- 1 For the sake of convenience and ease of reading, we use a collective pronoun when referring to rese (...)
1With rising academic interest in heterodox perspectives on international development interventions, including their unintended consequences and so-called “afterlives” (Rudnyckyj and Schwittay 2014; McKay 2012; Gez 2021), identifying suitable research approaches to capture such complex aftermaths emerges as a new challenge for development studies. In what follows, we share insights from our attempt to adapt an existing research approach to generate evidence on development’s afterlives, including the cascading and unintended effects of past development projects, both tangible and intangible.1 In particular, we explore the valence of qualitative-ethnographic methods for diving deeply into the plurality of projects’ lasting legacies, well beyond their formal temporal and geographic boundaries.
- 2 English: Rapid Collective Investigation for the Identification of Conflicts and Strategic Groups.
- 3 Out of 28 participants, 26 contributed essays and 2 participants did not contribute essays but assu (...)
2This article focuses on a collective qualitative research approach known as ECRIS (Enquête collective rapide d’identification des conflits et des groupes stratégiques),2 building on its core interpretative dimension and interdisciplinary exchanges as pioneered by development anthropologists Thomas Bierschenk and Jean Pierre Olivier de Sardan (1994; 1997). We share our reflections on a research field training focused on the remains of a Finnish-Kenyan water-infrastructure collaboration called Kefinco (early 1980s to mid-1990s). For nearly two weeks, we—this article’s authors—participated in a variation on ECRIS that centred the spirit of exploration and flexibility underpinning the approach. Our collective field research concentrated on ethnography, with special attention to questions of positionality vis-à-vis the field and vis-à-vis others within our highly heterogeneous team. We foreground the reflexive component of our team’s short ethnographic exercise, including analysis of on-site informal discussions and daily debriefings, and prioritise it over ethnographic data. In addition, after the workshop, all participants agreed to compose an essay in response to the open question, “What struck me in particular about the workshop experience?”3 Taken together, these individual essays—some of which were ethnographic, others evaluative, yet others conceptual—were used to engage with the following question: “Taking into consideration the growing interest in the study of—and, more broadly, engagement with—the afterlives of development interventions, how may such explorations be done in a manner that is collaborative, comprehensive, and sensitive?”
3Written from a methodological perspective, the article relays the epistemological journey that our team underwent trying to make out the “thing” called Kefinco, and builds on these experiences to make the case for collective ethnographic research of former development projects. In particular, we emphasise “multi-scalarity”—a term we use as shorthand for encompassing several modalities of temporal and geographic investigative expansions that are both methodological and epistemological. Our approach to bringing together multi-scalar material underscores the importance of resisting single-story simplifications,4 and highlights the need to present Kefinco’s legacy as a complex tapestry of both documented achievements and embodied experiences. The set of answers—and additional questions—provided in this piece are far from conclusive, but rather are an invitation for further creative adaptation of the ECRIS approach.
4The article is organised as follows: We start by exploring the ECRIS approach as originally conceived and providing a bird’s eye sketch of the formal “life cycle” of Kefinco as extracted from official records, concluding the first part by presenting our workshop, its rationale and structure. The second part explores elements unique to the application of the ECRIS to the context of development’s afterlives. In particular, we dwell on two crucial points: the introduction of multi-sited, multi-scalar external material (in our case, collected both in Kenya and in Finland) and the evocative potential of such a workshop to summon the ghosts of bilateralism, including a hazardous stoking of unrealistic hopes for the project’s “return.” Building on this multiplicity of perspectives and data points, the next section then takes a more ontological approach, asking: What, in fact, can we tell about Kefinco in its post-intervention form? To tackle this question, we draw on Claude Lévi-Strauss’ notion of floating signifiers (Lévi-Strauss 1968 [1950]) and on Edgar Morin’s writing against simplification and in embrace of complexity (Morin 1999).
5ECRIS is a collaborative, interdisciplinary, and ethnographic research exercise developed by Olivier de Sardan and Bierschenk in the 1980s and 1990s. The approach focuses on interfaces of miscommunication within a collaborative research context. It weaves together information and various stakeholders to create a variegated social tapestry. Drawing on their own exploration of development interventions in Western Africa, Bierschenk and Olivier de Sardan concluded that, “in our opinion, this method is particularly adapted to the analysis of development projects” (Olivier de Sardan 2005, 192). Focused on the analysis of power relations, the approach was initially defined around core interpretative disciplines in social sciences (anthropology, history, sociology, and political sciences), with the explicit intention of contributing to inclusive policy-making.
6ECRIS brings together three key notions: conflict, arena, and strategic groups, with the underlying idea that groups, conceptualised as non-essentialised collections of stakeholders, might be formed around shared interests regarding the appropriation of resources. Bierschenk and Olivier de Sardan conceptualise these strategic groups as a “working hypothesis for the researcher, a sort of ‘virtual group’ which helps us to understand the convergence of strategies between certain individuals who can be assumed to share the same position in the face of the same ‘issue’” (Bierschenk and Olivier de Sardan 1997, 241). However, strategic groups “are not formed for once and for all with a universal relevance for all problems” and “there are no rigid boundaries between strategic groups” (ibid.). Part of the ECRIS process, therefore, lies in the iterative refinement of strategic groups by participants in light of emerging data points and fresh reflections.
7In developing the ECRIS approach, its authors reflected on the importance of fieldwork as a gradual and intensive learning process. They were intent on transforming misunderstandings between researchers and interlocutors as well as among fellow researchers within a joint programme, turning interpretative differences into valuable learning opportunities. In addition, they emphasised keeping records of the field experience and analysing the material collected. The final product is the result of grounded theory (Charmaz 2006): an inductive process that draws on fieldwork data and, through an iterative process of joint collection and reflection, transforms into new findings. Far from the trope of the lonely—and, historically, primarily white and male—ethnographer, the ECRIS approach allows diverse teams of researchers to collectively identify as much primary information as possible within a short time. A variety of social actors are approached to capture the subject of inquiry from as many angles as possible. Findings are shared and discussed collectively to raise questions for further research. In practical terms, the larger team of ECRIS participants is divided into smaller teams that interact with specific strategic groups, subject to continuous and dynamic reassessment. Collective debriefings are key to the ECRIS experience: each team reports back, and a general discussion ensues. This exchange may, in turn, result in the reframing of the wider team’s perspectives, approaches, and questions. In other words, data gathering and debriefing are tightly intertwined, which allows for the continuous refinement of categories and of the guiding questions.
8During our ECRIS, while participants were at liberty to move between teams, and each team was free to redefine their target strategic group, in practice, teams and their compositions remained fairly stable throughout the week. Participants may have been inadvertently nudged by the facilitators towards pre-identification of categories due to their engagement with archival material and interview excerpts before fieldwork, more on which we discuss later. In any case, these ad hoc definitions of strategic groups helped to locate, and interact with, a wide variety of actors related to the past development intervention or affected by it, to grasp the heterogeneity inherent to each category, and to guide the complementarity of our teams’ discoveries. On the first day of our workshop, participants identified the following strategic groups: (1) direct and indirect Kefinco employees; (2) landowners, housing, and business owners; (3) water management institutes and ministries; (4) academic institutions; (5) water point users; (6) the Kefinco Estate and the project’s local management; (7) non-Kefinco Finnish-sponsored/Finnish-related initiatives; (8) unrelated community members. Workshop participants then opted for their strategic group of choice, with three to four participants in each team. Mindful of language, each team had at least one Kiswahili speaker.
9In applying ECRIS to a post-development context, our study faced particular temporal challenges. Whereas ECRIS typically explores unfolding dynamics in real time, our task was to uncover narratives and remains from a bilateral project that had concluded nearly three decades earlier. This required attention to the composition of our team and the introduction of a multi-sited and multi-scalar approach, combining historical records, archival research, and field interviews to capture both official and informal narratives about Kefinco’s legacies. Seeking to minimise knowledge gaps between those workshop participants who were new to the case study and those who had been involved since before the workshop, our ECRIS included a significant preparatory component as outlined below.
- 5 While plans to include Kenya in the sphere of Finnish development cooperation were in motion alread (...)
- 6 According to archival material from the Finnish Ministry for Foreign Affairs and interviews with fo (...)
- 7 It is noteworthy that, while much of Kenya consists of drylands, western Kenya and Kakamega in part (...)
10Before delving into the methodological reflections of our workshop, a short exposition of the Kefinco project is in order. Kefinco was a Finnish-Kenyan rural water supply development project for the provision of potable water infrastructure in western Kenya. The project, which was set up in tandem with the UN’s “International Drinking Water Decade” (Najlis and Edwards 1991; United Nations 1979) and speaking to Finland’s reputation in water technology and its portrayal as “the nation of wells,” ran from the early 1980s until mid-1990s.5 Divided into seven phases, it centred on the construction and rehabilitation of water points, including shallow wells with hand pumps, protected springs, boreholes, piped systems, and water plants. Set up as a bilateral agreement, the project was primarily financed by the Finnish Ministry for Foreign Affairs through the contracting of Finnish consultancies, with a smaller portion of the funds supposed to arrive from the Kenyan government.6 The project focused around Kakamega—a town in Kakamega County—and eventually extended to nearby counties including Bungoma, Busia, and Siaya, all in western Kenya.7
- 8 Information based on archival research and interviews conducted with a dozen former Kefinco staff a (...)
11Kefinco was highly ambitious, promoted through slogans such as “water for a million,” and at the time constituted Finland’s largest development investment in Africa. In preparation for implementation, a gated estate was set up in Kakamega Town—“Kefinco Estate,” located in a neighbourhood that is still known locally as “Kefinco.” The estate, which was composed of 20 detached residential houses hosting the Finnish workers—primarily engineers—and their families, was initially well kept. It consisted of two broad and green avenues and included a clubhouse that served as both a restaurant/bar and a cultural centre, a swimming pool, and a large detached sauna. To the best of our understanding, at its peak, the project boasted a fleet of about 30 SUVs, 50 motorcycles, lorries, trucks, farm tractors, and two drilling rigs as well as a recreational speedboat.8 On the Kenyan side, the project contracted dozens of locals as drivers, mechanics, and engineers, and many others as daily labourers to assist with construction. Overall, according to the Finnish archives, by the end of the project, it had dug and established a total of 3,714 water points, which include 1,177 boreholes, 1,229 shallow wells, 1,217 protected springs, and 91 piped schemes (Simojoki and Simojoki 2000).
- 9 From 1986 onwards, land easement forms were signed between a landowner and the Ministry of Water an (...)
- 10 The centralised model seems to be in line with a previous, comparable Finnish water project in Tanz (...)
- 11 Information based on interviews conducted with a dozen former Kefinco staff and related personnel i (...)
12Throughout the project’s lifespan, it transformed several times over. This was the time of growing professionalism within the international development sector globally (Mosse 2022; Unger 2018), and specifically in Finland (Kuhanen, Harju, and Hokkanen 2023; Koponen et al. 2012). This is evident in the archival traces that the project’s team left behind, which throughout the 1980s became increasingly elaborate. In its early years, the project appeared quite centralised and top-down, but over time refined a “demand-driven” model of interventions whereby people could apply and request Kefinco’s involvement to build a water point for the public good (indeed, many such water points were built in proximity to schools).9 Relatedly, the project developed a community participation component, notably through the establishment of well committees and the purposive involvement of women.10 This change of approach was driven by a bottom-up approach, in large part through the involvement of Julia Kunguru, a Kenyan woman who worked for the project as Community Participation Officer from the mid-1980s until the early 1990s.11 Over time, the project was supplemented by other Finnish-led initiatives: some were official, such as a network of health centres and a dairy farm project, and others as a result of private initiatives by Finnish employees, such as the opening of a school or a Pentecostal Church opposite Kefinco Estate.
- 12 The percentage of Finnish development cooperation stood at between 0,7 and 0,8 of GNI in 1990–1991 (...)
- 13 Agreement Between the Government of the Republic of Kenya and The Government of the Republic of Fin (...)
- 14 Archival material not available beyond that year due to classification rules.
13The project’s eventual termination and handover, which took place in the mid-1990s, can be attributed to several factors. One of them appears to be the economic crisis of the early 1990s across the world, and in Finland in particular. Prior to 1990, as the Finnish economy was booming, the project had a seemingly unlimited access to funds. However, on entering the new decade, Finland fell into a grave economic depression that lasted up to 1994, resulting in serious consequences for development cooperation.12 The crisis also eroded Finnish tolerance for the Kenyan government’s non-compliance with the provision of complementary funding. In the Finnish archives, one can see a marked change whereby, while in the 1980s the gap in local funding was frequently compensated through higher contributions from Finland, in the 1990s this was no longer an option. At the same time, it is possible that the project was heading towards termination even if it had not been for the economic crisis, given its already long life cycle. Indeed, the project well exceeded the original proposal of a five-year duration13 and was often greenlighted to continue to a new phase. Kefinco’s phasing-out stage was long, with Finnish actors still directly and indirectly involved well into the 2000s.14
- 15 This initiative was led by IFRA_Nairobi in collaboration with the IRD, INED and CNRS. It incorporat (...)
- 16 The framing of the event a training rather than as fieldwork was emphasised throughout the workshop (...)
14Our workshop built on previous methodological schools held in East Africa since 2006 and which were similarly inspired by ECRIS.15 Organised in the context of ERC Project AfDevLives, the workshop was designed with three key objectives in mind: testing and honing a variety of research skills and synthesising disparate approaches and information; collectively exploring the afterlives of international development projects through a specific case study; and team-building. Moi University, Eldoret, acted as a local host for the programme, and a grant from the IRD enabled the expansion of the workshop to include more East African participants. External trainees were selected through a competitive call.16 The emerging team included members of Project AfDevLives based at Iscte – University Institute of Lisbon—PhDs, postdocs, members of the advisory board, and the project’s Principal Investigator—as well as recipients of participation scholarships and a handful of guests. The resulting team was highly diverse in origin, academic seniority, and disciplinary orientation, a point that we considered a strength. As one trainee wrote in their essay, “this diversity enhanced rather than threatened the collaborative learning that we experienced in the program.” While, in the spirit of co-learning, all participants equally engaged in the activities and the atmosphere was largely informal, there was a distinction between the ECRIS’ organisers and its joining participants, whom we refer to here as trainees.
15To achieve our goals, we freely adapted the structure of the workshop and divided the activity into the following three sections: First, we began with three days of preparation at the French Institute for Research in Africa (IFRA) in Nairobi. On these days, the team was offered first-aid training, a presentation about the research project and about Kefinco (in the context of Project AfDevLives’ inaugural event), basic training in research methods, and a full day’s training in research ethics. This latter training, which highlighted real-life ethical dilemmas and employed hypotheticals and role play, was essential considering how the afterlives of development are tied to affects such as expectations, frustration, and nostalgia, as well as the trainees’ rapid plunge into this new field. In addition, some ethical dilemmas and challenges are specific to collaborative work. As one of the ethics trainers noted, “from an ethics perspective, collaboration means less opportunity to promote privacy and the risk of loss of confidentiality is high.” Less obvious but similarly significant was the first-aid training, which was not thematically connected to the workshop but was highlighted by trainees as particularly helpful for teambuilding. One trainee noted how, “throughout the day, participants were using first names, laughing together, and “treating each other’s injuries,” thereby helping in “establishing horizontal power relations, bolstering inclusion, and breaking the ice.”
16From Nairobi we travelled together to Kakamega by bus for the full-week workshop, during which we pursued a model of morning sessions followed by fieldwork in small teams, and an afternoon debriefing. The debriefing sessions were highlighted by our participants as exciting moments of learning and sharing, but some also noted that they had a tinge of stress-inducing, extractivist competition: which team produced most information or made the most extraordinary discoveries today? Beyond this basic structure, additional ad hoc activities were introduced in response to concrete needs and turned into learning activities (e.g., a silent writing session the morning after our first, intense full day of fieldwork). When a key Kenyan interlocutor made herself available to speak to our team online, we quickly shifted our programme to include a pre-interview session to discuss the structuring of interview guides as well as a post-interview session for analysing the interview dynamic.
17Throughout our week in Kakamega, and in line with Project AfDevLives’ emphasis on creative and phenomenologically inspired methods, many of the sessions were conducted “on the basis of shared but also diversified methodologies” (Charton and Médard 2006). Inspired by the diverse disciplinary and personal backgrounds of our participants, such exploration was prominent in the workshop—a fact that informed, perhaps, the feeling of open-endedness and freedom attested to by many participants’ essays. At the start of our week in Kakamega, we were led by experienced team members in on-site training in drawing and photography. In particular, the session on the use of graphic methods registered within the team and inspired participants to experiment with drawing in the field (Azavedo and Ramos 2016). Other visual methods used include 360-degree photography and 3D scanning of objects.
18Several participants used walking interviews, which have previously been highlighted as especially useful for studying the afterlives of development projects (Gez 2021), while other methods included observations, in-depth interviews, key informant interviews, unfocused observation, focus group discussions, and GIS mapping. For most participants, this was the first time they were being trained to conduct fieldwork in such a collaborative manner, and for the external trainees in particular, the very focus on the afterlives of development was brand new. As one trainee wrote, “Before the ECRIS training workshop started, I was mostly curious to learn about afterlives of development as I had never heard of the term till then. With a background in health and development studies, I had learnt about various interventions that had been proven to work in communities, but hardly did I think about the ‘afterlives’ of that intervention and whether the impact that had been predicted materialised or not.” The facilitators were struck by the intuitive resonance that the concept had amongst the East African trainees.
19In line with the ECRIS approach, our week in Kakamega was characterised by continuous back-and-forth between fieldwork and reflexive debriefing, where questions of positionality took centre stage. This is hardly surprising considering that the participants came from highly diverse backgrounds, and were divided almost equally between participants from the so-called South (including twelve from Kenya) and from the so-called North (including two from Finland). While the majority had a background in anthropology and sociology, a minority hailed from disciplinary backgrounds as diverse as geography, history, health, and even planning and architecture. Standing out as short-term visitors and largely viewed as outsiders, the participants were highly conscious of their positionality throughout the event. This became clear during our debriefing sessions, which featured common discussions about our team’s positionality in the field and its impact on the data collected and on ease of access, as well as about relations to the host communities and to each other (see Figure 1). This reflexive process was in no way limited to the participants being perceived as foreigners (wazungu), but also reflected inner-outer ambiguities shared by African participants, notably among non-Kenyan African participants and among Kenyan participants living abroad.
Figure 1
Drawing by Manuel João Ramos, produced in the debriefing room during the ECRIS and reflecting the tone of the discussion as well as his own feelings as a foreign (mzungu) researcher
- 17 While it is not our purpose, in this present paper, to comment on such perceived association betwee (...)
20Beyond stereotyping of contrasting positionalities among European and African researchers in African contexts, our team’s experiences in their strategic groups and in the workshop as a whole was more intricate. Certainly, one of us—a Kenyan university lecturer—was mistaken for the driver of his European colleagues upon entering Kakamega Golf Club—a private, exclusive members’ club that was once a central hangout for Finnish Kefinco staff outside their residential compound. Yet, as a rule, the configuration of the insider-outsider dichotomy shifted depending on the context. One of our teams, which consisted of Kenyans and one non-Kenyan African participant, shared their sentiments that their positionality as supposed insiders helped to navigate local interactions—in their case, in relation to water authorities. Another team, which investigated connections between Kefinco and the local university, felt that their positionalities as a mixed Kenyan-European team significantly intersected with their professional profile as academics. On the whole, we were under the impression that participants from the so-called North were assumed to be associated with the development apparatus, contributing to a perception that they were “returning” to revive the bygone water project, as we discuss later.17
- 18 For example, one team met a former technician who had worked with Kefinco in the 1990s before trans (...)
21Lastly, at the end of the workshop proper, we closed with a public symposium hosted at Moi University. The symposium brought together former employees of Kefinco on both the Kenyan and Finnish sides, external academic experts, and all members of our workshop, who presented their work in their respective strategic groups. While most symposium presentations were arranged in advance, several speakers were invited to speak spontaneously, following encounters in the field.18 The event had as its purpose not only to share our teams’ experience and lessons to a wider audience, but also to build a bridge of dialogue and kick-start a conversation with other researchers and with former project actors in both Kenya and Finland, several of whom joined us online. Among other things, it showcased the potential of ICT tools to reanimate conversations around bygone interventions.
22In their reflections during and after the workshop—including in their individual essays—the participants evaluated the experience as enriching and, at times, eye-opening. This assessment referred not only to the broad approach to projects’ afterlives, which some saw as an intriguing departure from the clear-cut classification of projects into either “success” or “failure,” but also to the experience of “learning by doing,” to the diverse backgrounds of participants and methods, to the balance between time for (collaborative) fieldwork and time for (collective) reflection, and to the gradual construction of the event that included both preparatory and concluding activities.
- 19 While the ECRIS design focuses on ongoing projects, it is interesting to note that Olivier de Sarda (...)
23Earlier, we discussed Olivier de Sardan and Bierschenk’s (1997) conception of the ECRIS approach and its application to the study of development projects. This application is oriented around contemporary dynamics, where it is assumed that a variety of actors are concurrently operating within the same arenas. But how might this approach be applied to a case like Kefinco, which has long given way to post-project entropy, and where “practices” long turned into “legacies?”19 In the case of our ECRIS, the adaptation to the study of a former intervention was marked by specific elements, some planned and others less so. It is to these adaptations that we turn our attention to in this section.
24One central, planned element in the adaptation of the ECRIS was the inclusion of multi-scalar perspectives, unbound geographically (Kakamega municipal limits, Kefinco’s reach) or temporally (present-day focus). We use the term “multi-scalar” as a shorthand encompassing several modalities of investigative expansions, both methodological and epistemological: a multi-level analysis up and down the hierarchy of the development apparatus and its so-called “policy chain” (Bierschenk 2014); a multi-sited study (Marcus 1995) that included previous fieldwork in both Finland and Kenya; and multi-temporal exploration of layers of history, uncovered through archival research, oral histories, and “tracing” (Geissler 2023) of cumulative layers of projects in and around Kakamega. Indeed, some members of our team had already begun to examine the afterlives of Kefinco the previous year, when they conducted a preliminary visit to water points and to the Kefinco Estate. Later that year, they embarked on archival research at the Finnish Ministry for Foreign Affairs and at the Kakamega Records Office under the Kenya National Archives, a parallel work that allowed the fleshing out of continuities and divergences of narratives. This archival work was complemented by in-person interviews carried out in Finland, with a dozen Finnish development workers and former diplomats who were involved with Kefinco during its implementation.
25While the facilitators among us saw the value of these entanglements and were curious to see how the material collected in one context would transpose into the workshop setting, they were also wary of unintentionally imposing an “authoritative version” of events on the trainees, which would in turn overshadow the diverse perspectives presented through fieldwork. Thus, they did not prepare a reference document recounting the history of Kefinco, but they did share one with a basic timeline for temporal orientation, followed by a selection of archival cut-outs. This material, distributed ahead of the event, was to provide participants a baseline knowledge about the project while also presenting them with a starting point for questioning and a point of reference for analysis. Similarly, the facilitators drew on the expert interviews from Finland to offer our first ECRIS exercise. In the exercise, thematic compilations were drawn from qualitative analysis categories of these (anonymised) interviews. Based on participants’ interests, they were divided into thematic teams, where they were asked to read their compilation of choice, discuss it, and write down follow-up questions. These two compilations—the archival compilation/official timeline and the interview-excerpts exercise—were a risky choice as they drew on material external to the ECRIS and on remote experiences that could potentially eclipse local ones. However, the variety of sources proved complementary. Archival records painted an objective-oriented picture of the project, often focused on technical achievements and infrastructure metrics, while excerpts from the interviews in Finland offered project assessments and anecdotes that fleshed out everyday experience around Kefinco as recounted from a standpoint that would have otherwise been inaccessible to the ethnographic team. By contrast, local voices in Kakamega—direct users of water points, officials, and witnesses to the project and its consequences—offered personal, relational, and contemporary perspectives that are not referenced in official reports. Line by line, our approach thus wove together an intricate tapestry of data points, some of which were complementary and others stood in tension with each other.
- 20 This refers to a rumour that we have not been able to substantiate whereby, after Kefinco’s closure (...)
26Throughout the workshop and in the final essays, the participants noted the uniqueness of this multi-scalar approach. As one wrote, “Instead of going out into the field in Kakamega with no prior knowledge of the situation at hand, we had so much information going into the fieldwork, and this was unlike any research I had ever done before.” Participants noted that the external material provided a rich background and painted a picture of 1980s and 1990s Kakamega, and of the project. This baseline information was also useful in providing pointers for an empirical strategy, including key locations and possible avenues for identifying informants. At the same time, one participant recognised that the prior sharing of material “could have created presumptions on a couple of issues, such as ‘the buried motorcycles’, such that even as we ventured in the field there may have been some biases.”20 Overall, however, the trainees seem to have appreciated how reliance on multiple types of sources offered enriching, constructive orientation. As one participant observed:
27“The layering of the archives, the interviews with stakeholders that have been involved in the implementation of the development initiative and bringing it back to the same place where it was implemented in the present—created these dense moments of contact in which simultaneously there was a historization of the present, and the meeting of the imagined future of those implementing the project and the presence that actually came to be. Coding the interviews with the staff and being aware of the discrepancies between the local and Finnish archival accounts of what took place in the same spot, inspired the conversations which we had with the community living there today and focused our gaze on specific objects within that space.”
28Another element that struck us as essential to the specific adaptation of the ECRIS to the study of afterlives relates to interactions on the ground specifically regarding the inadvertent awakening of affects and expectations among the local population through our team’s presence and questioning. This point, which—differently to multi-scalar work—we were hardly prepared for, is tied to considerations of positionality in the field, which, as we discussed, is an essential point of reflection in any ECRIS workshop, and not least in ours. However, some of the challenges that the participants encountered in this regard seem to be specific to the study of the afterlives of development and are therefore worth unpacking.
- 21 Such rekindling of (unrealistic) hopes—and, potentially, fears—among the local population is a key (...)
29Specifically, when our team arrived in Kakamega, local interlocutors were curious about our intentions, and many assumed that our questioning about Kefinco was driven by an intention to revive the bygone project. We sought to avoid stoking such expectations by stating the purpose of our visit as clearly as possible—supported by university letters confirming that we were actually in training, not even conducting formal research—and, to the extent that we were successful, our interlocutors responded with disappointment. The interest among the population in the question of the project’s “return,” accompanied by the fast emergence of a class of locals who volunteered themselves as brokers (Bierschenk, Chauveau, and Olivier de Sardan 2002) and as “privileged informants” (Lokot 2021), forcefully illustrated to us that, while the project may long be over, its potential for evoking aspirations of all sorts is undiminished.21
30In this environment of rekindling aspirations from the development sector, and on the backdrop of the bilateral Kenyan-Finnish axis of collaboration, positionality proved all the more prominent within our mixed group. As one of our trainees noted in their essay, throughout the workshop, “regardless of their actual citizenship, researchers from the North were collectively perceived and referred to as the ‘Finnish’ by both the local community and former employees due to Kefinco’s exclusive association with Finland.” In particular, our two Finnish participants—both social scientists and neither having direct connection to the Finnish development apparatus or to Kenya—became subjects of expectations. At times they were asked to ”come back” and fix unusable water pumps. While their interlocutors presented them with expectations, the Finnish trainees also expected of themselves for something familiar to emerge. For example, when visiting a Pentecostal Church across from Kefinco Estate—an initiative by a Finnish missionary who worked as a community participation officer during Kefinco—one of the Finnish trainees wondered if the white and blue colour scheme across the church was an intentional incorporation of the colours of the Finnish flag. At the same time, they also questioned these very expectations: “What felt the uncanniest throughout the ECRIS was the idea that a Finnish person had a special insight and eye to see the traces that Finns had left behind […] that others could not.” This feeling of the uncanny as straddling the thin line between “homely” and “unhomely,” familiar and unfamiliar (Mitchell and Petty 2020), is a mark of the “temporal disjuncture” (Lewis 2016; Davidov and Nelson 2016) that seems to accompany the ethnographic study of the afterlives of development.
31All in all, the workshop experience reminded us that even dormant traces of bilateral relations may be reawakened to unrealistic effects, with what we may call, following growing interest in the “ghosts” of projects and interventions (Aalders 2020; Edensor 2005; Johnson 2013), the “ghosts of bilateralism”: a lingering, “haunting” presence of past partnerships and their attendant expectations. Organisers of such future workshops should be especially mindful of just how triggering and evocative such residues of bilateralism can be for local populations. This, above all, implies operating carefully and sensitively, and considering how to maximise benefit sharing. At the same time, these experiences speak to promising potentialities to work consciously across the two sides of the former bilateral development collaboration, establishing a new bridge on the remains of another.
32Despite the preparatory work outlined in the last section, the trainees arrived in Kakamega with little knowledge about Kefinco and what they may find. For some of them, this was the first time reflecting on development interventions beyond a positivistic-functionalist approach. As the workshop progressed and evidence began to accumulate for Kefinco’s diversity of cascading consequences and shifting conceptions, this technical approach gave way to fundamental ontological questions: What is this “thing” called Kefinco? In what sense does it actually exist? And if it does, where are its limits? If such a line of questioning appears overly abstract, it spoke to the multiplicity of data points, perspectives, and positionalities, as well as tensions and gaps outlined above between material collected before and during the workshop. All of these made it difficult to pin down Kefinco in its present, post-intervention form.
33As the team went deeper in its training, we discovered diverging stories about the project’s afterlives that extend well beyond its original objectives. We learned, for example, that in Kakamega, one will not run into difficulties when looking for a good plumber—supposedly a legacy from water-related training that many locals accessed as employees of Kefinco and other water-related projects. Stories of occasional, life-changing scholarships, or of a person who received a camera as a gift from a Finnish worker and thereafter became a professional photographer, are also a reminder of the unpredictable ways in which development interventions affect people’s lives. Exemplifying the nuances of Kefinco’s afterlives is an encounter reported by one of our teams: Two interlocutors around a pump erroneously attributed to Kefinco near Kakamega Forest told us how they had used the pump until it had broken down a few years ago. But while the equipment was no longer of direct use, they acknowledged the Kefinco legacy that sensitised the community to come together around water infrastructure and to cede private land for the construction of public water points. At the same time, they pointed out the role of leaders and the difficulty in initiating and maintaining the collective spirit. All in all, their own water problems were not solved through such joint action. Rather, our two interlocutors hired former Kefinco employees who lived locally to build them private wells on their land. As one of them explained, “Kefinco taught us we could build our own wells.”
34In their essays, many of the participants presented their personal process of coming to terms with complexity. Those who started the training thinking about Kefinco in purely binary terms of success versus failure came to “embrace and acknowledge the diversity of narratives and developmental trajectories, culminating in a rich soundtrack,” to quote one trainee. Another trainee wrote:
35“The multiple strategic groups and interdisciplinary teams meant we unearthed different temporalities and perceptions of development. It was significant since no one narrative could account for the aftermath of Kefinco since the project was part of different experiences for people and was situated differently in the past and present for different actors. This discovery was only possible through the ECRIS. Kefinco wasn’t good or bad. Instead, it was part of individual experiences and realities. For instance, for the government, it was one of the many water-related interventions; for part of the community, it stood out as a memory of transformation in their pasts and possibly futures; for other community members, it was a reminder of failed projects relating to water access.”
36Complicating matters further was the fact that, on the ground, the afterlives of interventions are often characterised by a multiplicity of overlaying traces, and indeed, western Kenya has seen countless water-related interventions, many involving similar technologies and operating in close proximity. “How is it,” we asked ourselves, “that we are insistently focusing on the unassuming remains of one passing water intervention while actively pushing back other prominent ones?” This choice, which may seem odd and arbitrary to our local interlocutors, was referred to within our team as the “Kefinco bias.” As one trainee, referring to the particular pump they visited, asked in their essay: “Why was the ‘Kefinco 21×8×1992 NO 853’ water pump given a higher priority over others?” Illustrating the problem of this artificially imposed attention to Kefinco is the following description of an interview conducted by one of our teams with an official from the Lake Victoria North Water Works Development Agency:
37“One of the respondents at Lake Victoria used most of the interview time to talk about his personal history and how he ended up in his current management position. His intervention was not just about him but about how his personal life was associated with the emergence of water management authorities in Kenya, the politics behind it, and the challenges faced in the sector, including lack of personnel, funding, clear vision, and political disentanglement. Noticing that the interview was going in one direction, I kept asking the participant about Kefinco: ‘Did you have any experience with Kefinco? Did you work with anyone who worked for Kefinco? In the interviews [i.e., the excerpts from the interviews conducted in Finland], it was stated that Kefinco is helping in training the newly formed water authorities. Do you have any information about that?’ ‘It was one of the many development projects that existed in Kenya,’ he responded at one point.”
38The administrator’s unphased answer is telling. While Kefinco helped to construct thousands of water points around the greater Kakamega area, it was indeed no more than any of the other water-related interventions that had taken place in the region in recent decades. This multiplicity was especially noticeable in the case of school grounds, where our participants identified repeated and seemingly uncoordinated water-related interventions. Their proximity—and, too often, advanced state of disuse—revealed an unfortunate truth: in the field of development, a new construction is often more readily available than funding for the upkeep of one that already exists. Under these circumstances of multiple layers of interventions, local interlocutors often responded by identifying any number of interventions as Kefinco-related, whether or not they were historically connected to the project. One of the trainees wondered in their essay, “Why would people associate every water pump with Kefinco? Surprisingly, even a water pump that was established between 2009/2010 was a Kefinco water pump.”
39The fact that Kefinco focused on water, with its flowing and shapeshifting nature, provided an undercurrent of symbolism for the challenge of pinning down our subject of inquiry. This image of water shifting reflected the tensions in the data our team collected during the workshop, both with regard to facts—was Kefinco water free or charged?— as well as with regard to framing—were some pumps vandalised out of self-interest (parts stolen for use or for resale) or were they the target of “resistance practices” (Scott 1985)? Considering the breadth of the project, there is hardly a single answer to these questions. Furthermore, the project’s long period of execution, over 15 years, during crucial decades of professionalisation within the development sector, undermine the idea of Kefinco as single and coherent. Against this backdrop, the tensions between the precision of the (mainly Finnish) archival documents—filed with greater and greater meticulousness over the years—and the fluidity of the oral testimonies collected during the ECRIS were especially striking. This fluidity was further complicated by counter-narratives and rumours that, while far from unanimous, cast an altogether different light on the endeavour: In various places in Africa, concern with exploitation by foreigners finds its outlet in the circulation of rumours (White 2000), with many such rumours geared toward the extraction of precious minerals as the true motivation (Onneweer 2014; Gez, Fouéré, and Bulugu 2022). In the case of Kakamega, which has a long legacy of gold mines—including contemporary artisanal mines (Bohbot 2023)—we came across narratives questioning whether the project was not in fact a mere cover story for such an ulterior motive.
40The more our team, with its multiple positionalities and multi-scalar perspectives, discussed Kefinco, the more ontologically uncertain the term seemed to become. What is this “thing” called Kefinco? The term, some participants argued, is used carelessly by both our team and interlocutors, despite lacking a stable meaning. As a neologism based on an acronym, the term embodies both a subjunctive reality—an “as if” or “could be” universe (Seligman et al. 2008)—and a series of concrete interventions. During our debriefing sessions, where teams were running against the clock to share the large quantities of data collected on each day, we began to ask whether “Kefinco” pointed towards a single “thing” out of which, and with which, worlds could be represented, through which ideas could be bundled together, discussions could be had, inquiries could be carried out, and (field) notebooks could be filled in with both words and sketches. An intertextual and inter-graphic analysis of the 28 notebooks of the 28 participants would certainly lead to interesting observations, but the “things” inscribed in those notebooks might be no more than “floating signifiers” (Lévi-Strauss 1968 [1950]), through which 28 minds came together and produced strongly personalised memory supports that, for all intents and purposes, are qualitatively non-transferable (Ramos 2019, see Figure 2). Even though it might be impossible to fix “ultimate meanings” to floating signifiers, people continuously try to arrest the flow of meanings and to establish fixed signification (Laclau and Mouffe 2011, 111). Perhaps, if the ontological question leads us to a dead end, we should instead take a relational approach, recognising the dense proximity of water-related interventions around Kakamega and questioning Kefinco’s de facto interaction with, and impact on, countless other nearby development initiatives.
Figure 2
Drawing by Manuel João Ramos, produced in the debriefing room during the ECRIS and reflecting the tone of the discussion
41As our workshop advanced, questions multiplied and tensions between data points became more and more pronounced. Moving forward required us to work with contradictions: to reconcile with the irreconcilability of our data. In our daily debriefing sessions, after one of the participants suggested treating Kefinco as a floating signifier, another participant proposed the work of French philosopher and sociologist Edgar Morin and his theory of complexity as a lens through which to grapple with the multitude of viewpoints that we encountered. In his writing, Morin fought against what he called the “paradigm of simplification,” a deterministic way of thinking about the world and its parts (Morin 1999, 8) that creates “disjunction” or fragmentation and “reduction” in the form of falsely unifying distinct things (Lorenzi and Andrade 2023). Following three principles—the dialogical principle, the principle of organisational recursion, and the hologrammatic principle—Morin points towards a way beyond seeming contradictions. To some in our group, Morin’s rejection of attempts to enclose reality within a closed and coherent system brought to mind Scott’s (1998) famous observations about the ills of state-led and often coercive designs as governed by positivistic, top-down simplification. Instead of processes of simplification, through which a “synoptic view of a selective reality is achieved, making possible a high degree of schematic knowledge, control, and manipulation” (Scott 1998, 11), our epistemological trajectory in the ECRIS workshop was geared towards a restitution of the loose ends of complexity, unwieldiness and tussled open-endedness that are the legacies of Kefinco.
42These conceptual discussions, and especially Morin’s emphasis on non-linear relationships and unexpected consequences, provided our team with possible frameworks—and, perhaps, a permission structure—for working through the contradictions of the multi-scalar, dialogic nature of the project’s legacies. In our case, these conceptual discussions emerged well into the workshop itself and remained primarily a point of curiosity. In the future, however, such ideas can be integrated into similar events from the onset. With due preparation, frameworks such as Morin’s can help to chart a path for conceding positivistic conclusions without foregoing complexities, squaring formal development visions and positivistic-functionalist assessment paradigms with the unravelling of lived, lasting, and diverging experiences. Drawing on suitable theoretical underpinning, future workshops can work towards multi-scalarity and epistemological plurality, encompassing the tensions that make out the messy afterlives of projects: positivistic and interpretative; global and local; concrete and abstract; tangible and intangible; intended and unintended; formal and deeply personal.
43In this article, we have shown how the ECRIS collaborative research approach can be adapted to studying the afterlives of development interventions. As our experience shows, an adaptation of ECRIS holds potential for tracing bygone interventions in a manner that honours diversity and complexity. With its emphasis on methodological variations, the dynamism of strategic groups, reflexivity, and epistemological plurality, the ECRIS approach is suitable for capturing development afterlives with all their entanglements. The ECRIS’ heterogeneous team shares encounters and insights in a manner that would probably not be available to the single ethnographer, and participants’ positionality—whose local perception can be affected by the lingering ghosts of bilateralism—becomes a trigger for learning.
44In preparing such an adaptation of ECRIS, one challenge may be—as it has been for us—not only to bring together the diverging experiences and opinions presented to the team during the ethnographic training, but to articulate these findings in relation to the epistemological trajectory emerging from our multi-scalar perspective. This includes developing a multi-sited and multi-layered understanding of the project in question, thus weaving together multiple perspectives up and down the aid chain into an analysis that is unbound by the project’s official narratives and life cycle either temporally or spatially. While much remains to be done to stress local voices over official, external narratives, it seems to us that a rigid site-specific perspective comes with a risk of simplification. The fact that Kefinco remained, throughout the workshop, a somewhat floating concept was, perhaps, a sign that we did not overly reify our case study so as to throw out alternative epistemologies. Feeding this plurality, we consider positive, for example, that some factual questions about the project remained unanswered: What did the acronym actually stand for? “Kenyan-Finnish Cooperation?” Or, “Collaboration?” Or maybe it was something altogether different, “Company?” While we incorporated some material external to the field, we also tried to keep it in check, lest top-down voices became so authoritative as to narrow, simplify, and arbitrate what Kefinco “really is.”
45Lastly, in planning such future activities, special attention should be paid to ethical and political considerations. The ECRIS approach as originally designed, sets out to identify group dynamics and power dynamics in the hope of mediating and contributing to inclusive policy-making. In the case of an ECRIS focused on the afterlives of an intervention, this idea of mediation is particularly appealing, as we see the weaving together of multiple data points as a conscious countering of the post-development disconnects between “benefactors” and “beneficiaries”—disconnects that, as this article shows, run contrary to the many ways in which bygone interventions continue to preoccupy people and affect their lives. Such workshops, we suggest, can help to bridge the epistemological gap—solidified over many years of rupture—between the two ends of the bilateral development agreement. Conducted from an independent academic standpoint, we believe that such workshop-based studies can serve as learning opportunities for reassessing past developmental decisions in light of their long-term and cascading consequences. From a practical perspective, such renewal of ties may help to counter misinformation, offer closure, and possibly lead to new ideas and opportunities to “upcycle” former interventions.
- 22 In our workshop, only a handful of Kenyan participants had personal connection to Kefinco and none (...)
46Above all, our workshop underlined an imperative for responsible development action in a manner that considers the long-term, multi-dimensional impacts of projects. A wide range of actors stand to benefit from revisiting the dried-up waterbeds of bygone bilateral development projects. Above all, such workshops have the potential to foster more equitable, locally informed strategies for assessing the wider legacies of international interventions. They can help to counter the more or less abrupt ways in which projects dissipate, assisting the local population to better understand and process interventions in their midst and even to develop tools for effective claim-making. In this regard, working with local community members as participants in such workshops would be highly desirable, with careful attention paid to mitigating unrealistic hopes—or concerns—for “return.”22
