A Local Researcher’s Experiences of the Insider–Outsider Position: An Exercise of Self-Reflexivity During Ethnographic GBV and HIV Prevention Research in South Africa

Abstract

Examining the researcher’s position during the research process is important to enhance the representation of research participants and the local context where knowledge is produced. This article aims to reflect the experiences of a local researchers’ insider–outsider position during a qualitative study drawing on ethnographic methods. Familiarity with the research community can be advantageous for an insider researcher position while an outsider researcher position may be enable more observance of aspects of daily life to which insiders may be blinded during the inquiry, interpretation, and data analysis. Researchers who often strive for a balance between the two positions can find it challenging yet rewarding. In this study, research participants played a critical role in shaping the local researcher’s dual identity as the outsider position remained at the forefront of interactions with the research process despite familiarity with the local language, culture, and research setting. This view, however, shifted after the local researcher seemed to have “earned” the insider position. Local researchers need to invest in extensive self-reflexivity, acknowledge the vulnerabilities of dual positionality, and capitalize on the shared qualities and differences with research participants to enhance the representation of research participants in the process of knowledge production.

Background

Scientific research plays a critical role in understanding how to prevent gender-based violence (GBV) and HIV (Jewkes et al., 2003, 2009, 2010; Jewkes & Morrell, 2010, 2012) but can be criticized when conducted in precolonized countries for legitimizing and serving interests of the privileged in society and ultimately not solving the social, structural, and health problems of the underprivileged (Kessi, 2017b). For example, Black African women continue to be exposed to high levels of GBV and HIV infection in South Africa. Kessi suggests that social justice is an essential part of research; thus, research and activism should be necessarily intertwined to disrupt the legacies of racism, capitalism, and heteropatriarchy that intersect to shape women’s vulnerability to GBV and HIV infection in African contexts (Kessi, 2017b). Thus, researchers are called to ensure social justice by engaging in self-reflexive and participatory forms of research practice, acknowledging the local context, and safeguarding their representation of research participants (Kessi, 2017b). Representation can be guaranteed if researchers are conscious of not reproducing historical discourse and legitimization of Western thought and research practices. This is consistent with a decolonized framework that has been proposed for social research conducted among people of the global South (Kessi, 2017a, 2017b; Kessi & Kiguwa, 2015). To a great extent, it also concurs with Connell’s (2007a) proposal for knowledge production to credit and be representative of local stakeholders and local contexts in the global South.

While research methods may originate from Western anthropological scholars who explored indigenous cultures perceived to be “exotic,” the discipline has been accused of Western imperialism and efforts to synthesize elitist liberal consciousness with other cultures (Connell, 2010). Local ethnographers have since emerged from precolonized countries in part due to decreased funding for overseas studies and the recognition of biases such as overlooking the Western ethnographers’ own cultural foundations and dimensions at home (Van Dongen & Fainzang, 1998). The growing pool of local ethnographers in precolonial countries also amplifies the importance of improving the status of local knowledge production. This can be achieved by acknowledging Southern scholars’ arguments for the reflection of alternative knowledge systems borne out of the cultures, beliefs, and values of people in precolonial settings in research and for such knowledge to be put on the same scale as Western science (Kessi & Kiguwa, 2015). Doing so would give impetus to the knowledge systems that shape local people’s lived realities. Ensuring social justice in HIV and GBV research can begin through reflections of local scientists who lead knowledge production and an exploration of how they represent research participants during data collection, analysis, and interpretation of the research data.

In South Africa, local scientists have played a vast role in generating knowledge about HIV disease (Hodes & Morrell, 2018) and the epistemology underlying risks of HIV infection including women’s vulnerability to GBV and their sexual relations (Jewkes & Morrell, 2010, 2012; Jewkes et al., 2015; Morrell et al., 2013; Shai, 2018). Yet scientists, though focusing on the South African context, borrow from progressive theories emanating from the global North than from those developed locally (Morrell, 2016), and most of the research inquiries have been more quantitative and less qualitative. In Southern theory, Connell (2007a) proposes an emancipatory and democratic process of knowledge production that recognizes the local context and appreciates and legitimizes social theory generated from the global South. This is possible through dialogue, reflection, and engagement, which reflects Southern social thought has as much intellectual power and political relevance (Carrington, 2008). Qualitative research and ethnographies have a potential to contribute to this emancipatory approach. A few scholars have used ethnographic studies to understand research participants’ experiences of GBV research, thus building knowledge about its impact from the participants themselves (Sikweyiya & Jewkes, 2011, 2013a, 2013b). The profile of ethnographers in the GBV and HIV prevention fields is also changing from almost exclusively White, Western scientists from the global North and has begun to integrate Black African scientists who bring local researchers’ perspectives, interpretations, and reflections of the location, context, and representation of research participants to the fore. The approaches to ensure representativity of research participants should in addition provide in-depth reflections on the positions of power between local researchers and research participants.

The position of power of researchers may influence the development of a research partnership with community members, the research process, and its outcomes (Hennink et al., 2010; Mercer, 2007; Merton, 1972). Yet researchers must not be dismissive nor ignorant of cultural constructions of the social worlds under study and must not appropriate “voices” of research participants (England, 1994). Local researchers thus have an opportunity to champion the implementation of emancipatory processes in research, through self-reflexivity, to account for the complexities around their positionality and power relations in the field (Sultana, 2007). The fieldwork stage, in particular, as the space of research interaction with researched communities, becomes a critical channel through which the assumed researcher position is conveyed and a base for learning how to account for the “voices of others” in research. The researcher, being the instrument of translating the voices of others from the field into intellectual formats, thus must engage in processes of reflexivity to be able to clarify aspects of themselves that have great potential to influence the research implementation and findings (Hertz, 1997; Sultana, 2007).

Kessi (2017b) further argues that knowledge production should also focus on shaping the societies in which we live, thus foregrounding participants’ contextual concerns and knowledge including marginalized members of society. In her perspective, this creates meaningful conduits between the academic and lived experiences and mitigates the power differentials that lead to epistemic violence often produced and exercised against those who are researched (Kessi, 2017b). Other scholars share a similar view (Connell, 2007a). However, there is limited scholarship on local researchers’ experiences of conducting GBV and HIV research in marginalized contexts in South Africa. Reflecting on such work would broaden our understanding of how research participants’ contexts and knowledges can be represented. This is of particular importance when navigating women’s vulnerability in a context where their sexual relations are constructed and policed within a patriarchal system, and their vulnerability worsened by their curtailed power to determine and negotiate their sexual identity and expectations in sexual relationships with men (Lebona, 2016), including the circumstances under which to have sex, with whom and how (Shai, 2018).

Critical self-reflexivity enables a self-referent, self-reflective, autobiographical account to deconstruct the dualities of insider and outsider and reveal researchers’ representations of the world (Macbeth, 2001). It also mitigates power differentials between those conducting the research and those being researched (Naples & Sachs, 2000), highlights the dilemmas of fieldwork (van Maanen, 2011), and helps to deepen interpretation and analysis. Since power manifests differently depending on the researcher, the theoretical framework, and the local context, researchers must question themselves at the onset of their work and examine their position of power, which is inherent and inevitable part of a research process (Caprara & Landim, 2008; Hertz, 1997; Reinharz & Davidman, 1992; Wilkinson, 1988). However, there is a dearth of literature on the use of autoethnography in GBV and HIV prevention in South Africa despite extensive research in these two fields. Scholars claim that the autoethnographic approach, both as process and product (Ellis et al., 2011), constitutes an analytic integration of the researcher with the research aim, theoretical framework, and discipline of inquiry guided by the personal experiences and values about the cultural context (Atkinson, 2006). Researchers and participants are able to explore past and present experiences to gain self-awareness about interactions and their sociocultural effects (Hayano, 1979), through displays of multiple layers of consciousness that connect the personal to the cultural (Ellis & Bochner, 2000). Autoethnography ultimately acknowledges and accommodates the subjective, emotional, and other influences of the researcher on the research in understanding the cultural rather than hiding or ignoring these or assuming researcher neutrality (Ellis et al., 2011).

In this article, I explore my self-reflexivity on my positionality as an insider and outsider researcher drawing on my experiences while conducting ethnographic research in a rural setting in South Africa. Here, I explore my assumptions going into the field, and I present how these assumptions were challenged by the fieldwork process and interactions with research participants. The researchers’ self-reflection on their positionality relative to research participants is a significant first step in conducting credible qualitative research and advancing knowledge production on different phenomena. As a researcher drawing on ethnographic methods, an exercise in self-reflection about power is a useful tool to deepen ethnographic analysis and to highlight dilemmas of fieldwork (Naples & Sachs, 2000). However, autoethnography has inherent challenges that include among others, the researchers’ discomfort with “exposing themselves” in the process of finding and expressing their voice (Forber-Pratt, 2015) and representing the voices of others. I do not claim that this article is an “autoethnographic” or “native ethnographic” account as my doctoral study was not strictly ethnographic but drew from the ethnographic approach. My self-reflexivity is an exercise of exposing my experiences and learnings from the process of immersing myself in a familiar setting for the purposes of an academic inquiry. In this, I also briefly explore the considerations I made while attempting to optimally represent research participants’ voices during the writing up of the doctoral thesis.

Method

This article draws on my self-reflexivity on my researcher position during a qualitative research project for my doctoral studies. The study aimed at exploring constructions of femininity in the context of sexual relationships and its intersections with women’s vulnerability to GBV and HIV infection among women living in the rural Eastern Cape province, South Africa. I used ethnographic research methods, namely, overt participant observation, in-depth interviews (IDIs), and focus group discussions (FGDs). I undertook this research in an undisclosed, small rural town in the Mpondoland ¹ district of the province from May to August 2011.

The study setting was an indigenous social context of Mpondoland, a region and its people noted for having been one of the last ethnic groups that succumbed to (Hunter, 1933; Hunter Wilson, 1979), European colonial conquest in South Africa, and revolted against it (Mbeki, 1964). The Eastern Cape province is currently one of the poorest and least resourced provinces in South Africa (Statistics South Africa, 2017). Poverty in the province affects mostly rural households and women (Westaway, 2012). The 2011 Census data showed a 37% unemployment rate in the province (Statistics South Africa, 2011). The province has a very large HIV/AIDS epidemic (Connell, 2007b), and the GBV epidemic is also very high. Being the most rural part of the Eastern Cape, these socioeconomic and health conditions are greater in Mpondoland. The region faces the triple challenges of poverty, inequality, and unemployment. The unemployment rate was 44.8% in 2016, while the majority of people were dependent on social grants as their primary source of income with 23.08% of households living on under R2 500 (ZAR$168) per month (Eastern Cape Socio Economic Consultative Council [ECSECC], 2017). Educational outcomes are poorer in this region as 64% of schools are overcrowded (Rural Health Advocacy Project, 2018). 18.2% of the population in the district surrounding the study setting itself have had no schooling at all, and only 4% have completed post-high school education (ECSECC, 2017). The rural context of Mpondoland is also marked by the population’s difficulties to access health care and economic opportunities as well as their stronghold on their cultural roots, beliefs, and practices.

As a participant observer, I lived there for 2–3 weeks at a time and returned home to my family in Johannesburg and work in Pretoria intermittently. A local chief introduced me to the key informant to assist with community mobilization and access to research participants. I paid her R35 per hour at 30 or less hours worked per month. I also conducted two rounds of semistructured IDIs with 19 women focusing on personal backgrounds, aspirations, experiences, and notions of womanhood as well as sexual practices and relationship experiences. Over 40 women participated in FGDs that covered community constructions of womanhood. During that time, I took detailed notes of my interactions and experiences. For the duration of the study, I lived in a village near the study site that consisted of another village and a small neighboring township. The villages were made up of multiple homesteads scattered across green patches of land, and most homesteads had one or more buildings of different styles including a hut, a two- or three-roomed “flat,” and/or a modern house like houses found in townships such as Soweto (Wikipedia, 2019b). Having many different houses in one homestead was an indicator of relative wealth of families. The population in the village and the township is nowadays largely dependent on state social security having previously engaged in subsistence agriculture. Child social grants are the most common contributions to women’s livelihoods in the province at R455 (±$30) currently.

To ensure rigor in my study methods, I complemented IDIs and FGDs with overt participant observation to optimize my understanding of how womanhood is conceptualized, performed, and legitimized through personal participation in the lives of women (Paul, 1953), thus going beyond obtaining in-depth insights into their perspectives and experiences. Anthropologists often live longer in the field because they are unfamiliar with the setting and do not know the language. However, knowing the study setting mitigated some of the potential challenges of acclimatizing with the research context and participants. I also used the field notes as a self-reflexive tool to examine subjective insights from my experiences and interactions with research participants that could influence my interpretation of events and representation of research participants (Caprara & Landim, 2008). These research approaches ensured rigor and credibility of data (Davies & Dodd, 2002) and helped to attain some balance between being an insider and outsider researcher during data analysis. Ethics approval was obtained from the University of the Witwatersrand Human Subjects Research Ethics and the South African Medical Research Council. The findings of the doctoral thesis are documented in a monograph published at the University of the Witwatersrand Library (Shai, 2018).

Limitations

The research was undertaken in one community so that the findings may not be generalized to all the other settings in the Eastern Cape province or South Africa. I may not have seen as clearly what others from outside could pick up or draw from my interactions with research participants because the study site was familiar to me. My research focused on heterosexuality, so I could have missed opportunities to explore womanhood from nonheteronormative perspectives. Qualitative research also lends itself to social desirability bias, and this could have influenced research participants’ responses and relations with me in the field whether intentionally or unintentionally. However, this article seeks to expose some of the potential strengths and weaknesses of my dual positionality, which allows some interrogation of the experiences and interactions I had with participants.

Findings

The Insider–Outsider Identity

I conducted this study from the insider–outsider position and was able to reap the benefits, but I was also exposed to the challenges of either position. My insider status was shaped by a shared heritage with research participants in the study setting through my kinship. My mother was a daughter of a chief in Western Mpondoland and I grew up in a town near the study site, so I have in-depth knowledge of the Mpondo culture. I had been residing in the city for some 20 years but had limited contact with the community back home except with my maternal extended family. Growing up among chiefs and having both parents working—my mother a teacher and my father a clerk from the abaThembu clan—introduced a class difference with other people my age in the community, which could have influenced my Mpondo cultural understanding and expression. In 2000, I assisted an ethnographer working in my township, and she relied on my “insider” position to introduce her to the elders, chiefs, and traditional courts and interpret for her during meetings and interactions with young people (Wood, 2003). I have since played an outsider role as a professional researcher in HIV and GBV research (Jewkes et al., 2002, 2008).

Building and Fracturing Relationships With Research Participants

My relationships with participants revolved around the key informant, Nondu, her elder sister, Tembela, their friends and neighbors who lived and/or worked in the locality. Nondu started as a supportive key informant who organized daily visits to potential participants with expediency and engaged in numerous conversations with me about the life of women. I forged relations with Nondu’s social network of women aged 26–39 years. Many of her friends, who I have categorized as the townswomen, ran small enterprises whether selling fruit and vegetables, secondhand or stolen goods, or ran their own hair salons. They shared a common trading space in a shipping container located close to the center of the small town. The other townswomen worked at the government offices, local schools, and retail stores but visited around lunch or after work. I frequently met with many of Nondu’s friends around the container almost every winter morning of my fieldwork, sharing breakfast of amagwinya ² and tea with an old paraffin stove providing heating. A small kitchen housed in a similar shipping container nearby catered breakfast and lunch to commuters and civil servants, and the middle-aged woman who ran this kitchen often walked across to us with a large steel kettle to pour us coffee and tea. I established a daily routine of visiting Nondu, and the other women, from morning to evening.

In the beginning of fieldwork, my interactions with participants outside of the interviews were professional and yet friendly. Nondu, like other women in the study site, was curious about “what I was doing exactly.” At the end of Week 1, I had been introduced to a number of her friends, family, and neighbors one of whom was a key politician called Ellen. By Week 2, I had completed seven interviews among Nondu’s neighbors.

Community women’s perceptions on my role in the field were fluid and oscillatory and some recognized my dual positionality. At first, I was identified by the dual position and introduced like this: “a young woman whose parents come from this area but lives and works in Johannesburg and is here to study the lives of women in this village.” Soon research participants saw my role as a focus group facilitator and interviewer, which occurred in the first 2 weeks of my arrival in the field. As I got integrated into daily village life, I was seen as an investigator who was curious to know about women’s life experiences in the village. My inquiries into women’s lives were unfamiliar and marked the difference between me and the other young women I hung out with in the community. The elders also thought I was a social worker whose role is perceived as an outsider coming into the village to help. Over time, trust was established, and research participants began to share and discuss things with me as if I was one of them. This manifested during intimate discussions with married women, and with the young unmarried women who shared how and why they engaged in various risky practices with men, some of which included the practice of “eating men.” For instance, some women openly disclosed how they sometimes tricked men into paying damages for nonexistent children and asked their friends to stage phone calls pretending to be parents demanding child support. This was among a variety of transactional sexual practices reported in the study (Shai, 2018).

During fieldwork, I built good relationships with women in the study, I became even closer to some women than I was with Nondu, the key informant, though I spent most of my field time with her. I befriended Phendulwa, an older divorced woman, with whom I had fun and sad conversations and exchanged gifts. She was sick with TB and persistently tried to acquire a disability grant with great difficulty. Her travel to the government offices to enroll for the disability grant brought us together often as we usually met on a 5-km stretch of road on the way to or from town. The combination of her illness and exposure to dusty roads so soon into her diagnosis deeply concerned me, so we usually discussed her health and her life experiences. Our chats included her husband’s abusive and multiple sexual partnering behavior prior to him abandoning her for more than 10 years, his recent visits, expectations of sexual intimacy and her refusal, and the prospects of her dating again. Phendulwa represented women who do not necessarily fit into the model of ideal femininity in the community and exemplified the social notion of a “townswoman” as she had lived in Johannesburg townships for years selling alcohol, was divorced, and raised her children alone while dating married men (Shai, 2018). As shown in my thesis, other women were wary of her fearing she might take their husbands (Shai, 2018). Our interactions outside the interview helped to clarify my knowledge about the social positions of unmarried, divorced, or widowed women.

Nondu had introduced me to a group of friends: 12 unmarried, married, and divorced female friends many of whom I met at the container. These women struck me as independent, self-sufficient, and innovative. Being exposed to diverse personalities was relevant for my study to understand constructions of femininity. Most were politically connected, if not actively involved in the local political parties. They had endured disadvantaged backgrounds and tough life challenges, such as single parenthood, intimate partner violence, divorce, financial strife, breakups (not necessarily in that order), and seemed to have employed a survivalist approach to life. I also witnessed them being brash toward others and participating in relationships with men that put them at enormous risk such as partying with men newly known to them.

The power inherent in political or traditional leadership is not necessarily congruent with wealth or privilege. Ellen was a prominent local political figure in the small town and a friend of Nondu, her sister, and the other townswomen. She did not own a car and often relied on comrades for transportation. Being mobile put me in a position where I was seen as the local driver. Despite having explained that ethics surrounding my role as a researcher did not allow my explicit involvement in “politics,” though I was keen to hang out with Ellen and Nondu’s other friends, about 2 months into my fieldwork, Nondu and Ellen co-opted me to transport them to a village some kilometers from the study site to donate clothing and household stuffs as part of Ellen’s political role. This was posed as an opportunity for me “to meet people for my research,” even though the village was far removed from the study site. Around that time, I had come down with the flu and was staying indoors. So, I declined to accompany them on account of being unwell. When Nondu called trying to persuade me to change my mind, she revealed that the community visit was part of Ellen’s and her political party’s campaign trail to garner vote at the upcoming local elections. I reminded Nondu of my nonpartisan principle and how participating in their campaigning was unethical. She did not take my refusal well and threatened to sabotage my research. Her friendship with Ellen made me believe that she was capable of overturning the study. I challenged her and asked for an official letter from the chief and ward councilor, who had approved the study, stating the termination of my research project. But the threat to close my fieldwork was disturbing. Discussions with my supervisor, and family, were helpful in solidifying my resolve to stay impartial and to challenge Nondu. A termination letter or a corresponding phone call did not surface and led me to doubt whether Nondu had conveyed my demands to Ellen.

This experience brought up some important considerations for researchers in the field: the degree of power that a community under study may have over researchers, how divergent agendas between researchers and community members may lead to such impasses, and whether it is important to uphold principle at the risk of conflict with the community. This incident complicated my dual position, first, as a relative of the traditional leaders in the area, second, a “friend” of a politician (e.g., Ellen), and third, a friend with other research participants who identified with other political parties. I had also explained to Nondu my apolitical stance that is linked to my organization, and she was aware of potential backlash the study could face should any differences arise between politically aligned traditional leaders, ward councilors, and research participants since municipal elections were imminent. Her actions here implied that while others’ perceptions of my research position in the community may have oscillated or evolved over time, Nondu continued to see me as an outsider from whom she could demand to use my resources and threaten me when I refused. I did consider she may have initially thought of me as a friend save of saying the threats quickly surfaced when I did not relent.

A few days before the impasse with Nondu, I had planned to join her and some friends at an “after-party” following a political gathering to elect party political leaders. As is customary in the Eastern Cape, perhaps all of South Africa, after-parties usually conclude big functions such as weddings, funerals, and apparently now party-level elections. The after-party was an opportune time and space to learn about the municipal developments firsthand. After being threatened by Nondu, I also figured it would provide useful information in case I needed to reapply for permission to work in the area, but I was hesitant about going there. I was persuaded to go there after talking to Tembela who despite my despondency in lieu of Nondu’s threats insisted I attend. On my arrival, I was informed that Nondu had been given a “talking to” by the other women. Ellen herself seemed shocked when I relayed Nondu’s threats and attributed it to Nondu’s bullying tactics. The matter was diffused within a few minutes of my arrival at the after-party, and no real discussion was had between Nondu and me because I was told to forget about it and move on.

My relationship with Nondu suffered a setback after this incident, though it contributed to a shift in her friends’ perceptions of my researcher positionality. Challenging Nondu gained me a new degree of respect. The women who were aware of the incident supported my decision to stand up for myself. Not too many people got away with challenging key political figures, they said, and for “an outsider” to do so demonstrated a certain degree of agency. The accolade was also somewhat self-serving as some seemed to be settling a score with Nondu through me. Their comments revealed they considered her defeated, and me, untouchable. Some confessed the incredulity of my response because I came across as professional and maybe even “a cheese-girl.” ³ These attributes had initially set me apart from other women in the study. Being seen in another light shifted the attributes they ascribed to me. Soon afterward, some of Nondu’s friends started calling me Sfebe, ⁴ a term they called one another (Shai, 2018), and this indicated that I was now one of them. Sfebe is a derogatory term, but for these women, it also represented chutzpa and strength and being called Sfebe meant I was now in the inner circle!

I maintained pleasant but somewhat distant relations with Nondu. I missed her although I still held deep reservations about our relationship. She continued to assist with organizing the second interviews with women perhaps because she was getting paid for it. Her sister had played a key role in diffusing the tensions between us, and upon reflection, we all used her as a buffer in a fragile relationship.

Self-reflection on my conduct in the field revealed that I had been accommodating to research participants and had never needed to be oppositional—until the Nondu incident. I had taken conscious decisions to be kind and approachable and promoted myself as a confidante to many women. Playing down my outsider status seemed necessary not to set myself too much apart from a people whose lives I was there to observe and appreciate. I am of the opinion that by giving an interview, informants are letting us in to witness their lives. I felt that I could not do that if I obviously and intentionally upheld expectations that would seem judgmental or aloof. For example, I prefer food with a lot of vegetables though I love meat. During fieldwork, I struggled to cook evening meals regularly because I stayed out late into the evening and bought meals from the local vendor. My despair at how much oil, or how small the vegetable portions were, was perceived as being too “particular” (ukucekisa). I had grown up with similar cooking methods, so I did not criticize, but often found myself swapping meat (usually my favorite item on the menu) for more vegetables. Of course, being avid meat-eaters, my new friends usually enjoyed this at my expense. I began to appreciate the differences between mine and informants’ views about tasty food and nutrition.

In Mpondoland, meat is a highly valued cultural food that is present at every communal occasion, but not all families can afford it on a daily basis. Thus, every opportunity to consume meat is often appreciated by those who cannot afford it. While research participants observed my eating habits as being particular or preferential, they did associate it with my socioeconomic status much as my modernity, but this was not unique, as other women in the community were wary of unhealthy food and participants had previously engaged in a number of weight loss diet plans. The topic of food also came up among young women who expected a braai instead of cakes during their focus groups. So, I handed over the refreshment planning to Nondu so as not to offend participants.

Despite my familiarity with the culture and language, I was inadequately prepared for what it truly meant to live and enjoy the life of a contemporary Mpondo woman. Among Nondu’s friends, I got to spend more time with her sister Tembela and her friend, Anele, from whom I began to appreciate a great deal regarding the multiple ideological and practical influences on the concept of “womanhood.” The confrontation with Nondu did not only influence my identity in the field but also my research process. Since earning a new level of respect among the women and being seen to have some of their traits of a “strong” woman, many shared personal stories about having stronger personalities over our regular breakfasts. Listening to their accounts of bravery and courageous selves, I learned more about the underlying and sometimes undisclosed aspects of women’s lives, perhaps due to moderation by having recently entered their lives and their fear of being judged during the interview. I became privy to new insights about some of the women including the risk-taking practice of “eating men” (Shai, 2018). I had heard about some of the risk-taking behavior from Nondu or surmised from women’s reactions to certain topics during casual chats or interviews. Nondu was always curious about what her friends shared in interviews and sometimes “outed” those she knew to be secretive indicating that they had valuable stories to share. I did not encourage such behavior and always insisted on confidentiality between me and my informants. Sometimes I directly pursued the merits of Nondu’s distrust of some women during the second interview and found that some women had much more to share. It was at these moments that I appreciated Nondu’s strong character.

At the time of my impasse with Nondu, I had begun the second round of interviews with about at least 5 of the 18 women I had interviewed thus far. These revelations about the “other side” of femininity influenced changes into my scope of inquiry. I started to ask more direct questions about what “a woman with timing” (a calculating woman) was like and more about women’s own sense of bravery. Many women were receptive to respond to these topics: multiple sexual partners, transactional sex, bisexuality, and practicing witchcraft. I also began to notice the contradictions between what women purported to be, my initial assumptions, and my analysis of their interviews. Those women whom I had assumed to be despondent seemed to be more resilient, and those who were hard-core, were in fact timid, lonely, and sad than I had originally thought. These observations informed me a lot about the journey women go through in negotiating, establishing, and expressing their womanhood (Shai, 2018).

Sexual Objectification as a Woman Researcher

One of the challenges of daily interactions with local community members and family/friends of my key informant was my encounters with men. Propositioning women seemed to be a popular practice among men I encountered during fieldwork. Some of their attention was directed at me. Two men stood out, and they presented numerous challenges regarding my own femininity. Themba was the municipal worker who helped me access the ward councilor, while Biza was a combi taxi driver. Both were friends to Nondu, and we had regular interactions. Themba, who also happened to be married, was relentless in his pursuit of an affair with me, but his marital status did little to deter him and neither did mine. In fact, it seemed that neither male nor female married status effectively discouraged infidelity. The most fruitless excuse though was my suggestion that my work prevented me from dating authority figures in the field. Nondu and her friends were aware of this man’s attempts and observed this pursuit with keen interest. They joked about how dating Themba would solve my problems with sporadic supplies of water at the house where I was renting. While it seemed transactional and I was a little embarrassed by this, I did take him up on his offer to facilitate water delivery in my area. Even when I hadn’t disclosed the constant phone calls, Themba tried to use Nondu to get to me, and it became a sorry but hilarious experience at every chance meeting we had. He had expected I would be more interested in him after helping me. Suffice to say that his interest lasted only 3 weeks, and I never heard from him again.

Biza was much younger than me but had a hefty built. His close friendship with Nondu apparently intensified due to my presence. Biza’s continuous proposals to have an affair with me tested my credibility with Nondu and the other women. He was younger, attractive, and perhaps fitting the profile of an ideal man. He came from a wealthy home and owned his own taxi. His taxi business often took him on trips to Johannesburg and that made him seem worldly. Nondu and friends often sought details about Biza’s interest in me and jokingly referred to him as my “Ben 10.” ⁵ They used me to get him to buy us lunch, alcohol, or to sponsor trips to the beach and constantly put me under pressure to party with them and Biza. In the beginning, I obliged so I could learn more about my informants. However, as observed in my thesis, being used as bait and getting a man to finance our fun somewhat gave the suitor leeway to intensify his pursuit (Shai, 2018). So, I made attempts to minimize my interaction with Biza and changed my research tactics. In my youth, I had learned that certain men don’t take no for an answer and often find ways to counteract women’s rejection and such men’s reaction to rejection by women could be unpleasant. Biza was that sort. A simple “no” received many arguments to justify and nullify potential deterrents to having an affair. He blurted out his oblivion to my age or marital status and simplified his request. At one time, he proposed that he become my Weekend Special as per one of Brenda Fassie’s hit songs. ⁶ So, I regarded our encounters as “sparing” matches, and when I was aware I would meet him, I prepared for argument, being ignored, or being insulted for being too prudent.

Although I was aware of going into fieldwork, as any other day, that I could be propositioned by men met along the way, I had not factored in its intensity during my fieldwork. My interaction with these men also brought to focus the sexual context in research settings and reminded me how pervasive women’s objectification can be (and of how little regard was given to), whether women were interested or not. In fact, being uninterested made some men relentless in their pursuit. This experience also resonated with my observations and women’s reports in interviews about how sexual partnerships are formed; men’s perspectives on women’s desirability, respectability, and marriageability; the deeply entrenched use of sexual language in the Mpondo culture; and how comfortable women felt talking about sexuality in face-to-face formal and informal conversations with me (Shai, 2018).

Lessons About Women’s Self-Image and Sexuality

Biza’s frequent and candid sexual talks and ranting were inadvertently beneficial to my fieldwork. He would confront Nondu and her friends who were also my informants about their sexual exploits, some of which had not been mentioned in interviews. Though I could not ask the informants to explain Biza’s accusations against them, the information helped me piece together some of the discrepancies and clarified the meanings behind the hesitations, laughter and jovial attitudes toward certain sexual partners, and many other nuances I had observed from interviews. I also learned that with some women, the accounts they shared with me were merely a portion of the whole story while others, portrayed versions they felt would ameliorate their self-image. Meanwhile, Nondu had also been gossiping about her friends’ behavior, constantly filling me in on their sexual lives. She was rather perplexed at me not sharing what her friends told me in interviews, as I neither denied nor confirmed what I had been told. During one of Biza’s ranting episodes about friends’ sexual escapades, Nondu also confronts them to divulge specific details about male partners, some of whom were not mentioned during interviews.

Through Biza, I learned that participant observation as a research technique is valuable in that it gives the researcher complete exposure to a community and thus access to many streams of information (including sensitive information) through engagement in daily life. Sex is inherently a private area and depending on the relationship between the researcher and research participants during qualitative research, building rapport and disclosure of sensitive information are dependent on time, trust, and nature of the relationships established in the research community. However, an insider with access to gossip may be acquainted with multiple other versions, some of which individuals might have concealed in interviews in order to “save face.” Being an insider participant observer during my research gave me access to layers of “secret” information from the world of gossip, which shaped my understanding of important issues in my research. I came to understand that self-image was everything to women and did not always correlate well with actual behavior, and the disparities between self-image and actual behavior need to be analyzed in-depth. I did not take women’s reported behavior at face value anymore but attempted to merge with their self-image contained in how they communicated about themselves, what they hid, what they revealed, their desires in life, and the image I observed portrayed during our interactions.

Further reflections on my decisions in the field revealed my attitudes toward the role of communal alcohol drinking during fieldwork. My previous experiences of ethnographic fieldwork had taught me that drinking was the most common pastime in some communities, and though I had been part of that experience, I was conscious of the potential risks such as drunken driving, car accidents, and sexual assault. When I realized that informants wanted me to buy them alcohol I questioned myself about the ethics of doing so and drinking with them. Informants were ignorant about my concerns regarding undue influence, and my decision not to oblige was somewhat problematic. This led to tensions with some vocal and confrontational informants, but these did not last very long. When I supplied cake and tea at the first FGD with townswomen, the group protested so much. They negotiated I provide braai meat instead and suggested those who wanted alcohol could buy for themselves. Some informants believed I was a prude, having never seen me drinking alcohol, and were dismissive of my concerns about drunken driving. Instead, I got into a routine of providing daily morning coffee and amagwinya for Anele and her friends at their workplace. As a parting gift though, I gathered a few of the friends I had made in the study and took them for a drive to the beach, where we braaied fish, sang songs, danced, and I drank a bit of wine while they drank their favorite, Storm ⁷ ! They believed they had experienced the true me, the fun loving, loud, iyilo (bad singing voice), and a good dancer. I found it interesting that these women believed one’s truest self comes out during their drunken state. But it seemed I had passed some kind of test, as I realized that they may have felt that they had stripped me of the city-life researcher/professional pretentiousness and got to experience the “real” me.

My experiences with informants showed me that I had not adequately prepared for the extent of personal investment I needed to make for the participant observation to be a success. I also appreciated that there were some missed opportunities on subjects I could explore with women. Most of my time was spent with the townswomen, perhaps at the expense of my relations with the village women. Generally, discussions with all women focused on challenges women experienced in the family, community, and sexual relationships. As a researcher, I listened to women’s stories and asked “a lot of questions” for clarity. As a woman, I was also expected to contribute and sometimes did so. Partners, intimate partner violence, and partner unfaithfulness (of both women and men) were the most topical issues. My feminist responses to experiences women raised sometimes worried me, for I was unsure of whether I would respond in ways that undermined women’s own responses. So, in the beginning, I stayed as neutral as possible, but this did not last as some informants commented on my being too detached. On the other hand, the townswomen expressed their curiosity about my life experiences and asked their own questions, some of which were more poignant than the village women’s inquiries. Ruby, also in her 40s, was a married woman and a friend of Ellen’s. She was constantly preoccupied with acquiring men, to the extent that friends often “organized” men for her. Her sickly husband lived in the city with their children. Ruby’s unfaithfulness was attributed to multiple factors, the loneliness associated with the distance from her husband, his illness, or that she found him unattractive. Her inability to remain with one partner was often remarked on, and others believed her to have tendencies of impene (insatiable libido). Anyway, one night she got bored with my impartiality and directly asked about my marriage. Now it is difficult to do so without sharing specifics, so it was pretty uncomfortable. Though I shared some of my personal experiences, I downplayed them quite a lot in an effort to portray a less interesting life. On the other hand, married women living in the village were less intrusive and easier to relate due to the less personal nature of our talks, which focused on societal expectations of women.

The townswomen appealed to me because of the multifaceted aspects of their lives and identities. I attempted to retain the presence of the married sample of women I had interviewed and visited them at their local farming project. It helped me to increase the interaction, as I was later introduced to other experiences of married women. The most notable being the domestic violence episode at Lizeka’s house. The incident occurred just 2 weeks before the end of fieldwork. Nondu and I were wrapping up the day’s work when we were called to intervene at Lizeka’s. Her husband was physically assaulting her, and she was also fighting back. Community women had responded to her cries, the sound of which we heard from the other side of the valley. Women belonging to a forum set up to respond to domestic violence cases were the first on the scene. Lizeka’s husband was given a nickname that reflected the community’s perception of a combative masculinity. For the purposes of this article, I shall refer to him as Gonondo. The forum women confronted Gonondo as they stormed into his house. I was also among the women summoned to provide a helping hand. In what seemed like an effort to reconcile them, both Gonondo and Lizeka were given the opportunity to explain what had happened. They gave conflicting accounts, but the gist of the story rested on Gonondo assaulting a child upon his arrival at home that evening, and Lizeka hitting him first in anticipation that she was next. Confronting Gonondo did not appear to be helpful in the situation as he did not take responsibility for his actions including emotional, verbal, and physical abuse of his wife and children. The forum women tried to talk him down and warned him about the consequences of physical violence, but he remained defiant. Lizeka already had a protection order against Gonondo but was hesitant to get him arrested again. He also insisted that Lizeka had hit him first, and he hardly had a chance to touch her. Gonondo also joked about his need to shake things up, ahlokohle, just to create fear and respect from his wife and children. I can recall his facial expression that portrayed his cavalier attitude about his abusive behavior. Perhaps it was a defense mechanism as well, not to appear to be perturbed by having a rondavel surrounded by outraged women. I was also asked to talk to Gonondo to dissuade him from being abusive. I tried unsuccessfully. Lizeka has also told us she didn’t want to get him arrested. I recognized this as the trap many women in my family had fallen into and on which their partners relied as they continued to abuse their wives and children. I made referrals to multiple sources of help, though I suspected no concrete actions would be taken. The forum women also attempted to appeal to Gonondo and kept referring to the multiple visits they had had at his home for the same problem.

A few days after this incident I discussed it with some of the community women who were at Lizeka’s house that night. It was most revealing to learn that they blamed Lizeka for “provoking” her husband to anger and suggested it was in her nature to be confrontational. Their interpretation was astounding and demoralizing to me as it contradicted their coming to her aid when she cried for help. I could not confront them about this contradiction but merely realized that while it might seem obvious that partner violence is a terrible event that should never be experienced by anyone, some people including women believe it is justifiable and usually blame women for it. I realized that many abused women, like Lizeka, do not report abusive husbands to the police, so they will not be blamed for destroying their lives. The feminine attribute of being constructive and not destroy rang true for Lizeka and she wanted him to be stopped (Shai, 2018), but not through incarceration. Later, when I asked why she did not pursue the protection order, Lizeka mentioned her despondency with the criminal justice system as previously Gonondo was only detained overnight, and Lizeka was lectured by the police about how to treat her husband. The lack of support from her community of women also magnified the problem; hence, I understood that although she was not necessarily dependent on her husband’s financial support, she lacked the support system to end the violence.

Dual Positionality and Relationships With Research Participants

The duality of my identity in the field produced complex relations among married and unmarried women. It was easier to negotiate access to elderly women in the village due to my marital status. Sometimes married women are more receptive to other married women, despite age or being an outsider. I was allocated a public symbolic space that was differentiated from Nondu’s (the key informant) who was unmarried at the time on the basis of acquired knowledge about notions of respectability and maturity that are presumably inaccessible to an unmarried woman. This was only possible due to the elevated status afforded to married women consistently with notions that regard marriage as the main conduit to women’s access to ideal femininity (Shai, 2018). On some occasions, it seemed that notions about married femininity and ageism hindered disclosure about sensitive topics such as sex and sexuality among a few older women who were shocked by related questions. The study further revealed that due to deeply entrenched social expectations on women to maintain notions of dignity and sexual repressiveness, some older women were uncomfortable to discuss sexuality issues with younger women and believed it undermined their seniority and dignity and promoted socially inappropriate behavior among women (Shai, 2018). Despite resistance from a few older married women, engagement with other older married women was extremely successful due to their general openness and fun way they approached discussions about sexuality.

Drawing on my knowledge of community expectations toward young women, I paid attention to how I presented myself. I dressed modestly, wore mostly skirts, and avoided pants during community functions, such as funerals or ancestral ceremonies in accordance with community expectations of married women. When I inquired about the acceptability of my dress during a focus group, participants said that wearing pants “depends on your husband. If he likes it then it’s ok.” This response was unexpected and suggested a possible transition in the community mindset and some acceptance of modernity. I also drew from personal and research experience of interacting with elders and portrayed mannerisms that denote respectfulness. This also applied to how I spoke, phrased concepts, and interpreted misconceptions or ideologies that seemed off course. Having worked in rural settings prior to this fieldwork, it was often advised that “fieldworkers must go ‘down’ to the level of the villagers,” to simplify concepts and minimize any perceived socioeconomic differences.

Expectations among unmarried and younger women who lived in town produced different experiences compared to older married women in the village. Unlike older married women, the group of younger unmarried women I interacted with in town had more liberal expectations of my conduct in the field. Sharing breakfasts with them at the container provided me with opportunities to access young unmarried women’s unguarded accounts about femininity and sexuality, sexual histories, strategies, and other experiences. These women were characteristically ambitious, street smart, and financially savvy. Interactions with the women outside of their workspaces involved hanging out at their homes, and a few times drove around looking for u-line (direction where the party is at), and drinking. My association with the women was challenging at first; it was obvious that I was an outsider, I asked too many questions about their lifestyles and some of the seemingly insensitive language they used toward one another, including calling one another Sfebe. Yet my familiarity with the setting, being of the same age group, speaking the language, and open and nonjudgmental attitudes toward their challenges and experiences helped to bridge that gap. As we grew closer, we had more in-depth conversations about what it was like to be a woman. Conversations often centered on sexual partners, present, past, and those fancying or being fancied; increasing access to income and child social grants; children and their fathers, cheating partners, and cheating on partners. These topics came in no particular order and were solely dependent on mood, or burning issues women brought that morning or afternoon. While some of these topics were sad, there was usually laughter and strategizing to get women out of the messes.

Some researchers are concerned about the social dynamics brought about by an insider researcher (Ganga & Scott, 2006), wherein differences between the researcher and participants come into focus as a result of shared cultural knowledge. “Diversity in proximity” is where insiders are better able to recognize both the ties that bind, and the social fissures that divide, researchers and participants. Differences in my language use, dress sense (youthful compared to my age-peers), “driving a car that I probably owned” and presumed higher education, people I knew and my relations to Mpondoland likely influenced community perceptions about me in the field. At the broader social level, community members exercised a shared social principle of openness to visitors. Normally, visitors to the area ran developmental projects, and despite my work being vastly different from those projects, the community hoped my presence brought about some kind of betterment.

Negotiations with participants highlighted mild and serious issues of diversity, and these challenged my resolve and confidence in the field. Focus groups were revealing of the extent of agency women exercised within their own spaces and outside. I had overestimated the difference in marital status made to conversations despite my background in sexuality research. This was the first time I was confronted by the notion of “blindness to aspects of our culture” (Fulu, 2009) and tendencies to take our experiences or backgrounds for granted. For instance, I was not surprised by the few older married women’s resistance to sexuality talk in the FGDs and interviews but had not been prepared about the extent to which the Mpondo culture allows for open discussions about sex. The study further demonstrated that conversations about sexuality were commonplace among adults and the language could be quite explicit. This posed challenges in moderating the FGDs as they succumbed to often explicit and hilarious sex talk among research participants regardless of their age or marital status. What were meant to be once-off focus groups ended up being two focus discussions with each group.

Discussion

This article demonstrates my self-reflexivity from the insider–outsider researcher position, the challenges my dual positionality introduced, and how positionality shaped my experiences and decisions during data collection in the rural Eastern Cape province, South Africa. There were multiple advantages to being an insider researcher such as having a shared cultural, linguistic, ethnic, national, and religious heritage with research participants (Ganga & Scott, 2006) and in-depth insights into concepts, meanings, and experiences of participants thus producing even richer data (Dwyer & Buckle, 2009; Merton, 1972). The outsider position enabled me to realize gaps in my assumptions and meanings I attached to constructions of womanhood, for example, I learned that modernity was accommodated among villagers’ expectations of a married woman through a patriarchal reference system was maintained. The unmarried femininity was more complex, that is, marginalized due to perceptions about not fitting into the ideal married femininity but resourceful in ensuring survival and resilience in their complicated sexual relations with married men (Shai, 2018). However, my attempts at exercising political impartiality (Merton, 1972) introduced threats from my key informant. The threat itself was a complex challenge against my dual position. The key informant had violated the social expectations of ubuntu toward “outsiders,” and as a researcher, I likened this to a common behavior of retaliation that can be seen in the competition among women and tactics some women used to access the resources of others (Shai, 2018). By declining, I thus knowingly participated in the standoff as an insider to safeguard my interests of detachment as an outsider. My personality and response could have jeopardized my fieldwork.

The ability to achieve objectivity can be tricky for local researchers, as sharing cultural, linguistic, and ethnical similarities with research participants may not necessarily qualify them as insiders per se due to significant structural differences such as a higher socioeconomic and educational status. In my case, my assumed knowledge of the local context was a potential hindrance to a real understanding of research participants’ lives, challenges, and motives for their interactions with me. This changed as my understanding of the current context and lived experiences of women gradually developed and research participants began to demonstrate trust toward me over time. The evolution of my interactions with participants instigated this change, as they challenged my assumptions with the reality of their lived experiences. This necessitated a shift to apply the concept of the “space between” (Dwyer & Buckle, 2009), which is a multidimensional, oscillatory, and fluid positionality to strike a balance between the insider and outsider positionalities to allow shared attributes between the researcher and participants to mitigate the differences brought about by the researcher’s outsider status while maximizing on commonalities inherent in the insider status. The “in-between” status developed as my empathy for the women’s struggles, choices, and behavior was enhanced over time and I embraced both the shared experiences and structural differences between us. This eased my experience of their inner circle and improved my learning of the complexities in their risk for GBV and HIV infection. For instance, allowing myself to be “used” by research participants to benefit from men who financed fun broadened my interpretation of the dynamics of transactional sex in which many women in my research had been involved without judgment or the need to sanitize the tactics women utilized in these sexual exchanges with men (Shai, 2018). I also found myself occupying the space between regarding the violent incident at Lizeka’s house. Though I believed in reporting the incident to the police, I quizzed myself for not insisting she did so. My reflections led me to a realization that I was actively identifying with her lack of agency owing to the criminal justice system and the community’s failure to support her and had chosen a subjective position of providing the emotional support Lizeka lacked at the time.

This article also highlights the potential gains of engaging in research in one’s “local” community, but the insider position is earned and attributable to the researcher by the research participants. The process of earning the insider position from participants can be intensive but is necessary to gain trust among participants exemplified by being brought into the inner circle where aspects of women’s lives are shared covertly. Being brought into the inner cycle was markedly beneficial to the research process, as it enhanced my objectivity to understand among others the motives behind research participants’ risky choices and behavior and apply this in data collection and research analysis. I was encouraged to balance my insider and outsider positions by appropriately conveying the voices of research participants in ways that respectfully represented their ideas and reflections on their lives and yet analytically interpreted how their constructions of femininity intersected with women’s sexual risk behavior and exposure to GBV and HIV infection, for instance (Shai, 2018). This was done with much care to ensure fair representation of the local setting and its contemporary history and accounted for the prevailing patriarchal gender order shaping gender relations and women’s access to income and resources. Further, such understanding also explained how and why certain constructions of femininity and high levels of risky practices posed GBV and HIV risk for women and emphasized the need for structural interventions that embed gender transformation to mitigate both GBV and HIV risks.

A Summary of Lessons From Self-Reflexivity

I drew some important lessons from my self-reflexivity exercise:

An insider positionality does not imply an automatic understanding of the underlying meanings that research participants attribute to certain phenomena. How we deploy fieldworkers needs to guard against taking research participants’ lived experiences for granted because of familiarities with the research context as that may limit the researchers’ ability to deepen understanding and may see research only skirting the surface in addressing public health problems such as GBV and HIV.

A balance between what is familiar and what is not intimately known about research participants’ perspectives and lived experiences is crucial and requires the local researcher to recognize their dual identity from the onset of the research process. Participant observation by local researchers involves complicating both one’s identity and one’s academic interpretations, as local researchers most likely face similar exposures and lived experiences as the research participants. While I could draw from my own experiences when writing the doctoral thesis, I constantly had to compare my experiences with those of participants and draw out where similarities or differences could lead to misrepresentation or undermining of research participants’ voices.

It is important to acknowledge and account for the structural differences introduced by the researchers’ residence and academic positions to the research participants, and the researcher should continuously endeavor to create a workable balance that represents both the academic and research participants’ interests in knowledge production.

Gaining research participants’ trust is a process, and when trust is conveyed (with the help of participant observation) through being brought into the circle of secrets among research participants, it should be highly valued. The insights I obtained after being brought into the informants’ inner circle greatly shaped the learnings around constructions of femininity and cannot be obtained unless researchers invest time, patience, restraint, and the highest level of confidentiality and recognize the risks that some woman informants take to share their lived experiences about their sexuality in a context where women’s sexuality is judged harshly through a patriarchal lens.

Conclusion

My 18-year experience as a researcher in South Africa attests that many Black Africans play a key role in data collection but usually as research assistants and coordinators and rarely as principal investigators. This also implies that the insider position is often occupied by less experienced and less educated research assistants and the outsider position by experienced and more educated researchers from urban areas and the global North. Having crossed the threshold as a doctoral student in a context where Black South Africans have immense difficulties accessing and maintaining educational opportunities (Connell, 2007b), owing to the legacy of a dehumanized existence due to apartheid which systematically emphasized racial, class, and cultural differences, I could appreciate the urgency to act responsibly and introspect on the extent to which I represented research participants in my research. Moreover, the recent calls for decolonized higher education, represented by the #FeesMustFall campaign in South Africa (Wikipedia, 2019a), could be advanced by exploring the Black South African researcher positionality, academic knowledge, and experience simultaneously with the research communities’ lived experiences and knowledges to provide a representative critique of researched phenomena (Kessi & Kiguwa, 2015). Local researchers need to invest in and interrogate their researcher identity, acknowledge their vulnerabilities, and account for the shared and different qualities with research participants as doing so may enrich the research process and interpretation of research findings. This may ensure that both the researcher and research participants contribute toward an emancipatory and democratic process of knowledge production.