(De)colonising outcomes of community participation – a South African ethnography of ‘ethics in practice’

Consent-assent for legal minors

When research involves legal minors, age-related power dynamics must be considered. Obtaining parents’ consent and thereafter the legal minor’s assent (agreement), is the expected procedure (Graham et al., 2015; Yanar et al., 2016). ‘Minors’ are people who are not considered legally capable of deciding for themselves because they are under the age of legal majority, which in South Africa changed from 21 years to 18 years in 2007. The consent-assent approach is underpinned by several unsubstantiated (and increasingly problematised) notions. These include that (1) legal minors are incapable of undertaking the moral reasoning required to consent for themselves (Montreuil and Carnevale, 2018), and therefore (2) need their parent or legal guardian (Nakkash et al., 2009) (3) to consent for them as a protective measure (Graham et al., 2015).

The assumption that parents or legal guardians are best placed to consent for minors reflects (romanticised) Western social relations, which have been problematised even in Euro-American contexts (Graham et al., 2015). Many adolescents are capable of making informed decisions about participation in research for themselves. The requirement for parental consent may present either as a barrier to young people exercising their right to participate (Fisher, Arbeit, Dumont, Macapagal, and Mustanski, 2016) or a form of coercion, because young people depend on, and are typically expected to obey, their parents (Montreuil and Carnevale, 2018). In postcolonial contexts (including African countries and indigenous settings in high-income contexts), there are further reasons to problematise parental consent. Children may not live with their biological parents and non-parent guardians rarely have legal status. The most ‘culturally appropriate representative’ in these contexts may be their primary caregiver, non-legal guardian or a community leader or teacher (Nakkash et al., 2009). Imposing the assumption that parents are best placed to consent implies the Western family is the only legitimate form (Bourdieu, 1996). Assuming universality of Western family relations is an example of how ‘coloniality of power’, a central concept in decolonial theory, persists in postcolonial contexts.

Decolonising

Decolonial theory provides a broad lens through which to understand social power inequalities as legacies of colonisation (Chilisa, 2017). The central premise is that an unjust ‘coloniality of [unequal] power’ persists despite the removal of colonial authorities. It exerts its effects through four inter-related domains: economy, authority, sexuality and knowledge/subjectivity (Mignolo, 2007; Quijano, 2007). The knowledge/subjectivity domain is particularly relevant to decolonising research. It articulates the ways in which academic knowledge creation historically buttressed colonisation and continues to perpetuate power inequalities (Mignolo, 2007), for example, by presenting Western knowledge (including knowledge regarding what constitutes ‘ethical’ practice or a family) as superior (Dawson, 2020; Sabati, 2019). Colonial knowledge systems privilege ‘scientific’, ‘modern’ (Euro-patriarchal) ways of knowing as superior to all alternative paradigms (Quijano, 2007). Scientific knowledge is presented as fact, in ways that colonise people’s subjectivities, that is, shape their perceptions of themselves (and others) (Bourdieu and Wacquant, 1992; Mignolo, 2007). Because scientific knowledge creation assumes Western superiority, the ‘facts’ created by scientists have often focused on the supposed ‘deficits’ of non-Western societies and people. They have been used to rationalise injustices (including colonisation) and marginalise non-Western viewpoints (Mignolo, 2007). Decolonisation is needed to enhance scientific knowledge by ensuring it acknowledges and gives equal treatment to diverse perspectives and experiences. This in turn enables alternative, more socially just and sustainable futures to be imagined and achieved (Dawson, 2020).

Pretending that scientific methods are detached and neutral is an important technique for situating colonial science as superior to other knowledge paradigms (Bourdieu, 1988; Mignolo, 2007). Hence, decolonial theory calls for ‘epistemic disobedience’, that is, delinking scientific enquiry from the illusion of neutrality (Mignolo, 2009). Community-participatory research is a form of ‘epistemic disobedience’ and a potentially important approach to decolonising knowledge creation (Ndimande, 2012; Wood, 2017; Zavala, 2013). It focuses on ‘engage[ing] in knowledge making’ (Mignolo, 2009: 176) people with lived experience of the phenomena being studied (Wood, 2017). Participatory approaches advocate academics working in partnership with community co-researchers who shape the research (Zavala, 2013).

In relation to VIAC, numerous community-participatory strategies have the potential to decolonise. Community members might participate in determining the best social spaces for conducting VIAC procedures and/or deciding whether (and which) adults should be present (Montreuil and Carnevale, 2018; Nakkash et al., 2009). Community participation could also inform decisions regarding whether information is most appropriately provided to individuals or groups, verbally or in writing and/or enable research to be conducted in Indigenous languages that affirm the cultural identity of the people being researched (Dawson, 2020; Ndimande, 2012). Group-based processes could situate VIAC within a relational ontology, which appreciates Indigenous people’s relationality (Chilisa, 2017). Working with co-researchers from the study community to design and implement ethical procedures might make them more culturally appropriate and sensitive (or disruptive) to power dynamics (Zavala, 2013).

Setting

Qwaqwa is a former ‘Bantustan’, that is, a racially segregated area that Black people needed permission to leave during Apartheid. The densely populated settlement consists of an urban centre and an expanse of peri-urban villages. The population of 336,000 are predominately Black (98%), ethnically Basotho (81%) and relatively youthful (33% are under 15 years of age) (Stats SA, 2018).

Qwaqwa has developed little since the introduction of democratic rule in South Africa in 1994. Unemployment is widespread (42%), especially among youth (53%). Most households (80%) have an annual income of less that R40,000 (~USD2800), are headed by women (51%) and lack basic services, including refuse disposal (75%) and piped water (68%) (Stats SA, 2018). Despite lack of material change, the ideological context of postcolonial South Africa is conducive to redressing the unequal power relations. For example, decolonising, community-engaged scholarship is enshrined in higher education policy (Williams, 2018). Many South African students have embraced the notion of ‘decolonising’ higher education (Lewis & Hendricks, 2017).

Study design

Community-engaged process

Our VIAC process initially involved establishing links with two youth organisations, one church-based, the other an artists’ cooperative. The leaders of these two organisations introduced us to their VYWs. Thereafter, we participated in the two organisations’ special events (e.g. youth month celebrations) and regular weekend activities. We invited the leaders and VYWs to participate in university events, including campus-based seminars and community-based activities for visiting international academics. They introduced us to the leaders of two other youth organisations.

Thereafter, Michelle developed and facilitated the workshop-based Participatory Research and Ethics Short Course. Eleven VYWs (aged 20–40 years, including the RAs) from four youth organisations participated in six workshops, which were hosted in a university seminar room. The VYWs were transported to the campus in a university-branded minibus and provided learning materials and meals. They received a certificate of participation upon completion. Each workshop involved 2–3 participatory learning activities (e.g. debates and role plays), designed to elicit the VYWs knowledge of relevant community norms and enable Michelle to introduce academic principles, procedures and norms. Michelle attempted to position community norms as the yardstick against which to judge academic norms, for example by asking participants to present their ideas first. However, she felt compelled to ‘teach’ the VYWs academic ethics, because of concerns cited by the university ethics committee and human resources department.

Ethics approval was initially withheld due to concerns about us (1) inviting community members to the campus, and (2) not providing (a) names and ‘qualifications’ of community RAs, (b) signed declarations showing the RAs had been ‘sensitised’ to academic ethical norms and agreed to ‘uphold good ethics standards throughout the research, including the protection of privacy’ and (c) professional Sesotho translations of participant and parent information statements. The ethics committee eventually approved our application, after we agreed to provide the requested information and documentation. We requested the committee reconsider their refusal to allow research in which participants entered the university campus (which in South Africa remains a gated space). They did. Thereafter, our human resources department initially refused to employ community-based RAs (because they did not have master’s degrees). They permitted their employment after we outlined the rationale for community participation in writing.

Following ethics approval and the short course, VYWs participated in recruiting youth participants. The three RAs led recruitment sessions at the four youth organisations. They introduced the research verbally and encouraged and answered questions in Sesotho and distributed English-language hand-outs. Youth were invited to submit a written or oral expression of interest (EoI) in Sesotho, Isizulu or English language. They were informed that participants would be selected from a hat (i.e. selection was not merit-based) and would need their parent’s written permission (but not money, transport, specific knowledge or experience) to participate. Michelle participated in all recruitment sessions, including by transporting the RAs to youth organisations in a university-branded vehicle. Other VYWs helped organise 16–17-year-old youth from their surrounding communities to attend. For example, at one organisation when youth had not arrived at the scheduled start time, VYWs went to their houses to hurry them up. In addition, one VYW who was not an RA (i.e. had not signed a contract and declared their ethical competence) partnered with a friend and independently door-knocked to recruit youth in their community. They also hosted a group information session at another community member’s house, during which they organised youth to submit written EoIs.

Finally, we organised VIAC procedures with youth and their parents. Thirty-three youths submitted an EoI. Six could not be contacted, and one later opted out. The remaining 26 were contacted by an RA to arrange a parent to undertake the VIAC procedures in a group session at a youth organisation’s premises (N = 11 parent–youth pairs) or individually at their home (N = 15). An RA and Michelle drove in a university-branded vehicle to meet potential participant-parent pairs and conduct VIAC procedures. The RAs typically distributed both Sesotho and English language information statements then read the English-language information statement aloud and translated it to local Sesotho vernacular, stopping frequently to invite questions.

Other VYWs participated in organisation-based VIAC sessions in numerous ways. For example, they introduced Michelle and the RAs, transported parents, took cell-phone photographs, shared their positive experiences participating in the Participatory Research and Ethics Short Course workshops and/or showed-off their certificates of participation. The VYW and friend who independently organised a recruitment campaign led (instead of an RA) five home-based VIAC sessions in their community.

Disrupting procedural ethics norms

Working with VYWs who were long-term (mostly lifelong) residents of Qwaqwa ensured that community norms were incorporated into the VIAC procedures.

Perpetuating procedural ethics (academic) norms

Disrupting academic norms

Community participation in the VIAC procedures for our qualitative study disrupted several procedural ethics norms, by demonstrating that they failed to achieve their intended purposes, in social realities to which they were poorly matched. This finding concurs with the results of other studies of community-participatory research processes (e.g. Kamuya et al., 2013). It extends previous research by analysing the mismatch theoretically in terms of postcolonial power dynamics.

Presenting study information in potential participants’ vernacular is an important way of shifting power to community members (i.e. decolonising research) (Ndimande, 2012; Wood, 2017). The procedural ethics focus on written information ignores the (originally) oral form of many African (and other Indigenous) languages and the persistent orality of indigenous languages in contexts like South Africa where a colonial language is typically used for written communication. Whether the aim is to ensure understanding or symbolise lay people’s vernacular as equally legitimate, imposing a colonial form (written) of an Indigenous language failed to achieve these aims in our study. Communicating in vernacular meant communicating orally. Reading the imposed form of Sesotho presented in the professional translations would reasonably have increased misunderstandings while simultaneously symbolising lay-people’s vernacular and the knowledge it represented, as inferior (Bourdieu, 1991; Ndimande, 2012). It would have eliminated the (albeit minimal) identity-affirming opportunities that communicating about research in potential participants’ vernacular provided (Ndimande, 2012).

Informal presentation of study information is apparently common in practice and accepted by some ethics committee. Our ethics committee’s insistence on seeing a professional translation implied that they wanted study information provided in standardised form. We interpreted their willingness to adapt (and potential to decolonise) some academic norms (e.g. by enabling community participation) as conditional upon us abiding by others (e.g. providing standardised information and maintaining privacy).

Ethical norms regarding confidentiality are based on assumptions that people volunteer to participate in research individually and secretly, want their identity kept private and may experience harm if they are identified (Graham et al., 2015). These assumptions were unfounded in our qualitative research about migration, in which participants put forward their names in groups and being part of a group was fundamental to having the courage to disrupt norms, which exclude most Black youth from the physical and social spaces of universities. Following academic confidentiality norms reinforced distinctions between university and community by vesting the power to know (e.g. youth’s names and phone numbers) only in the university-employed researchers (Bourdieu, 1988).

Treating information about potential participants confidentially was futile because it was never a secret. Even if it was, our potential participants were proud of participating in a project that associated them with the university (and all its actual and symbolic power) and wanted to be identified. Participants’ desire for recognition rather than secrecy has been noted in other participatory research processes in South Africa and high-income countries (e.g. Wood, 2017; Yanar et al., 2016). Our results concur and suggest that when information is not sensitive, qualitative researchers could enhance the ethics of their research by incorporating mechanisms for participant recognition, rather than assuming confidentiality is universally the best practice (Guillemin and Gillam, 2004).

Although VIAC is not a legally binding arrangement (Librett and Perrone, 2010), procedural ethics norms require legal guardians, typically biological parents, to consent for minors (Montreuil and Carnevale, 2018; Nakkash et al., 2009). This requirement was inappropriate in our study, because some youth did not live with their parents, and parents were not always considered the most appropriate consent provider. Again, this is not a novel finding. It concurs with evidence from other African (Nakkash et al., 2009) and high-income country (Hopkins, 2008) research settings. Given that researchers are not required or advised to verify identities in VIAC procedures (Librett and Perrone, 2010), our results suggest that insisting on parental consent is futile. It may contribute to parents and/or participants misunderstanding VIAC as a legally binding arrangement from which they cannot withdraw (Vreeman et al., 2012) in postcolonial contexts where formalities like signing are uncommon and/or symbolise formal legal arrangements (Brear, 2020). Undermining such misconceptions is an important step in decolonising qualitative research in contexts such as South Africa, where one legacy of colonisation is considerable family separation due to work-related migration. In our migration research project, obtaining consent from non-legal guardians known to VYWs, decolonised by enabling us to include the youth who had direct experience of parental migration. It also shifted power by legitimising different forms of family, which may have subtly been stigmatised, if we had denied non-legal guardians the right to consent for youth who they primarily cared for (Bourdieu, 1996).

In procedural ethics, autonomy is often also conflated with deciding for oneself, without external influence (Campbell, 2017). This is one reason why individual VIAC is considered inappropriate in indigenous societies, with relational decision-making norms (Chilisa, 2017). Participants in our study considered relational decision-making appropriate and normal. Our theoretical lens, which conceptualises all individual choices as socially structured by habitus (i.e. embodied historical knowledge of unequal power relations) (Bourdieu, 1990), situates people influencing each other’s choices (i.e. relationality), from various positions of power, as not only normal but inherent. It indicates, as do our results, that relationality in decision-making is unlikely to be specific to Indigenous people. The interactions between the authors (one Black Indigenous man, one White Western women), ethics committee, human resources staff, RAs, VYWs and youth demonstrate a high degree of relationality for all participants in this study. For example, Michelle’s decisions were influenced by the ethics committee’s concerns, the practices of the RAs and historic experiences conducting informed consent procedures in postcolonial research settings (as described in Brear, 2020). Situating relationality as an exclusively Indigenous norm perpetuates the colonial focus on intrinsic differences. It implies that social differences are inherent (biological, unchangeable) rather than logically constructed responses to social, cultural and economic circumstances, especially unequal power dynamics. In showing that relationality is inherent, our findings suggest that other dichotomies between Indigenous and non-Indigenous people and practices are socially (scientifically) constructed. Our theoretical interpretation of a community-participatory approach disrupts the assumption that VIAC is or should or could occur in the absence of social influence. It suggests the need to reconceptualise autonomy as deciding for oneself in a relational context, rather than without external influence (Campbell, 2017). The inherently relational nature of individuals should at least be acknowledged, probably embraced, in research ethics.

Relationality is an accepted feature of community-participatory approaches. They purposively utilise existing relationships and the subjectivities inherent to them, to enhance knowledge creation and disrupt unequal power relations (Wood, 2017; Zavala, 2013). Existing relationships we had formed with VYWs made important contributions to enhancing the ethics of our study. Particularly important are the ways in which the RAs and VYWs influenced (encouraged) participation in a project they perceived would provide youth opportunities to acquire valuable forms of social and cultural capital (e.g. new friends and certificates of participation).

Perpetuating hierarchies

Enabling VYWs and youth to access a symbolic form of cultural capital (i.e. a certificate) potentially represents a decolonising outcome. However, only to the extent that it legitimised the value of participatory co-learning and disrupted norms that conceptualise knowledge-worthy-of-a-certificate (i.e. legitimate knowledge) as something created only by experts and reproduced only through didactic teaching (Mignolo, 2007). Our results show that participatory learning was not legitimised, and only minimally acknowledged, in the community. Even after experiencing participatory co-learning for themselves, some VYWs assumed and indicated to potential participants that our research would involve ‘teaching’. Potential participants and parents informed by them assumed that the participants would be subjected to a ‘merit-based’ selection process and ‘taught’, when we intended to learn from youth selected ‘lottery’-style. Community participation thus had the unintended and undesirable effect of presenting academic knowledge as something participants should be striving to acquire, and as superior.

In this context, getting a certificate or travelling on the university bus became a source of pride, but not because VYWs and youth conceptualised these as symbols legitimising their Indigenous knowledge. Youth and VYWs wanted to access (albeit minimal) actual and symbolic forms of capital/power available in the academic field, rather than to shift that power to the community field, as accounts of community-participatory research imply should happen (Zavala, 2013). Theoretically, it is unsurprising that colonised people (i.e. those with limited power) would participate in academic research with the intention of gaining capital/power (Bourdieu and Wacquant, 1992). The youth’s and VYWs’ subjectivities/habitus had been shaped to accept the superiority of academic/scientific knowledge, showing how ‘coloniality of power’ (Mignolo, 2007) continues to exert colonising effects, decades after the official end of colonialism in South Africa.

In postcolonial settings where many lack economic and cultural capital, participation in health research has previously been shown as motivated by expected economic/material (e.g. healthcare or food) gains (Kamuya et al., 2013). Gains in social and cultural capital, which our study showed also to be important, have received less attention. Participatory research principles imply that accessing social and cultural capital by working in partnership is valuable for community members and academics. However, the tensions have not been drawn out and are glossed over in conceptualisations of participation as decolonising research by catalysing community ‘ownership’ (Zavala, 2013). Academic institutions and their epistemologies maintain great (social, cultural and economic) capital/power in postcolonial contexts (Mignolo, 2007; Wood, 2017). Community members wanted to participate in our research in order to gain social and cultural capital that might enable future opportunities (including economic) (Brear, 2020). Community participation was driven as much by a desire to secure a place in the academic hierarchy, as to usurp ownership of the academy or disruptively problematise its knowledge creation processes.

The Bourdieusian conceptualisation of capital/power having multiple (social, cultural and economic) exchangeable forms has important implications for decolonising research. It suggests that decolonisation will be limited while ever efforts focus narrowly on the academic field. Universities occupy an elite position within broader hierarchical relations of (social, cultural and economic) power (Bourdieu, 1988; Wood, 2017). Our results suggest that economically marginalised lay-people will valorise the social and cultural power of universities, while ever scientific knowledge can be exchanged for relatively large amounts of economic power (e.g. relatively high salaries), and that academic research cannot truly be decolonised, without broader shifts in economic power relations (Wood, 2017). Paying community-recruited RAs who did not have the expected formal qualifications represented a minimal shift in both economic and cultural capital. It provided them actual economic power (i.e. money) for a short period of time and symbolised the (economic) value of their lay-knowledge. It enabled them to dedicate time to participating intensively and develop dispositions that better enabled them to ‘compete’ in the university field. However, colonial ways of thinking and doing research were not intentionally challenged; rather, the select group of RAs and VYWs tried to assimilate to university norms (Dawson, 2020).