Revisiting VET research paradigms: Critical perspectives from the South

ABSTRACT

This paper reflects on a large multisite funded VET research project conducted by a large and diverse research team. Reflecting on two of our case studies, from Uganda and South Africa, we consider both the need for broadening the VET research agenda to incorporate more research on non-formal sites of vocational learning and work, and the imperative of continued critical reflection on modalities of researching the formal sphere. What we offer is a very fallible attempt to open up the debate about the future of VET research further through, we believe, a critical reading of some of our failures as well as successes in trying to ground our research ethically, ontologically and axiologically and not just methodologically. We advocate, where possible, a radical embeddedness of VET research in communities, whilst acknowledging that this is applicable only to some parts of a comprehensive VET research agenda. We also acknowledge that employers and the state are also legitimate stakeholders who should be part of research but point to the need for a more critical reflection into the patterns of power implicit in researching with/on these constituencies. We believe that our reflection on our successes and failures in these two cases and the project as a whole offers useful provocations regarding ways of making VET research more reflective of diverse settings, less extractive from those being researched and more equal in the participation of members of the research team from the South.

KEYWORDS:

• Participatory research
• VET research methods
• VET Africa 4.0

Introduction

‘International’ journals for vocational education and training research have seen a growth in the past decade in the numbers of accepted articles both from Global South authors and, more broadly, on Global South topics (e.g. Aldinucci et al. 2021; Maitra, Maitra, and Thakur 2022; McGrath et al. 2020). In a recent special issue of the International Journal of Educational Development, McGrath and Yamada (2023) explore the nature of and reasons for this growth. Crucially, they argue that some of this work is not just taking Northern theorisation about vocational education and training but is contributing new theoretical directions to the field. Alongside recent special issues on race and VET (Avis, 2023; Avis et al. 2023; Avis, Mirchandani, and Warmington 2017), this work has a potentially important contribution to make to questions of ‘knowledge and expertise’ (Horden et al. 2022), a longer standing strand of the Anglophone VET literature.

One of the important aspects of the growing presence of Southern literature in the international VET journals is that some of it disrupts orthodoxies that are rarely contested in the North, regarding the place and role of VET in society. Whilst there remains a large Southern literature that considers formal public vocational education and training institutions and/or large formal private firms, there is an important strand of this literature that looks at a far wider range of vocational learning sites and workplaces (Jjuuko, Tukundane, and Zeelen 2021; Tukundane et al. 2015; Alla-Mensah and McGrath 2023; Brown 2023; VET Africa 4.0 Collective 2023. This does not stand as some exotic adjunct to the mainstream literature on VET but, increasingly, as a contribution to critiques of assumptions that Northern approaches to VET research (whether Germanic, Anglophone or of some other form) are generalisable or, rather, deeply contextual.

This is not to argue that a conventional VET focus is not relevant in the South (cf. Allais 2020; Otchia and Yamada 2021; Sibiya 2023). Rather, it is to suggest that there are alternative approaches that must be considered, which may provide insights into Northern research also. Hence, we consider here how a concern with researching more complex vocational learning and working contexts led us to engage with longstanding discussions about power, participation and partnership in research, as seen for instance in the development studies literature (White 1996). Equally, longstanding decolonial concerns with the nature of formal research (Smith 1999; Hall and Tandon 2021) led us to consider whether there were issues in our more conventional practice of VET research in industrial settings marked by high levels of inequality and low levels of trust, and where key identity markers such as race and gender play crucial roles in assigning status and access.

In what follows, we draw on our experiences as researchers in a large internationally funded VET research project (discussed at further length in the next section). We consider two of our fieldwork settings in particular (out of four): one a site characterised by informality where the team were grounded in adult, youth and community education traditions; the other a more ‘classic’ VET location where the research was conducted by a leading VET research centre. In presenting the two very different cases, we try to provide an argument that the non-standard approach to VET research brings valuable new insights to VET research but we are not arguing that this approach is superior to more traditional ways of doing VET research. Nonetheless, those of us who are more versed in conventional approaches did find the alternative standpoint a useful one in advancing a critique of our own practice. However, our argument here is that reflection on both approaches is valuable for the development of a more inclusive and reflective VET research practice. We now move into a brief methodological discussion about the paper and the wider project from which it is drawn before critically reflecting on two of our case studies and then proceeding to discuss what implications these might have for the wider VET research field.

Methods

We wrote this paper as a group of researchers reflecting on a project funded by the UK Global Challenges Research Fund (with significant co-funding by Southern partners). The project was, with a no-cost extension, 33 months long. Before that, there was an initial 10 months when some of us were involved in developing the bid and then planning for its starting. Ethical approval for the research was obtained at a British university and from relevant national ethics bodies in Uganda and South Africa. All participants in the research provided informed consent in a language that was familiar for them. At the time of writing, we are 2 years post completion during which we have written a book and continue to collaborate in smaller groups around several other papers.

This paper is authored by eight members of a research team of 20. Whilst all 20 are subsumed under the corporate authorship of the ‘VET Africa 4.0 Collective’ for the related book, here we have followed accepted journal authorship practice of including only the active participants in the paper’s development. We are racially and gender mixed, with seniority ranging from a Vice Chancellor through to a junior researcher who achieved a postgraduate qualification whilst we were writing the paper.

Our approach has been broadly reflexive and the number of authors makes it a very particular exercise. We are conscious our interactions are imbued with power, and there inevitably will be things we couldn’t or didn’t say or write. Moreover, not all of this is accessible through the dialogic process we strived to use in writing together. The reasons for some team members’ non-participation in cowriting here were often ostensibly about unavailability, both of senior academics and those in more precarious positions. We cannot know for certain what is absented by these absences.

This paper has its origins in planning for a conference presentation. An initial team of four, the UK-based PI and an early career researcher from each of the three African universities, wrote short initial reflective narratives on the research cases and our positionality. We drew pictorial representations of our interpretations of the research and then reflected on them together. We then drafted a longer narrative that formed the basis of a blog and conference presentation. Following this, we revisited our narratives and asked colleagues for their reflections as well. Some of these became involved as co-authors though not part of the writing core. Along the way, a change of job and continent led one of the original team to have to pull back from working further on the paper. Subsequently, we have met (almost entirely online) and discussed, written, revised, checked, co-authored a book with others in the team, made more presentations, and crafted this constantly shifting piece. There has not always been consensus in the writing team and others in the wider project pushed back in response to some elements of our critique, leading to further elaboration of our storyline, with critical but supportive reviewer comments prompting a further round of internal debate. That the paper has taken something like 2 years in its development is actually a rare but welcome opportunity for the main authors to engage in an extended dialogue about its nature, in stark contrast to the rapid extractivism of much contemporary ‘collaborative’ writing.

The research project this paper reflects on consisted of four cases, two each in Uganda and South Africa, and sought to challenge traditional formal sector only perspectives of VET in development by understanding the aspirations, challenges and intersections of life, living and work, with an emphasis on designing opportunities for socially and economically marginalised people (McGrath et al. 2020). In all four cases, at least one team member had some form of insider status. However, as we will reflect below, there was a significant difference between those cases where teams were already significant actors within local skills ecosystems and where it was more a case of a few team members having some involvement, perhaps more peripherally. This led us to choose two of the cases as being most emblematic of some of our internal tensions and dynamics.

In what follows, we first present the context of the two cases we are primarily drawing on (Gulu and eThekwini). We are mindful of Eyben’s (2004) five-fold notion of power: power to, power over, power with, power as knowledge, and power structure, as we interrogate the authenticity of participatory aspects of our research process. In a range of contexts, authors have argued that education and training institutions that do not engage with questions of justice are part of the problem (Hall and Tandon 2021; Smith 1999; White 1996). This has given rise to calls for reframing of understandings in ways that repoliticise roles by recognising the time spent building change and connecting with communities (Jordan and Kapoor 2016). This highlights the importance of ‘bottom-up’ processes that take a nested approach to intersecting inequalities and pay particular attention to local communities (Velasco-Herrejon and Bauwens 2020). We draw in part on Participatory Action Research (PAR), which offers one alternative way forward that foregrounds epistemic justice through cocreation of knowledge. Its power-sharing approach involves the deliberate creation of reciprocal relationships between researchers and communities to facilitate ‘strong and sustainable transformation’ (Benjamin-Thomas et al. 2018). PAR also recognises that researchers and participants both have situated and experiential knowledge that can benefit each other (Ozano and Khatri 2018). However, McDonald (2021) cautions that simply bringing people together for a ‘participatory activity’ does not guarantee authentic participation. Neither does it equate to shared power and knowledge democracy. Likewise, Lynch et al. (2012) caution that participatory research approaches are embedded in power-full relationships that could be positive or negative and consequently require a great deal of care. In this paper, we are mindful of such cautions and draw on the depth of participatory research in the radical adult education and decolonisation fields (Clover 2011; Freire 1996; Hall and Tandon 2021; Odora Hoppers 2021) to guide our reflexive work. We seek to present these alongside a reaffirmation that an important part of VET research will continue to focus on interaction with policy and industrial elites who remain key VET stakeholders.

Case narratives

Gulu

Gulu is a city in northern Uganda, which is in a period of transformation and development following more than two decades of civil war. The economy is mostly informal, and there are high dropout rates from primary school. Gulu University, established in 2002 during the war, plays an important role in the development of the region. Wedekind et al. (2021) described the university as an important anchoring institution in a regional social skills ecosystem. The university has a strong tradition of community engaged teaching and research. The lead researchers from Gulu, the Vice Chancellor, and the then UNESCO Chair came from the radical adult education tradition and the research was embedded in their ongoing work and scholarship related to youth, learning and work (Jjuuko, Tukundane, and Zeelen 2021; Monk et al. 2020; Openjuru 2010). This has critiqued formal VET but needed further development regarding the informal and non-formal dynamics of learning and work. The research was embedded in ontological and epistemological foundations of interdependence and co-creation as well as a participatory research culture embodied by the researchers (Monk et al. 2020).

The Gulu case was influenced early on by the work being done in the Alice case in South Africa (VET Africa 4.0 Collective 2023). On encountering their work through the project, the Gulu team quickly recognised the resonance for their theoretical, methodological and empirical contexts. The Gulu team learned from their earlier work and oriented our research towards entering and supporting the development of a local network similar to theirs. In particular, they borrowed their network mapping approach (slightly modified from Metelerkamp, Drimie, and Biggs 2019), selecting this as a first research activity.

For the first 2 years, the research was much akin to a number of balls being juggled in the air, each representing a different dimension of the research. The research office, located in the back of a mechanic’s yard in town, was a mess of poster boards, jamboards, and maps as the team attempted to situate the various networks and pathways of young people navigating their lives and integrating learning to match their aspirations with practical day-to-day livelihood needs. The team spoke with indigenous communities related to medicine and metallurgy; youth art groups; market women; roadside food vendors; directors and students of NGO and CBO subsidised village training programmes; NGO programme directors for VET initiatives; owners of non-formal VET schools of varying sizes; small business owners in tailoring and welding; and farmers and agriculture businesspeople. From the formal sector, they spoke with directors, students and teachers of large vocational schools and Gulu University, and policymakers.

In the process of collecting data with a diverse team of researchers and pursuing multiple and diverse themes and networks of learning, the team was able to provide a very different perspective of the social skills ecosystem in Gulu than had traditionally been presented. The process required allowing for the emergence of new and different actors, allowing multiple perspectives to emerge. This flipped traditional hierarchical understanding of VET away from the model of formal education seemingly designed by omniscient leaders for helpless youth, to a more robust understanding of the dynamic and complex interactions among policy, formal educational opportunities, informal economy, aspirations, dreams, innovations and survival mechanisms influencing and changing the society. In our book chapter reflecting on informality in Gulu and Eastern Cape, we reflected:

Therefore, it was unsurprising that in both case studies, we observed that the nature and structure of relationship superseded the specific nature of content. Yes, access to useful information was an important driver, but relationships were ultimately the starting point for the value created for participants across the different types of networks in both countries. Given young people’s need to remain highly adaptable, opportunistic and resilient in the face of unexpected shocks, it was relationships that allowed people to assemble, repurpose and reconfigure knowledge into dynamic responses.

(VET Africa 4.0 Collective 2023, 96)

Our research approach came to mirror the chaos of the informality which we were immersed in. In what follows we reflect on some key points that enabled us to enter the fray.

The research team comprised three academics (the two professors and a lecturer), three recent graduates of Gulu University, and one local businessperson and grassroots political leader. Only one was a woman. An advisory team with experience in VET was established. All are Ugandan, except one of the professors and the lecturer, both of whom are white males.

From the outset, there was an emphasis placed on equality across the team. The senior researchers assumed a trusting and non-obtrusive mentoring role, providing space for others to flourish, make mistakes, follow their interests, and learn together. This was important, because the team came from diverse backgrounds with very different readings of the world. Their diversity and participation in the city’s predominant informality allowed for entry into informal settings at a grassroots level and with grassroots framing, and it also encouraged the study of different VET pathways. For example, one of the researchers was particularly interested in indigenous knowledge, which led to a whole subset of research into traditional practice and innovations in VET. Another was interested in experiences of people with disabilities, another in opportunities in the arts, another on environment and climate change, and yet another in the experiences of women in small markets and roadside vendors. All the non-professorial researchers were youth, partially living and acting in Gulu’s informal sphere. Following them as leaders naturally brought us into the informal spaces, bringing out a diverse range of stories that gradually began to contest narratives of youth inadequacy and immobility and helped to shape an understanding of VET as fluid, relational and interconnected to other parts of learning and living. This reflects the suggestion by Cook et al. (2019) that for new voices to emerge, old voices need to be disrupted. Facilitating spaces of leadership amongst the younger researchers led to disrupted narratives and a new frame from which to understand VET. The research process became exciting as the team began to make connections across the diverse strands being followed.

Multiple seminars and training workshops were organised to support team members in community engagement, coding, analysis and writing. This allowed team members to develop professionally and individually. For example, the female team member who was initially hired as a project administrator showed interests and promise in conducting research, she was mentored and given an opportunity to work as a researcher. This allowed her to flourish, continuing to work on multiple other projects as a researcher and eventually enrolling for a masters at one of the South African partner universities, which she has now successfully completed. Another decided he wanted to write his own journal article, which he was supported to complete.

The team’s diversity facilitated entry into different parts of VET-related learning and living. We found that which questions were asked, how they were answered and how they were interpreted was significantly different depending on who was asking. This became particularly noticeable in the different stories collected by the female researcher, particularly when she was engaging women. This realisation led her to assume a lot more responsibility within the research and helped us to reflect on how we were entering the different communities, and the relationships we developed in the process.

Benjamin-Thomas et al. (2018) remind us that reciprocal relationships are essential to co-creating knowledge and sharing power in research. In Gulu, as we demonstrated an openness for co-constructing knowledge through deeper situated learning in the communities, the communities, particularly the youth in the informal spaces, guided us. We came to learn that the informal is based on a deep relationality that required researchers to be within the research context, not remaining aloof. We will switch to first person singular to offer an illustration in a more authentic voice.

As a component of developing pathways in fashion design, I biked a fair distance to go and meet with a few young women to discuss their pathways and life aspirations. I distinctly remember being in a hurry and hoping to complete my life grid and be on my way. Upon arriving, however, I found the young women had prepared some food for me, something they would not normally be eating because of the expense, they introduced me to their families sharing the homestead and to their children. We shared some food and some stories and then later in the afternoon they brought out some of their drawings and explained to me what they had learned from different places, how they had moved around the country, tried out different businesses, sought out support from different people. They invited me into their homes and into their lives. They were generally interested in sharing their innovations, particularly in the realm of scavenging and recycling to save on costs, but also to clean up their environment. It was from those few hours that we embarked on a new direction of research questions and decided to try out a VET ecochallenge. These women also happened to be connected to the grassroots art scene in Gulu, and thus emerged a new connection and node in the social skills ecosystem. I am now a board member of a small social enterprise they developed, and a member of their (and countless others’) community organisations. They have presented as guest speakers in my university classes, and we are friends.

This example demonstrates the power of engaging with people in reciprocal relationships. As the participants came back to us and asked us to participate in their work-lives, demonstrating trust, we entered a dynamic learning space in the real world and had the opportunity for our research to be impactful as evidenced by the continuity and growth of most of these networks and relationships. We naturally need to value people, their voice and the intersections of their life pathways. Research becomes more embedded in everyday life, with longer term goals of mutual benefit, which reduce power hierarchies and knowledge assumptions. Reciprocal relations are closely intertwined with authentic participation in shared lives that must be thought of more broadly than a research project. Reflecting on McDonald’s (2021) cautionary note related to participation and power, we are mindful that the above example remains full of power – where the researcher is in a position of considerably more power – to choose or not to choose how and when to engage for example.

Jordan and Kapoor (2016) explain that engaging with communities in a bottom-up process is important for developing equitable relationships. In Gulu, in the formal sphere, we started the engagement process by forming an advisory committee of people immersed in VET scholarship, and with a large network mapping exercise based on their contacts. We also later developed some communities of practice oriented around pilot practices. These engaged largely with decision makers talking about and for ‘youth’. These were important. However, as the research progressed, and relationships developed, we shifted from engaging with community to engaging in community. It is difficult to say which happened to us first, relationships or spaces for youth participation. Upon reflection, they seem to go hand in hand. Many of us were already engaged in the informal spaces of youth, and so some relationships existed, which made it easier to add the research component. At the same time, being open to new relationships and trusting youth to lead us in new research directions helped to engage in deeper relationships and learn more about our dynamic youth world. An example of such participation is how we worked with an informal youth group who led a series of ‘community learning cafes’ using painting, theatre, and many other activities to engage the community in debate about challenges and opportunities related to decent work, learning and livelihoods. During the COVID-19 pandemic, they turned to radio programmes. Youth shared vibrant stories of dynamic learning, catalytic community change and the potential for VET to ‘catch up’ to their needs and dreams. These youth asked of each other, and of their community, different questions which were not always seemingly related to VET, and yet which in fact could not be taken out of the broader equation of learning and living. In doing so, they demonstrated that lives are complex and to understand VET we cannot limit the study of it to education.

We also participated in youth-led initiatives becoming contributing community members. For example, we engaged with a small tailoring business taking on interns and doing on-the-job training. The owner decided that as she expanded her business, she required more qualified workers. After conducting a rather hurried interview in a hectic workplace with her, where she shared some problems she had with the quality of learning programmes, she came back to us and asked us to help her develop a quality fashion design school. Over the course of the research, we participated in a series of meetings with her and helped to think of a business model, develop an initial curriculum, train some of her teachers in student-centred teaching, participate in certifying and (post-project) accrediting the small non-formal school. Some researchers continue to teach basic digital marketing skills in the school programme, and one is a board member. As we engaged in the process, we were able to observe, reflect, experiment and continually dialogue with the workers in the factory, the students, and the value of an alternative, empowerment-centred, project-based, curriculum, and the significant challenges involved in the process. We also provided direct, research-based impacts in our community.

From the outset, we had strategic goals of developing a network and building a movement for both shifting the stigma associated with VET and improving society through the process of research as praxis. This aligns with Benjamin-Thomas et al. (2018), who explain that participatory research should work towards justice and benefit the community involved. In Gulu, this guided our openness to relationship building.

We purposefully created spaces for such a network to develop. For example, we chose to experiment in a virtual reality (VR) programme to overcome lack of resources such as tractors in formal VET spaces. We brought a diverse group of actors together from NGO, University, formal vocational institutions, and a start-up youth-led VR company in a community of practice. Together, we decided to run a pilot programme on tractor repair, developed a curriculum, filmed a series of short videos, and invited students and teachers alike to test it out. We held a small reflective forum on it afterwards with remarkable discussion related to student centred pedagogy. One of the NGOs involved then took up the use of VR in a slightly different way on their own. There was considerable interest amongst the practitioners, and one vocational training institution is also working to take it up in different contexts.

However, the community of practice has lost traction without the researchers who were initially at the centre bringing it together. So, while the intention was good and some ideas were taken up, the more direct impact goal of developing a community of practice has faded. Nonetheless, in the end, we have become contributors to an existing social skills ecosystem full of nested networks (VET Africa 4.0 Collective 2023) in which this community of practice plays a role. For example, a young social entrepreneur who participated in the forum has recently engaged with our research team to use VR in promoting her cultural tourism business. One of our team acquired filming and editing skills in the process and may move into this space. Such outcomes are unpredictable, but creating the spaces for them to emerge is essential. This was made possible by aligning the research goals with the process of the research.

Reflecting on attempts to engage in authentic participation and power in the Gulu case, there is opportunity to engage in critical discussion about possibly hidden power-ful discourse in participatory research processes. For example, while we had a push towards investigating the experiences of people with disabilities (a substantial portion of the population in Gulu), we neglected to forge any substantial networks and relationships with the various unions and community organisations for people with disability in the region. This reflects a shortcoming in our research, as well as the power-full relations that are nested in participatory research that need to be made explicit. Likewise, we can reflect on our team composition, which was composed of six men and only one female – none of whom identify as minorities or vulnerable. This has an impact on which communities we engage with, what questions we ask, and the responses we receive.

Participatory approaches to research emphasise social justice and aim to integrate epistemic justice and knowledge democracy in the research processes by valuing and raising the voices and epistemic contribution of those which are often ignored or undervalued (Hall and Tandon 2021; Monk et al. 2020). Participatory research processes are, however, mired in the messiness of everyday life. The researchers have significant power that needs to be acknowledged, including in decisions about what to research, who gets to participate, and what is socially just. Because participatory research is ideally embedded in the community, this is not so straightforward. Relationships are messy, communities are diverse and there remains significant influence on the broader social context from which the research emerges. For example, a reviewer of this paper, pointed out that much of the orientation of the research in Gulu was oriented towards understanding capitalist economic understanding of youth livelihood opportunities. This would seemingly be at odds with the authors and researchers’ perspectives, which claim to be critical of the ‘capitalocene’ perspective of development. Correct or incorrect, this demonstrates that not unlike other forms of research, there are issues of power and social influences that are inherently embedded in the participatory research processes that must be reflected on and navigated. It becomes especially important to reflect on these critically in participatory research approaches because of the claims of social justice that are made by participatory researchers.

eThekwini, Kwa-Zulu Natal

While the Gulu case demonstrates a move towards a participatory approach to VET research, the eThekwini case represents a more conventional political economy of skills research modality. In contrast to the context above, this represents the very different relational world of the formal sector in a low trust setting. Reflection on this case enabled an understanding of complex institutional arrangements and did so by spotlighting the nature of macro-institutional pressures in operation. These pressures led to a range of trade-offs, some explicit and others implicit, that were the result of forms of control and regulation over resources including access to information. To understand actors at levels other than the macro, the research team working on the eThekwini case attempted to bring the political economy skills mode of research into conversation with ideas around just transitions. This presented challenges of its own which the team overcame in some ways but not others. In this section, we examine these matters further, showing how they affected the research process.

Providing a background of the issues raised in the eThekwini case demonstrates why and how they present classic political economy research questions. South Africa’s (problematic) version of a ‘developmental state’ has seen interventionist regional industrial policies, including Strategic Infrastructure Programme (SIPS), Industrial Development Zones (IDZs) and Special Economic Zones (SEZ). Given eThekwini’s position as a major international port (Durban) serving the main industrial region inland via rail and road links, it had been targeted by these policies. We wished to investigate the place of skills within these policies and their implementation, driven by theoretical concerns that this would give a different angle on debates about skill supply and demand, and the role of skills in development.

The eThekwini case presented limitations for research into skills ecosystems early on. In 2019 initial fieldwork was carried out. The fieldwork revealed that when examined as a single case, there appeared no significant skills impact from the spatial and structural investment programme. Following on from this, a strategic methodological decision was made to restructure the single ‘case’ and present it as a set of embedded cases with different units of analysis. Together, these would provide combined insights into the functioning of a wider skills ecosystem.

This restructured methodological approach began with an extensive degree of initial desktop research. Academic literature and policy papers on national and provincial development strategies were reviewed, alongside labour force data and education and training strategies and data. These informed the development of occupational workstreams (rather than a single job, a workstream includes a broad range of occupations with vertical and horizontal linkages between them). This way of undertaking the research was conceptually useful in order to establish an examinable site of impact, in a research context where impacts seemed limited. We expand on how this informed our choice of interviews in a later section.

Whereas our initial design saw Gulu both as a site of research and a location of a key research partner, these processes were separate in the eThekwini case. In this case, the long standing partnership between English and South African teams was to the forefront, reflecting their leading roles in African VET research. The decision to focus on eThekwini was a design choice, although the team did have links to the area. Thus, whilst the University of the Witwatersrand’s (Wits) Centre for Researching Education and Labour (REAL) is one of the world’s leading centres for political economy of skills research, it is also situated 600 km away from the case site. Moreover, its history is one of white liberal elitism, presenting a clear formal orientation. Nonetheless, REAL does have a growing component of community-oriented research, challenging the historical orientation. Managed by the Director of the Wits REAL Centre, our fieldwork team consisted of two black early career researchers, one (male) at the start of his PhD and the other (female) towards the end of hers, as well as a white male senior researcher who was not part of the REAL core staff but who was based in eThekwini and who had many years’ experience within the metropolitan skills system.

Given its policy-related research concerns, the study had an elite-facing aspect. As is typical in such research, semi-structured interviews and documentary analysis were core elements. Interviews were conducted with a range of national departments (Trade and Industry, Environment Forestry and Fisheries, etc.); the Kwa-Zulu Natal Premier’s office; the eThekwini Municipality’s Economic Development Department; the eThekwini Maritime Cluster; the Kwa-Zulu Natal Provincial Training Academy; public TVET colleges; private colleges and training centres; one community NGO and one environmental NGO, to name a few.

The elite-facing nature of the study raised challenges with respect to the data collection process for our Wits research team. The elite of government officials and state-owned enterprise employees in the eThekwini case plays a facilitating and directive role in relations between the various actors that constitute the eThekwini skills ecosystem. These elites are also the enforcers of the ‘rules of the game’ within the ecosystem. Their status gives them the ability to exert influence through social networks, social capital and strategic position within social structures. At the same time, elites are also difficult to access in comparison to other groups. In the eThekwini instance, barriers to access existed for two main reasons: the first was that elite informants tended to be reluctant to be interviewed for fear that they might share sensitive information, putting themselves in a position that could harm them professionally later on; the second was that the Wits researchers were perceived as distant because they were conducting research from an entirely different province and coming from an elite university.

In contrast to the Gulu case, then, building reciprocal relationships of trust with informants who felt this kind of vulnerability and distance proved exceptionally hard. It also meant that the use of techniques like snowball sampling was more difficult. As Eyben (2004) states in her five-fold conception of power, power over determines the ability of informants to participate in research. In the eThekwini case, informants’ vulnerability indicated the power of institutional arrangements over them. Further to this, our distance made it hard for the two early career researchers on our team to fully understand the context and the need for research methods that could overcome the barriers that this created.

However, in the light of difficulties around access, the personal networks of the expert researcher on our Wits team proved invaluable. As a former senior official in the municipality and a long-standing member of both municipal and provincial planning commissions, he could take an affective approach to securing interviews through the use of the relational commitment that was felt between him and other ‘insiders’ in local institutions.

This is an illustration of the complexity of conventional research approaches. Although on the one hand it may appear as though this kind of research can be carried out in the absence of human relationships, the eThekwini case indicates otherwise. At the same time, the senior researcher’s white male positionality also played a role in the distance between himself and the same informants, who were usually black. The privilege and power associated with white masculinity sometimes meant that there may have been concerns that the expert researcher’s intention was to ‘catch out’ those who could be caught out for any misconduct or abuse of their institutional power. This was exacerbated by the context of the municipality being under legal investigation for fraud and corruption. Together, this and the legacy of research in South Africa as surveillance or as a kind of auditing exercise (Powell and McGrath 2014), created a relationship of suspicion between informants and the university researchers. The data arising from this set of interviews tended to conceal areas of discussion that were regarded as contentious with the use of euphemistic language, demonstrating the need among informants to protect themselves.

The fraud investigation and a wider spread public concern about the increasing depth of corruption also amplified previously existing tensions between the state and other actors, most visibly the private sector. In the eThekwini skills ecosystem, this was manifested by an almost total disconnect between the two subsystems. Whilst the private sector paid the required levies to fund public training, they also continue to invest heavily in their own parallel skills programmes to train current and potential employees. Interviews carried out with private sector key informants produced only a small amount of data that spoke to the public programmes available through the state. Instead, they tended to produce data indicating the reasons for which these informants viewed their in-house programme offerings as more competitive than those provided through the public education and training system. The narrative that public TVET institutions lack the capacity to deliver critical education and training programmes was echoed by both private sector employers as well private TVET providers, who argued that they boast better equipment and incentivise their lecturers by paying them three or four times more than those in public TVET institutions. Relationships across the skills ecosystem were seen as being so bad that one public official described his experience working in a Sector Education and Training Authority as ‘trauma’ because of the level of mistrust and blame.

Thus, while the South African government may claim that this skills ecosystem is co-ordinated, such points generated from the data contradict this. The fragmentation between actors in the ecosystem and the expressions of informants’ frustration at this fragmentation was useful. It provided lessons around the main sources of tension in the eThekwini case, giving our research team some initial themes which we would later draw out in our analysis.

Together, the sections presented so far have shown the complex macro-institutional dynamics of the eThekwini case and how these shaped the data collection process, as well as the kind of data generated. The cultural context of the eThekwini case added another layer to these dynamics, again shaping the data collection and generation processes. In this regard, the positionality of the black, female early career researcher in the elite interviewing process also needs reflection.

A young black woman from a Johannesburg township, she participated in interviews alongside our Wits team’s expert researcher. The experience of the interviews was therefore not entirely a reflection of her single positionality, but the positionalities of the two researchers combined. Crucially, eThekwini municipality is one of the most traditional parts of South Africa. The dominant isiZulu culture is still underpinned by rural, traditional values. Respect for authority (defined typically by sex and age) is highly valued. This determines how relations are organised even with those outside of the isiZulu culture. Young women are expected to adhere to the rules of engagement, which generally require docility, even within formal institutions like government municipality. In the interviews, the black, female early career researcher experienced invisibility, often being talked past. Her positionality hence limited her ability to participate fully and collaborate in interview dialogue. The cross-cultural nature of the eThekwini case led to such outcomes in the data collection process, where inequalities were reproduced through the research process. Research into just transitions ought to enhance opportunities for both researchers and research participants to recognise and overcome their marginalisation, respectively.

Some of the cultural aspects of the case also surfaced in the data, providing explanation for certain features of the case. For example, our data suggested that attracting women into the education and training, and employment opportunities around Durban’s port has been a significant challenge. This could be because of deeply entrenched patriarchal views regarding the kind of work that women can and cannot do. Port work is associated with manual occupations needing, for example, mechanical or electrical skills among other technical skills, which remain predominantly viewed as male-oriented skill sets. This determined the extent to which women featured in the workstreams conceptualised within the research method on the eThekwini case, explaining why it was hard for our team to access certain demographics for interviews.

Up to this point, the focus has been on the elite interviewing process, with an emphasis on powerful actors in the conventional sense of power and the intersections between the positionalities of these actors and those of the researchers in the eThekwini case. However, Eyben’s other forms of power were present amongst actors trying to make the stated vision of infrastructural and capacity development work. This was mainly concentrated within civil society organisations. The early decision to conceptualise the study as a set of embedded cases with different units of analysis allowed the Wits research team to deep-dive into the role played by smaller networks of individuals within otherwise hidden parts of the eThekwini skills ecosystem, for example the South Durban Community Environmental Alliance. This organisation works with the artisanal fishing community, among other marginalised groups within the eThekwini municipality. Their work created opportunity for a co-learning site wherein innovative, informal approaches to skills development were being exploited. So, while our Wits team initially took a research approach that was mainly linear, being in dialogue with organisations like the South Durban Community Environmental Alliance presented opportunities to allow our research process to take more of a reflexive slant. As we learnt more about the organisation, it became apparent that its place in the skills ecosystem positioned it well to serve as a bridge between powerful actors and marginalised communities. If, based on policy rhetoric, the Wits team expected the state to drive the bridging process between actors, the data disrupted the expectation by showing a disjuncture between rhetoric and practice. What emerged was that rather than being embedded in the eThekwini skills ecosystem, the organised environmentalists in the community went unacknowledged by the municipality. This could be because environmentalism is viewed by some as having nothing to do with the port, which is associated with industrial maritime activities. Put differently, the former is regarded as part of a set of soft interventions, and the latter hard interventions. The result of this perception is that environmentally oriented initiatives are not systematically and institutionally supported.

As our research team moved through the range of interviews, it became clear that there was a need to develop themes to analyse our data for more bottom-up skills development opportunities within the case. To do so, we drew on the literature around just transitions, working with concepts that illustrate the complex pathways between learning and work. These complex pathways bring into focus the semi-formal nature of education and training arising from organisations like civil society.

Although the research developed in this way, in focusing mainly on the core municipal skills ecosystem in the earliest stages of the fieldwork and thematic analysis, and particularly on the highly formalised and internationalised maritime sector, on reflection our research overprivileged the realities of the more conventionally powerful and lacked the disruptive power that we began to unleash in Gulu.

Our distance from those we were researching was confirmed not just by our race and gender, but by class, culture and language. This was most apparent when we were engaging with public college students. This was done by our young male researcher, who, though of Zulu origin, had also grown up in a Johannesburg township. As he sought to engage with potential respondents, he became very aware that most VET students in eThekwini experience multidimensional poverty, which was being exacerbated by COVID-19 at the time of the research. The pandemic and lockdown also made research access much more challenging and contributed to the distance we have already described between researcher and informants. Through social media, he managed to contact the President of the relevant Student Representative Council and 11 other students for online interviews.

However, in the process of accessing the students, he became aware of a linguistic challenge. Whilst he was much more confident in conducting interviews in English, he was aware that their English was quite weak and that it would be better to conduct the interviews in isiZulu. This made him very conscious of his limited fluency. Moreover, in the interviews though he tried to ask open-ended questions in order to create the space for the students to speak freely, he was hampered by his lack of fluency.

In writing this paper, we discussed whether this lack of fluency had served to overturn the power imbalance between researcher and informant. However, he thought this effect was limited. Rather, his position in an elite university took centre stage, and students positioned him as a success as a result; whilst they saw themselves as failures, reflecting the low status of public VET provision.

However, finding themselves engaging with someone who was identifiably like them in terms of age, race and township upbringing did do something important to the implicit nature of the research interview contract. Much of the interviews’ time was deflected from the extraction of data from them and towards trying to extract from him information about how he had ‘made it’ (cf. Powell and McGrath 2019 for a similar experience).

More negatively, we might describe this as a desperate expectation among the VET students that the research project might create opportunities for them to advance socio-economically. As highlighted in the Gulu case, there is always some amount of disparity between the expectations of the researchers and informants that they interact with. Again, similar to the Gulu case, while the research team may have attempted to express the goals and objectives of the research, a few individuals still expected more than what the research could offer.

Discussion

What then do these two very different cases of research processes tell us? First, of course, they highlight how deeply contextual and contingent research is. Whilst parts of the same project, the cases took on designs, interactions and outcomes of their own that reflected the complex agentic sets of the case teams and wider project collective, the complex structuring of the case contexts and the institutional and disciplinary traditions of the teams, and the agency of informants who themselves came with very different resource sets.

The socio-economic context shapes the context of research, and research has potential to problematise and shift the socio-economic narratives related to the livelihoods and lives of the participants, both researched and researcher. VET is deeply connected to the lives and livelihoods and wellbeing of societies. In both cases, we found narratives that built stigma about VET and those participating in it. This was at least partially founded in the hierarchical formations of decision-making, in which researchers occupy elite positions and make judgements from the outside about the meaning and aspirations of the VET participants, usually without them. The research was designed to answer particular questions from the particular perspective of the researchers, thus creating boundaries that isolate other interpretations.

It appears that the different methodological traditions that the two teams worked from had important effects. In both cases presented here, as researchers began to engage with the cracks, a different story emerged, although more quickly and clearly in Gulu. Particularly there, we grew in understanding that boundaries had to be crossed, different questions had to be asked, new pathways explored, and the complexities of life better considered. This leads towards a fuller and more realistic picture.

Rather than a tightly controlled project designed by a Northern-based PI, we had considerable negotiation and local decision-making in the research design, reflecting the seniority of the Southern project leadership team. How this played out depended crucially on the make-up of the case study teams. However, how the research unfolded was hugely shaped by interactions in the field, and how these were shaped by various forms of power that were deeply contextual. There was a big difference in the responses we received in our research according to who was deciding which questions needed to be asked, and who was posing the questions to whom. Considerations of language, cultural and social identity, position and gender were reflected strongly in both cases in relation to the framing of the research.

It is possible that research can purposefully pay attention to questions of power and positionality from the outset. In so doing, it needs to build in opportunities and spaces for reflexivity amongst the team and with the stakeholders. Whilst this has classically been done through advisory committees, the Gulu case points to the powerful effects of learning through ongoing participation in the daily lives of communities that researchers are embedded in.

In so far as VET research is for the youth, the unemployed, or the marginalised, then we need to include them, to know them, to work with them beyond the temporal and contextual boundaries of a research grant. Engaging in an openness and attentiveness to difference from our own habits of perception and communication can help uncover hidden and important aspects of research. However, to truly engage, research needs to be more than a fixed term activity. It requires a concerted effort to learn together, be part of the social ecosystem, develop and facilitate networks, concerted and strategic goal of exacting change, focus on strong and dependable relationships. Researchers and institutions must be dependable. While the social skills ecosystem allows for the formation of communities of learning, it is important to recognise the existing learning networks and build on them over long periods of time.

And time itself is an important element of research. There is a strong tendency for research projects to be time-bound, reflecting the exigencies of funding. At the same time, research exists within wider temporalities. In the more formal setting of eThekwini’s municipality and maritime industry, time is carefully parcelled out into the conventional hour blocks of meetings and interviews, with the length of time ‘granted’ by elite participants to interviews a reflection of their power. Not in this project but at other times in South Africa, members of the team have been granted 30 or, on one memorable occasion, 15 minutes for an elite interview. In Gulu, as we have seen above, time is less formal, particularly away from international and state agencies. Respect here is not about minimising the time cost to the respondents but giving them the time due to them.

In the more informal setting of Gulu and in the precarious spaces occupied by public VET students in South Africa, research interactions were not simply about the extractive impulses that can infect conventional research. Rather, those consenting to be interviewed were implicitly or explicitly also looking for knowledge and other resources. Many of the two research teams were not so distant from the respondents in terms of age, race or background but the most important effect of this was not a reduction of power imbalances but a raising of the question of what strategies had worked for them.

Moreover, Gulu University’s greater closeness to the community, and not just spatially, meant that the research was also more about an engagement between actors in a skills ecosystem, with reasons to respect each other. In eThekwini, by contrast, the ecosystem was fractured by mistrust and the personal embeddedness of the research consultant contrasted with the outsider status of an elite university from another city.

There is a danger in VET research that it can concern itself with formal knowledge and a knowledge hierarchy in which VET, its students and its workers are placed in low positions, in contrast with state officials and business representatives. Although much of the literature in JVET seeks to position itself against this tendency, we argue that our reflections here help further develop an argument about the need to see VET as existing in a much richer knowledge economy. In both the Gulu case and the Alice one not featured here, there was a very strong sense of the knowledgeability of the actors being engaged with that, in turn, transformed the researchers’ knowledge of VET towards a more expansive view. Even in the eThekwini case, there were hints of this in the way that some of the students at the public college were informal information brokers. However, unlike in the more inclusive skills ecosystems of Gulu and Alice, they were invisibilised.

Conclusion

The worlds of formal VET and formal workplace learning should and will continue to be important sites of VET research globally. However, we contend that this is only a part of a field in need of a broadening to include other learning and working locations.

In reflecting on our own more formally oriented VET research, we have been concerned to be self-critical and point to some of the challenges that we faced in doing such research. We think that these challenges are more widespread but relatively understated in the current research literature. We believe that these reflections add a further dimension to the recent questioning by Avis and others of VET research that head in a decolonising direction. Our vantage point of being a diverse team working in the complex and contested context of South Africa is a valuable one as it makes such issues difficult to ignore.

Alongside this, we offer a case of a very different type of VET research. We are conscious that this might sound less self-critical and we are aware of the very particular set of circumstances that facilitated the research process. Nonetheless, we argue that there is a need for more research of this kind that is about longer-term commitment to and engagement with communities of workers, learners and citizens.

What we see is a growing move in VET research in Africa to work with vulnerable and marginalised groups seeking to overcome intersectional disadvantage. For such research to be meaningful, it needs to spend more time understanding the life-crossing challenges and aspirations of youth seeking out learning to build their futures. Whilst our cases are African, we believe that many of the issues are more widely relevant.

This paper is a very fallible attempt to open up the debate about the future of research further through, we believe, a critical reading of some of our failures as well as successes in trying to ground our research ethically, ontologically and axiologically and not just methodologically. We are conscious that radical embeddedness of VET research in communities is an ideal position, which is applicable only to some parts of a comprehensive VET research agenda. We also acknowledge that employers and the state are legitimate stakeholders who should be part of research. Nonetheless, we believe that our reflection on our successes and failures in these two cases and the project as a whole offers useful provocations regarding ways of making VET research more reflective of diverse settings, less extractive from those being researched and more equal in the participation of members of the research team from the South.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

The work was supported by the Global Challenges Research Fund.