender of the researcher, translator and respondent is generally recognized as being pertinent to the amount and type of information that an ‘outsider’ can elicit in an interview (Herod 1993), the social interactions of the interviewing process (Miles and Crush 1993), the ’packaging’ of the information (Mosse 1994; Rose 1997) and the interpretation of these ’interactive texts’ (Rosaldo 1989).

Gender relations (both between women and between women and men) are dynamic processes, and the ways they shape behaviour and power relations, and thereby the research process, change over time and space and underpin the very context within which the interview itself takes place (Finch 1984; Herod 1993; Mather 1996; McDowell 1992b; Miles and Crush 1993; Nast 1994; Oakley 1981; Schoenberger 1991; 1992), as Jean’s account clearly illustrates. The cross-cultural interviews in these research projects contained a plurality of ’gendered assumptions’ by all parties (see McDowell 19881, further complicating the unravelling of meanings and understandings. These dynamics of gender and power relations both within and between cultures significantly affected the type of information gath- ered and the subsequent interpretation of this infor- mation (compare Radcliffe 1994; Rose 1997). These are important issues, and were considered before,

during and after the fieldwork process; they also had to be explained to the translators if they were to be followed in the interviewing process. One advantage of using the same translator throughout the field- work, as in Chasca’s case, was that they became highly skilled and sensitive to the dynamics of the interview and translation process.

In both cases, the researchers felt the need to present themselves in a way that emphasized their powerlessness in relation to the government and any potential benefits that might be assumed to result from the research. Both researchers were young, white and female, and both shared experiences associated with this ascribed status. Requests for gifts or lifts were commonplace but not overwhelming, though each researcher dealt with these situations slightly differently, drawing upon different discipli- nary backgrounds to reconcile these complex issues. The ‘positioning’ of the two researchers and their translators influenced how the interview and thus the ‘data‘ were shaped and constructed. The next step is therefore to consider how to unravel and understand these ‘data’ in light of the above. This shifts the focus to the problems and possibilities of cross-cultural translation and interpretation, and the rearticulation of meanings.

The lexicon of translation, interpretation and meaning

According to Sturge, translation is a practice of intercultural communication, a task in which we:

understand other cultures as far as possible in their own terms but in our language, a task which ultimately entails the mapping of the ideas and practices onto Western categories of understanding. (Tambiah 1990, 3, quoted in Sturge 1997, 21 )

This notion of mapping ideas and practices provides an interesting framework through which to view translation. By viewing translation as a mapping process, both the theory and practice of translation can be brought together. It is this process of map- ping one language onto another that should be subject to the same self-reflexive approaches as interpretations of other texts, however they are produced.

Smith problematizes the process of transferring concepts and meaning through language and trans- lation, and through this analysis she argues that, by examining the mapping process, these ‘problems’

Autobiography, reflexivity and interpretation in cross-cultural research 32 1

can give rise to ‘ “in-between” forms of understand- ing’ ( 1 996, 165). She gives examples of transferring concepts from German to English, yet in the context of this paper, concepts are being translated from source to English through a translator. This makes the earlier discussion of research experience and the notion of positionality and reflexivity of both researcher and translator directly relevant. In such cases, the researcher is further ‘displaced’ and ‘decentred’ (and even distanced) from the source language, raising problems of ‘authority’ in translation, but perhaps allowing the potential, through the agency of a translator, for a more obvious view of the ‘hybrid spaces’ mentioned by Smith (1996, 163).

According to Frenk ( 1 995), Derrida ( 1 991 ) argues that every translation creates a new text rather than a copy of another text, this being the essence of deconstruction or post-structuralism. Taking this view, Frenk says that:

every utterance in our interviews becomes ‘untranslat- able‘ since it is shaped, framed, configured by and for both its immediate, contingent contexts and its cultural context for which there is no equivalent in another language . . . [and thus such subtle nuances are easily] lost, distorted or exaggerated when the utterance is translated, reconfigured, rearticulated. (Frenk 1995, 138)

At the other extreme, some translations are simply presented as finished products and made to look ‘natural’ in the final language, having concealed all the difficulties. Frenk ( 1 995) argues that ‘transcultur- ation’ requires translation to include the cultural context and to explore a more dynamic notion of the transformation of text by carefully describing the process of translation and rearticulation (Mosse 1994). According to Lefevere and Bassnett (1990, 1 l ) , ‘translation as an activity is always doubly contextualized, since the text has a place in two cultures’. It is clear therefore that the value-laden processes of translation and transcription need to be examined fully in the contexts in which they were constructed: an integral part of these contexts is the autobiography of researcher and translator and the implications of power relations on the production of texts.

In the context of these research projects, ethnic diversity heightened the problems of translation. Anthropologists generally prefer the bounded uni- verse of a small cultural group, but for the purposes of both research projects it was important to look at

different people’s responses to environmental and socio-economic change. This proved particularly challenging given the number of languages required to conduct the research, and both researchers, as illustrated, were heavily dependent on their translators.

For Chasca, discussing the translation process with the translator while transcribing proved to be invalu- able. It was soon decided that literal translation in the interview was not possible (compare Hatim and Mason 1990). The translator would therefore sum- marize as accurately as possible what was said by the respondents, enabling the researcher to conduct the interview. Full literal translation was then attempted during the transcription process, where further discussion could take place. It is much easier to convey the meaning of what someone has said quickly than it is to recount word-for-word what someone has said in another language. This situation is not ideal, as the researcher is further ‘distanced’ from the original ‘text’. However, given the diversity of languages needed to conduct this research there was little alternative. By transcribing in the field it was possible to feed back and follow up on issues in later interviews.

Transcripts were produced in English in both projects. In both cases, transcribing was done mainly with pen and paper in the field or in the nearby town. The transcripts produced by Chasca and her translator incorporated all that was said in the inter- view, including the translator’s translations at the time. By doing this, though time-consuming, it was possible to see the dynamics of the interview struc- ture and the translation process. In Jean’s case, it was more appropriate to translate chunks of text from interviews (life histories or focus discussions) rather than follow the methods adopted by Chasca. In most cases, it would have been virtually impossible to produce transcripts in the original languages as many are non-written, or written solely by academics in complex linguistic notation. Transcripts could not have been checked with respondents, as many were illiterate and certainly could not read in their first language. Consideration had to be given to the production of texts in English rather than source language, and the mapping processes entailed in the production of these texts.

Several of these issues can be illustrated in the following example. The dialogue below is taken from an interview conducted by Chasca and Lefenya about hunting and hunting methods in one of the settlements.

322 Twyrnan et al

C h a m (language spoken-English): And for the small ones, the duiker and steenbok, do they use the trap for them? Lefenya (Setnvana/Naro): For hunting the duiker and steenbok, do you use, what do you call it, w i n d Respondent (Naro/Setswana): Qwine, mogala? L (Setswana): Yes. R (Setswam): No. Those things were used by the people from the older days. L (English): OK, in the older days, people used those things. C (English): Tell him that we have seen people with the trap, in Bere they still use it. L (Setswana): She says that she has seen people using those things. R (Setswana): Those people are the first people. L (English): He says they are still old-fashioned. C (English): So how does he catch the duiker and steen bok? L (Setswana): How do you catch them? R (Setswana): I catch them with a gun.

In this instance Lefenya translates on-the-spot that the people he is referring to are ‘still old-fashioned’ and therefore use those methods. However, when we were transcribing and listening carefully, she noted that he referred to the people as ’the first people’. In Setswana, he said ’tsimologo ya ntlha’, which literally translates as the ‘the beginning. the first people’, but his use of Setswana was poor and grammatically incorrect The key question here is whether he was referring to the ‘first people’ in a respectful way, the usual way of using this phrase, or whether Lefenya’s initial reaction and translation at the time of the grammatically incorrect Setswana phrase is more in line with what this respondent was trying to convey.

Two very different interpretations could be drawn from this piece of dialogue. First, this young man could be trying to distance himself (metaphorically) from the people further south, whom he views as backward and old-fashioned in their hunting methods. Though such traditional hunting methods are still widely used there, he could be trying to ‘displace’ and ‘decentre’ himself from the traditional institution of hunting. Yet a second reading could see the reference to ‘first people’ as explicit acknowl- edgement that he holds great respect for these traditional institutions. However, given the comment about hunting with a gun, this might also suggest an element of discordancy as to where he ’centres’ himself within society. Using these hybrid spaces of interpretation helps us to understand both how the mapping process of translation works, and also how

the shades of meaning through this process can be interpreted or unwittingly (or deliberately) moulded, shaped and manipulated. Further interviews with this respondent about other ‘traditional activities’ such as veld burning suggest that, in this case, the second reading might be more appropriate. This respondent sees inconsistencies in the practices from the past, which he understands, and the government direc- tives, which often contradict his traditional knowl- edge. His incongruent views illustrate the shifting of critical values within society, but also by individuals. To make sense of the views he holds, he begins to devalue the importance of burning the veld or hunting, discarding his tradition and therefore reconciling the discordant perspectives.

It also became clear to Lefenya, Chasm’s trans- lator, that when people did not speak in their first language they often constructed whichever language they were speaking as if it were their first language. Often words from other languages would be used to describe certain issues, hence detailed knowledge of a range of languages was necessary, as well as a recognition of the politics of who used specific languages at particular times. Whenever such changes in language occurred, they would be noted in the transcripts and discussed. Through these dis- cussions with the translator and with respondents in interviews, the research project highlighted the diffi- culty of transferring concepts across languages. jean had a similar experience with translation and tran- scription, and similarities and differences were often discussed. The detailed transcription process and the discussions between researcher and translator proved invaluable in understanding the types of texts we were producing. However, the researcher can never have access to the ‘original’ text, only the secondary text produced with the translator, and this has significant implications for the interpretation of meanings and the power and authority of the texts.

These issues of translation and meaning are there- fore clearly linked with issues of representation and power. As we have seen, language plays a central role in the construction of meaning, yet when one language is mapped onto another there is a compen- sation of these issues. Ultimately, the multiple texts produced in such instances have the potential to distort languages and construct new meanings (Derrida 1991). To utilize these ‘distortions’ and ‘new meanings‘, the ‘in-between‘ forms of under- standing must be addressed and the ‘hybrid spaces‘ investigated. Though problems of ‘authority’ in trans- lation can be raised, there is potential to harness this

Autobiography, reflexivity and interpretation in crosscultural research 323

hybridity through the agency of a translator, in order to open what Smith ( 1 996, 163) calls ‘new spaces of insight‘.

Conclusion

One of the driving themes of this paper is that accounts and interviews must be interpreted in relation to the contexts in which they occur. This is not a new statement in geography, but, within the context of cross-cultural research with a translator, these issues have usually been given little promi- nence. How the researcher, translator and respon- dent interpret their own and each other‘s actions depends on a number of factors, as illustrated in the discussions above. Mosse believes that,

information does not just exist ’out there’ waiting to be ‘collected‘ or ‘gathered‘, but is constructed, or created, in specific social contexts for particular purposes. (1994, 499)

Interviews, meetings and focus groups are all differ- ent types of social contexts and thus will elicit different types of information or accounts (Burch 1993; Goss and Leinbach 1996). Opposition has also grown to the idea that it i s possible to carry out value-free objective research (Rosaldo 1989), which can ultimately lead to a crisis in representation (McDowell 1992a). Instead, there is a recognition that we are bound up with questions of authority, communication and representation and, in the con- text of cross-cultural research, the power relations that exist between academics and their research subjects (Radcliffe 1994; Rose 1997; Staeheli and Lawson 1995).

This paper arose out of common concerns about the complexities of cross-cultural fieldwork, and the obvious importance of the positionality of researchers and translators in the research process. These concerns build upon debates within the social sciences about autobiography, reflexivity and the research process, as well as notions of validity, reliability and truth (Chambers 1997; McDowell 1 9 9 2 ~ Miles and Crush 1993). This paper has reexamined some of these debates in the context of crosscultural research by focusing on praxis: actual experiences in the field. Researchers are aware of the constant need for rigour in methodology, juxta- posed with the need for flexible methodological adaptations in the field (see Mosse 1994). This highlights the critical need to recognize the realities

of fieldwork as part of the contextualization of the fieldwork process. If interpretation is four-fifths of the ’truth’, perhaps autobiography, positionality and reflexivity in the contextualization process go some way to making up the final fifth.

Acknowledgements The research for the main project was funded by the Economic and Social Research Council. The doctoral research was funded by a scholarship from the ESRC, with additional funding from the University of Sheffield. The authors would like to thank respondents, translators and all who were invoked with the research in Botswana. Thanks are also extended to the other members of the research team for valuable discussion throughout the fieldwork process, and to the useful comments of two anonymous referees.

Notes

1 ESRC Global Environmental Change Research Project,

2 ESRC-funded.

References

Behnke R H, Scoones I and Kerven C (eds) (1993) Range ecology at disequilibrium: new models of natural varia- bility and pastoral adaptation in African Savannas (ODI, London)

Bloch M (1 991) ‘Language, anthropology and cognitive science’ Man 26(2), 183-98

Burch E S (1993) ‘Comments and reflections: more on the reliability of oral and traditional history‘ American Anrhro- pologist 95(3), 651-2

Burgess R G (1 984) In the field: an introduction to field research Contemporary Social Research Series 8 (Routledge, London)

Chambers R (1 997) Whose reality counts? Putting the last first (Intermediate Technology, London)

Croll E and Parkin D (1990b) ’Cultural understandings of the environment‘ in Croll E and Parkin D (eds) Bush base, forest farm: culture, environment and development (Roudedge London), 11-36

Crush I (1 993) ‘The discomforts of distance: post colonial- ism and South African geography’ South African Ceo- graphical journa/ 75(2), 6&68

Derrida J (1991) ’From “Des tours de Babel”’ in Kamuf P (ed) A Derrida reader: between the blinds (Harvester Wheatsheaf, Hemel Hempstead), 243-53

Devereux S and Hoddinott I (eds) (1992) Fielhvork in developing countries (Lynne Reimer, Boulder, CO)

Durham D (1995) ‘Soliciting gifts and negotiating agency the spirit of asking in Botswana‘ journal of the Royal Anthropological Institute 1, 11 1-28

Edwards R (1 990) ’Connecting method and epistemology: a white woman interviewing a black woman’ Women’s Studies International Forum 13(5), 477-90

Ref No 13202531 74.

324 Twymanetal

Farmer B H (1983) ‘British geographers overseas, 1933- 1983’ Transactions of the lnstitute of British Geographers 8, 70-79

Finch J (1 984) ’ “It‘s great to have someone to talk to”: the ethics and politics of interviewing women’ in Bell C and Roberts H (eds) Social researching: politics, problems and practice (Routledge, London), 70-87

Finnegan R (1992) Oral traditions and the verbal arts: a guide to research practices (Routledge, London)

Foucault M ( 1 981) ‘The order of discourse’ in Young R (ed) Untying the text‘ h poststructurdist reader (Routledge and Kegan, London), 46-78

Frenk S (1995) ‘Representing voices: what‘s wrong with our life histories?’ in Townsend J G (ed) Women’s voices from the rainforest (Routledge, London), 137-43

Candy M (1996) ’Crumbling land the postmodernity debate and the analysis of environmental problems’ Progress in Human Geography 20(1), 2 3 4 0

Gilbert A (1987) ‘Research pOricy and review 15. From Little Englanders into Big Englanderr: thoughts on the relevance of relevant research‘ Environment and Planning

Goss J D and Leinbach J R (1996) ‘Focus groups as alternative research practice: experience with trans- migrants in Indonesia’ Area 28(2), 1 15-23

Hatim B and Mason I ( 1 990) Discourse and the translators (Longman, London)

Hendry J (1992) The paradox of friendship in the field analysis of a long-term AngleJapanese relationship’ in Okely J and Callaway H (eds) Anthropology and auto- biography (Routledge, London), 163-74

Herod A ( 1 993) ‘Gender issues in the use of interviewing as research method‘ Professional Geographer 45(3), 305-1 7

Hobden A ( 1 995) ‘Paradigms and politics: the cultural construction of environmental policy in Ethiopia‘ World

Johnston R J ( 1 984) ‘The world is our oyster‘ Transactions of the Institute of British Geographers 9, 443-59

Katz C (1994) ‘Playing the field: questions of fieldwork in geography‘ Professional Geographer 46(1), 67-72

Kay C (1993) Whither Africa and development studies?’ Area 25(2), 13741

Lancaster R N (1996) ‘The use and abuse of reflexivity’ American Ethnologist 23( l ) , 13G32

Leach M and R Mearns (1 996) ‘Environmental change and policy: challenging received wisdoms in Africa’ in Leach M and Mearns R (eds) The lie of the land: challenging received wisdom on the African environment (James Currey, Oxford), 1-33

Lefevere A and Bassnett S (1990) ‘Introduction: Proust’s grandmother and the thousand and one nights: the “cultural turn” in translation studies’ in Bassnett S and Lefevere A (eds) Translation, history and culture (Pinter, London), 1-1 3

Mather C (1996) The view from outside? Interpreting oral testimonies from rural South Africa’ South African Geographical /ournal78( 1 ), 13-1 9

A 19, 143-51

Devdopment 23(6), 1007-21

McDowell L (1988) ’Coming in from the dark feminist research in geography’ in Eyles J (ed) Research in human geography (Blackwell, Oxford), 154-73 - (1992a) ’Multiple voices: speaking from inside and

outside “the project” ‘ Antipode 21(1), 56-72 - ( 1 992b) ‘Doing gender; feminism, feminists and research

methods in human geography‘ Transactions of the Institute of British Geographers 17, 399-41 6

McFarlane F ( 1 996) Strangers meeting (African Publishing Services, Gaborone)

Miles M and Crush J (1993) ‘Personal narratives as inter- active texts: collecting and interpreting migrant life- histories’ Professional Geographer 45(1), 84-94

Moore D S (1996) ‘Marxism, culture, and political ecol- ogy‘ in Peet R and Watts M (eds) Liberation ecol- ogies environment development, social movements (Rodedge, London), 1 2 5 4 7

Mosse D ( 1 994) ‘Authority, gender and knowledge: theor- etical reflections on the practice of PRA‘ Development

Nast H J (1994) Women in the field critical feminist methodologies and theoretical perspectives: opening remarks’ Professional Geographer 46( 1 ), 54-66

Nelson N (1998) ‘From neighbourhood study to network analysis: 25 years of working with female heads of household in Nairobi‘ Paper presented at the African Studies Association of the UK Biennial Conference ’Comparisons and transitions’ School of Oriental and African Studies, London, 14-1 6 September

Oakley A (1981) ‘Interviewing women: a contradiction in terms’ in Roberts H (ed) Doing feminist research (Roudedge, London), 3 0 4 1

Okdy j (1 996) Own or other cuhure (Roudedge, London) Okely j and Callaway H (eds) (1992) Anthropology and

autobiography (Routledge, London) Peet R and Watts M (1993) ‘Introduction: development

theory and environment in an age of market trium phalism‘ Economic Geography 69(3), 227-53

- (eds) (1 996) Liberation ecologies: environment develop- ment social movements (Routledge, London)

Peterson N (1993) ‘Demand sharing reciprocity and the pressure for generosity among foragers’ American Anthropologist 95(4), 860-74

Porter G (1 995) ‘ ‘Third world’’ research by “first world” geographers: an Africanist perspective’ Area 27(2), 1 3 W 1

Potter R B and Unwin T ( 1 988) ‘Developing areas research in British geography 1982-1 987’ Area 20(2), 121 -6

Quicke J (1994) Ethnography (Research Methods Unit, University of Sheffield)

Radcliffe S A ( 1 994) ’(Representing) post-colonial women: authority, difference and feminism‘ Area 26(1), 25-32

Reay D (1996) ’Insider perspecb’ves or stealing the words out of women’s mouths: interpretation in the research process‘ Feminist Review 53, 57-73

Rosaldo R (1 989) Culture and truth: the remaking of social analysis (Roudedge, London)

and Change 25,497-526

Autobiography, reflexivity and interpretation in cross-cultural research 325

Rose G (1 997) ‘Situating knowledges: positionality, reflex- ivities and other tactics’ Progress in Human Geography 2 1 (3), 305-20

Said E (1 993) Culture and imperialism (Vintage, London) Schoenberger E (1991) ‘The corporate interview as a

research method in economic geography‘ Professional Geographer 43, 18&90

- (1992) ’Self-criticism and self-awareness in research: a reply to Linda McDowell’ Professional Geographer 44,

Shiva V ( 1 988) Staying alive: women, ecology and survival in lndia (Kali for Women, New Delhi)

Shostak M (1981) Nisa: the life and words of a !Kung woman (Earthscan, London)

Sidaway J D (1992) ‘In others worlds: on the pditics of research by “first world” geographers in the “third world“ ’ Area 24, 403-8

Smith F M (1 996) ‘Problematizing language: limitations and possibilities in “foreign language” research’ Area 28(2), 160-66

Spivak G (1993) The politics of translation’ in Spivak C (ed) Outside in the teaching machine (Routledge London), 179-200

Spradley J (1979) The ethnographic interview (Hdt Rhinehart and Winston, New York)

21 5-1 8

Staeheli L A and Lawson V A (1995) ‘A discussion of “women in the field‘: the politics of feminist fieldwork’ Professional Geographer 46( 1 ), 96-1 02

Sturge K (1997) ‘Translation strategies in ethnography‘ Translator 3(1), 21-38

Tambiah S J (1990) Magic, science, religion, and the scope of rationality (Cambridge University Press, Cambridge)

Thomas D (1993) ‘Sandstorm in a teacup? Under- standing desertification‘ Geographical /ourna/ 159(3),

Thomas D S G and Middleton N J (1994) Desertification: exploding the myth (Wiley, Chichester)

Thomas D S C and Shaw P (1991) The Kalahari environ- ment (Cambridge University Press, Cambridge)

Thomas D S C and Sporton D (1997) ’Understanding the dynamics of social and environment variability: the impacts of structural land use change on the environ- ment and peoples of the Kalahari, Botswana’ Applied Geography 17( 1 ), 1 1-27

Twyman C and Hurst J (1996) ‘Social and environmental profiles of the Okwa Wildlife Management Area and the Matlho-a-Phuduhudu Wildlife Management Area, Chanzi District’ Unpublished draft report, Department of Geography, University of Sheffield

318-31