Telling ruins: the afterlives of an early post-independence development intervention in Lake Victoria, Tanzania

Abstract

In the early 1960s, three pilot agricultural and settlement schemes were set up along the shores of Lake Victoria in the north-western region of Tanzania with the involvement of Israeli development agency Agridev. One of these sites was Mbarika, where the experimental project ran for three years and had mixed results before being discontinued by the young Tanzanian government. This article explores the story of that scheme and its long-term legacies some 50 years on. Unpacking the representational and material ruinations that outlived the project's official timeline, we examine the memories and rumours that continue to haunt the site to this day and their entanglement with successive development experiences and shifting political ideologies. Through interviews, ethnographic observations and archival research, we shed light on the complex, deeply ambiguous legacies and ‘afterlives’ of a development intervention set between expectations of modernity and a sense of exclusion.

‘There is no yard stick to measure the [project's] accumulated self confidence of the farmers and its influence on the new generation, but there is no doubt of its existence and its progressive influence.’ (Proposal Report by Agridev to the Tanzanian Ministry of Land, Settlement, and Water Development, October 1966)¹

INTRODUCTION

The Tanzanian village of Mbarika is located in Mwanza region, about 100 kilometres south of Mwanza city.² Fishing is a major activity as the village is located on the shores of Lake Victoria. Local livelihood also relies on animal husbandry and agriculture, as well as food crops (maize, cassava, sweet potatoes, paddy, groundnuts, millet and sorghum) and cash crops (notably cotton). Since the early 2000s, small-scale gold mining activities have brought newcomers from as far as Shinyanga, Geita and Arusha. Alongside these recent activities, however, Mbarika is still known along the southern shores of Lake Victoria for a four-year pilot cooperative scheme implemented from 1962–1966 in collaboration with the Israeli Ministry of Agriculture. This agricultural project, which is referred to locally as ‘the Israelis’ project’ (mradi wa Waisraeli in Swahili) – henceforth the Mbarika pilot – introduced new crops such as onions and new varieties of cotton, and employed advanced machinery for irrigation, pesticide and fertilisation. At the decision of the Tanzanian government, the Mbarika pilot came to an end in late 1966, just a few months before the Arusha Declaration announced that Tanzania would opt for African socialism (Ujamaa) and self-reliance (Kujitegemea) – a turning point in the history of the country.

Fifty years after the end of the pilot, the traces of the short-lived scheme are still visible, both as physical ruins around the village landscape and as memory narratives that continue to inform local debates about the past and expectations of development to come. The old lakeside field is still referred to as ‘the Israeli farm’ and the Lake Victoria waterfront is known as ‘the Israeli port’, though neither site is under cultivation today. The ruined structures that once stored fertilisers, cotton and two water pumps, along with the massive rusty irrigation infrastructure with its metal pipes and water taps, remind residents of an attempted transformation that never took off. In the decades that followed the Mbarika pilot, a long list of initiatives came and went – some sponsored by foreign countries or by the Tanzanian government, others by NGOs, and yet others by for-profit actors.

Drawing on interviews with villagers and ethnographic observations,³ as well as archival material,⁴ this article reveals the complicated, deeply ambiguous legacies of development in Mbarika. Its focus – a post-independence, pre-Ujamaa development initiative – has largely been overlooked by the recent literature on legacies and memories in Tanzania, as opposed, for instance, to the late colonial Groundnut Scheme (Rizzo 2006; Ahearne 2016) and the Ujamaa villages (Askew 2008). Central to our analysis is the notion of ‘afterlives of development’, an ambiguous concept that has been used both to transcend dramatic proclamations about ‘the end of development’ in the 1990s (Sachs 1992; Rahnema & Bawtree 1997; Rudnyckyj 2010) and to analyse the unfulfilled promises of development – whether a project was formally evaluated as ‘successful’ or as ‘failed’ (Ferguson 1990, 1999; Li 2005; Gez 2021). Drawing on the critical view of a development project ‘as an ongoing socially-constructed and negotiated process that goes beyond the time/space frames of intervention programmes’ (Long 2001: 4), such afterlives are often replete with unintended consequences (Bierschenk et al. 1993; Olivier de Sardan 1995; Koch & Schulpen 2018) and ironies (Leve 2009). Implied in this perspective is the idea that, rather than consisting of bygone ruins – that is, reified and ineffective things – the past's continuous relevance should be deciphered using the prism of processual ‘ruination’ (Stoler 2013). We argue that, while on the formal side, the Mbarika pilot holds lessons about the limits of technology-intensive rural development, its ongoing tangible and intangible ruinations, as lived and interpreted by the local population, are a telling expression of ideological and practical tensions in post-independence Tanzania. In particular, by dwelling on the circulation of rumours, we explore the dynamics of perceived exclusion as a central legacy dominating the project's afterlives.

The article consists of two parts. The first part, which includes the first two sections, draws on archival material to present the rise and fall of the Mbarika pilot – first by identifying it within the context of the three Israel-led schemes around Mwanza, and second through focus on implementation in Mbarika proper. The second part consists of three sections in which we draw our ethnographic data to explore the ambivalences and ironies of conflicting narratives regarding the project's long-standing legacies. Focusing especially on local rumours and oral histories, we set these legacies within the context of top-down project design and, more tentatively, ex post facto reconstruction of narratives. In the conclusion, we return to the concept of ‘afterlives’ by synthesising a long-term view on the Mbarika pilot and its multifaceted local consequences.

THE ISRAEL-LED PILOT SCHEMES IN LAKE VICTORIA

At independence in 1961, Tanganyika's peasant economy was dominated by hand-hoe technology. The persistence of these practices flew in the face of years of centrally directed development schemes aimed to transform farmers into modernising agents and to impose villagisation, terracing, agricultural extension work, veterinary services and marketing operations (Iliffe 1971; Kjekshus 1976, 1977). The president of Tanganyika (Tanzania from 1964), Julius K. Nyerere, inherited, as he said in 1962, a ‘country of peasant farmers’ (Nyerere 1966: 183). The agricultural policy set in the country's First Development Plan of 1964–1969 was a continuation of the colonial development policy supported by the World Bank with a view to encouraging cash crop production and to fashion a ‘progressive farmer capable of setting in motion the progress of the whole agricultural community’ (Dumont 1969: 11). Tanzania's Village Settlement Scheme was launched in December 1962 to boost agricultural productivity through ploughs, tractors and fertilisers, and by creating villages where, up to that point, people had lived in scattered homesteads (Moore 1979). According to Nyerere, ‘the first and absolutely essential thing to do … if we want to be able to start using tractors for cultivation, is to begin living in proper villages’ (Nyerere 1968: 183). This initial phase of villagisation was seen as ‘a precondition for the introduction of novel and costly technology financed … by foreign aid’ (Kjekshus 1977: 275).

One of the countries that provided significant foreign aid in the spirit of Nyerere's vision was Israel. That country's historic Labour Party, with its long history of experimentation with rural collectivist communities (notably the kibbutz movement), found natural allies in the leadership of socialist-leaning leaders such as Kaunda in Zambia and Nyerere in Tanganyika/Tanzania (for example, see Schler & Gez 2018). Between the late 1950s and the early 1970s, Israel was strongly invested in Africa, notably through construction projects and the introduction of agricultural technologies, but also through social organisations such as cooperatives and youth movements (Bar-Yosef 2013; Yacobi 2016). These relationships deteriorated following the 1967 Israeli–Arab war and then largely collapsed following the 1973 Israeli–Arab war and the Organisation of African Unity's decision, following pressure from Arab member states, to sever all ties with Israel.⁵ Prior to that, during the so-called ‘honeymoon phase’ (Oded 2018), Israel – mostly under the Center for International Cooperation within the country's Ministry of Foreign Affairs (MASHAV, Hebrew acronym) – sent over 2700 experts to Africa for development initiatives, and tens of thousands of Africans were trained in Israel. While this Israeli investment had an element of ideological sympathies with the political awakening and long suffering of colonised people, it was first and foremost an attempt to promote political and economic peripheral alliances in the face of the Israeli–Arab conflict (Oded 2018). East African countries were of particularly high geo-strategic priority because of their proximity and their access to two key water sources: the Nile and the Red Sea (Carol 2012 [1977]).

In the first years after Tanzania's independence, its relations with Israel were close. Despite their limited familiarity with the local context, Israeli experts were commissioned to write two reports on the country's economic development – reports that inspired, in part, the Tanzanian government's approach towards the technological and attitudinal transformation of its rural population (Kaplan 1961; Schneider 2014: 23). Other collaborations related to areas such as security and construction.⁶ In 1962, Tanganyika's government hired the services of Israeli experts to develop three pilot rural agricultural schemes along the shores of Lake Victoria, in the vicinity of the town of Mwanza. The government's choice of location was not coincidental. Mwanza lies in the heart of Sukumaland, a largely flat savannah associated with the Sukuma people, Tanzania's largest ethnic group. After the Second World War, the British colonial authorities, having recognised the area's strategic importance and the supposed ‘homogeneity and malleability of the Sukuma’ (Maguire 1969: xxiv), used the region as an ‘“experimental laboratory” for late colonial policies of economic, social and political reform in Africa’ (Schuknecht 2010: 9). In particular, cotton, which was first introduced to the region by the German administration, became Sukumaland's principal cash crop in the years after the Second World War. It was partially in defence of the interests of local cotton growers, and under the leadership of figures such as Paul Bomani – later to become Tanzania's first Minister of Finance and a prominent politician – that the region thus became ‘the largest producers’ cooperative movement in Africa’ (Maguire 1969: xxiv). Considering the feeling of top-down directives and absence of community participation that, as we show below, dominate the memories of the Mbarika pilot scheme to this day, it is especially worth highlighting the contrast with Sukumaland's legacy of late colonial cotton-growing cooperatives: as scholars note, in Sukumaland between the Second World War and independence, the surge of cotton cultivation and the organisation into cooperatives were both driven from below – that is, not so much by the colonial administration as by peasants’ entrepreneurial drive and their response to market logics (Maguire 1969; Schuknecht 2010).

With this regional reputation for cotton cultivation and for cooperative organisations, the three pilot schemes – to be implemented by Israeli agency Agridev⁷ – were to build on the region's existing strengths and reputation. These would be complemented and improved upon through the introduction of new seed varieties and a diversification of crops, through the advancement of modern machinery for cultivation, irrigation and fertilisation, and through the incorporation of cooperative knowledge drawn from Israel's extensive experience. The schemes involved no forced resettlement of the local populations, but rather relied on the selection of volunteers from within the region. Participants received an allocation of plots, and were given loans in the form of subsidies for the collective purchase of seeds and fertilisers, which they eventually had to repay individually, under favourable conditions.⁸ They were to enjoy collective access to advanced agricultural technology including irrigation pumps, sprinklers and tractors – operated by Agridev – as well as agricultural training and guidance on cooperative management. Upon harvest, the villagers were to sell their cash crops in bulk to a central shop operated by Agridev, who would then sell the produce in Mwanza. Apart from several varieties of cotton, farmers were encouraged to grow onions and to experiment with groundnuts, maize and rice.

Overall, three schemes were developed along the lake shores: Mbarika, established in 1962, followed by the lakeside schemes of Kalamera and Nyatwali – both east of Mwanza and therefore far from Mbarika – the following year. As pilot schemes, the settlements were each slightly different, and each confronted specific environmental and social challenges. On the whole, from a formal implementation perspective, the schemes proved moderately successful. They introduced new technical know-how such as double cropping and advanced tools that improved yield substantially. Yet their reliance on advanced technologies in a low-infrastructure setting implied tremendous running costs while their experimental orientation did not satisfy the government's expectation for quick, robust and replicable results. As political scientist Abel Jacob, in his evaluation of the project shortly after its termination, sums up, ‘To begin with, there was simply not enough time given for the project to succeed. Three years is a very short time … The fact that the Tanzanians began to reappraise these schemes after such a short time indicates impatience – and a lack of the funds necessary for a long waiting period’ (Jacob 1972: 193).

In addition, the early stages of the project revealed several design flaws, such as suboptimal distribution of water pumps and poor choice of farmland. Though unfamiliar with the area, it was not until 1965 that Agridev invested in comprehensive soil surveys and generated maps of the sites using aerial photography. By then, however, it was already too late. That same year, the Tanzanian government ran its own evaluation and concluded that the schemes were too expensive to keep. They decided to let the schemes’ contract expire, which it did in 1966. In addition to consideration of costs, the government's decision seems to have been influenced by an unflattering independent report focused on the project's early phase, as well as by the lack of enthusiasm with the project among many of the ‘settlers’ as they were called.⁹ In their internal correspondence about the project's disbandment, Agridev's experts admitted that it took them too long to find paths to viability and lamented that, by the time they managed to solve the fundamental flaws, they were already shown the door. As one of them candidly wrote: ‘We have come here – Agridev and the [Israeli] Ministry of Agriculture – at the request of the Tanzanian government, as experts and not as managers of an experimental farm. And yet, we have erred as if we were novices … For every concern that was raised we had but one answer: This is but an experimental project, and we must have more trials’ (Israel National Archives, our translation).

THE MBARIKA SCHEME

Of the three Lake Victoria schemes, Mbarika was regarded as the most complete exemplar of the designers’ vision being implemented. Starting in the 1962/63 cultivation season, Mbarika first involved 15 families who operated on a 40-acre farm. By 1964, it had grown to 50 families and some 225 acres, in addition to 25 TANU Youth¹⁰ who cultivated another 10 acres of onions. While the final number of families stayed around 50, there was a turnover of members, and in early 1966 Agridev was writing with relief about the departure of the ‘worst’ 20 families, a move that they claimed improved social dynamics: ‘Slowly but surely, we see the arrival of good members instead of the bad ones who have left’ (Israel National Archives, our translation). Mbarika's 1962/63 and 1963/64 yield was fairly modest, but by 1964/65 the scheme began to show real promise. During that year, project participants were able to return all their costs, including depreciation charges for farm machinery – thereby reassuring Agridev about handing the scheme over to Mbarika's local cooperative. At the same time, Agridev acknowledged that, being geared towards experimentation rather than mass production, the project's success was tenuous. According to Agridev's director, Amiram Sprintzak, ‘Even Mbarika, which is considered successful, is not viable according to the Tanzanian government's expectations … one cannot expect both learning experience and economic viability; there is no such thing’ (Israel National Archives, our translation). This experimental thrust manifested, for example, in a 1965 collaboration between Agridev and the nearby Ukuriguru Agricultural Research Station – one of the oldest agricultural research institutes set up in the 1930s primarily for cotton production – with the purpose of identifying additional suitable crops to complement cotton and onions.¹¹

In late 1965, Agridev began coming to terms with the handing over of Mbarika to the government. From the Israeli side, the belief was that withdrawing from Mbarika would allow Agridev to concentrate its limited resources in Nyatwali and Kalamera, where challenges were greater and viability was still a long way away. Agridev also knew that time was running short and preferred to take the initiative, noting in an internal correspondence that, ‘We wish to prove the Africanisation policy of our project in practice, and before we will be reminded of that [by the government]’ (Israel National Archives, our translation). Thus, despite Agridev's concern with the reversibility of the scheme's fragile and modest achievements, in June 1966 it reached an agreement with the Tanzanian Ministry of Land, Settlement and Water Development that Mbarika ‘is ripe to be transferred’. Acknowledging the quality of the village's cooperative, minutes from the meeting explain the decision using the following justifications: ‘(i) the area is limited and precludes much expansion (ii) it is distant from the other Agridev schemes, and therefore expensive to administer (iii) the farmers are of a high quality’ (Israel National Archives).

Debates ensued about what the project's termination would mean for Mbarika and how this transition should be carried out in practice. In particular, the project's cost-intensive agriculture implied a complicated transition, one that not only required the integration of well-trained locals into management positions, but also necessitated abundant yield returns to cover high running and maintenance costs. These high costs were acknowledged by one Agridev expert, who – writing about the Mwanza schemes in general – warned that, ‘if the farmers of our schemes will suddenly become fully independent, their income will be lower than that of farmers outside the schemes, who would have no machinery and irrigation system’.¹² In March 1966, Israeli Ambassador to Tanzania, Yitzhak Pundak, voiced his concern about the future of the Mbarika scheme, arguing that,

they [the Tanzanian government] interpreted [our] suggestion that they take over the scheme as the dismantlement of the whole business there: taking out the watering system and the agricultural machinery, repositioning the local management staff – in other words, returning the territory of Mbarika to its state prior to our efforts. We, of course, oppose this approach, as the impact of such a step on the region and on the Tanzanian government requires no explanation. (Israel National Archives, our translation)

Fearing that the handover would mark the end of the scheme, Agridev pushed for a gradual transition, and proposed reducing the experts’ presence to a single weekly visit over an interim period. Other solutions were floated as well, such as integrating the scheme into the Tanzanian National Service, whose establishment followed the Israeli Nahal model of combining agricultural settlement and military training, and where Israeli experts had already been involved in the management of multiple experimental farms (Jacob 1971). Agridev clearly had an interest leaving things in good shape: one expert wrote to his peers that ‘it would be unseemly if we leave Mbarika with an unusable irrigation network’ (Israel National Archives, our translation). At the same time, some within Agridev suspected that the scheme would be doomed once they had left, and discussed the possibility of saving the machinery by transferring it to the two remaining schemes.

Throughout these negotiations regarding the future of the project, one voice that seemed suspiciously absent was that of the villagers themselves. While tensions between locals and Agridev across the Mwanza projects were at times pointed to as a weakness and as an argument against continuous sponsorship, there seems to have been little attempt to increase participants’ involvement in charting the future of the scheme. This disconnect seems to have been mutual, and according to the villagers we spoke to in our fieldwork, their trust in President Nyerere and their own limited exposure led them to accept the decision to terminate the scheme without dispute. Benjamin, a Mbarika elder who used to work for the local court at the time of the project, recalls in 2015 that, ‘you know, community members had no voice. Most of the citizens are just normal people, if they are told to do this, they simply do. If they are told stop, they stop what they have been doing’ (Dotto family 2015 Int.). Mbarika's workers were apparently given only minimal and unsatisfying explanations as to why a high-investment project that appeared to be on track for success was suddenly discontinued, an opacity that – as we discuss below – continues to nourish speculation to this day.¹³

The centre of Mbarika, which gathers both mud houses with a thatched roof and brick houses roofed with corrugated iron sheets, is arranged around a large tamarind tree. It is around this tree, surrounded by a few shops, that community meetings have been conducted for decades. It is there, too, that Mwalimu Nyerere, the architect of post-independent Tanzania (Fouéré 2015), is remembered to have addressed the villagers about Ujamaa and collective work during a visit he made in the late 1960s, not long after the Israel-led pilot was abandoned and Ujamaa started on a national level. At the time, Nyerere was indeed touring the country to introduce Tanzania's new development path based on villagisation and collective work (Nyerere 1966, 1968), which was aimed to provide ‘services, such as schools, safe water, and health facilities’ (Schneider 2007: 12). The shores of the lake that were chosen as a site for cultivation for the Israeli scheme are situated further down from the village centre. Michael, who was a school pupil in the 1960s, remembers Nyerere's visit to Mbarika, and in a 2020 interview recalled that Mwalimu did not even visit the lakeside site: ‘It had already collapsed’, he shrugged, implying that, by that point, nothing of interest was left there for Nyerere to see (Michael 2020 Int.).

It is true that the end of the Mbarika pilot marked an end of a mass investment, but Michael's shrug is likely to have been overstated. After the project ended, the state seized most of its moveable material, including machinery for grinding groundnuts and crushing maize. What could not be taken away – the underground pipes, the water tank, the sparse structures – was left on site. This was the time of Ujamaa Vijijini, when new villages were created and their inhabitants summoned to cultivate the fields collectively and to set up cooperatives (Von Freyhold 1979; Hyden 1980; Schneider 2004). In Mbarika, the Israeli scheme already brought people together, but after its termination and the arrival of Ujamaa Vijijini to the Mwanza region in the early 1970s, more farmers gradually joined, in line with a national policy requiring Tanzanians to live as wajamaa in planned villages, at times even through forceful relocation (Scott 1998). Mabula, a Mbarika elder who participated in the pilot, emphasised in a 2015 interview that the state was initially able to provide ‘fuel for the water pumping machines to continue with the farming’, but when the fuel ran out, the initiative fell through ‘because most of the people couldn't afford to contribute, and they didn't even harvest’ (Mabula and Paul 2015 Int.). Although scholarly literature shows that the state made massive investments in the Ujamaa villages, notably making ‘important expenses for material incentives’ in order to cushion ‘the peasants’ move into the village’ (Kjekshus 1977: 279), the sheer size of the programme made it impossible for the government to sustain it adequately. State bureaucracy was too weak, and in some instances resources were misappropriated (Hyden 1975, 1980).

While having the advantage of proximity to the lake, no intensive agriculture could be maintained once the Mbarika pilot's irrigation system gave way, and the farm area was converted back to traditional hoe farming for food crops. Today, the cultivation site is barely in use. For the visitor, it is a quiet, scenic and widely barren shrub land, save a number of small fields of maize and a handful of ruins. At the edge of what used to be the collective field, on top of a hill overlooking the lake, one can see a high-roofed, long-collapsed brick structure, which was once used as the village's temporary buying post, storing the picked cotton until it was transported to Mwanza by Agridev's own experts. Several other collapsed buildings are also recalled as the residence of the project facilitators, the main office and the fertiliser storage areas (Figure 1). All along the sloping field is the remaining irrigation infrastructure: a rusty, incomplete system of both over-ground and underground rewelded pipes with their rotating taps for irrigation emerging from the surface (Figure 2), originating from two lakeside pump stations kept in concrete barracks (Figure 3) and once used to conduct the lake water uphill to the field. The two pump stations still bear a plaque with the insignia ‘W.D. & I.D. 1964’, an acronym that stands for the Tanzanian Water Development and Irrigation Department (Figure 4).

Figure 1. Remains of the project's main storage facility.

Figure 2. Rotating taps for sprinkler irrigation.

Figure 3. Rewelded pipeline between the pump station and the irrigation site.

Figure 4. One of two pump stations, carrying the insignia ‘W.D. & I.D. 1964’.

AMBIVALENT AFTERLIVES

Today, the Mbarika pilot continues to serve as a reference point for the elders who participated or saw it from up close as well as for young people and newcomers. Jane, a 28-year-old mother of five who moved to Mbarika in 2009, recounted in 2020 that, ‘what I heard about Mbarika is that it is a historical village, it is the village known to be the village where irrigation farming was firstly undertaken by Israelis. The Israelis came and established the irrigation project with the community members, they were cultivating vegetables’ (Jane 2020 Int.). Such knowledge is not limited to Mbarika's residents: according to Masanja, a local farmer, ‘everyone who passes by the area says, this was the Israeli project’ (Masanja 2020 Int.). Yet what exactly is associated with this heritage is ambiguous – a reminder of the gross simplification of dividing a project along a simple matrix of success versus failure. Scholars highlight that a development project may be regarded as (un)successful in its formal mode of execution or in meeting its set indicators, while having reverse long-term effects on the underlying problems that necessitated the intervention to begin with (Ika et al. 2012; Ika 2015; Andrews 2018).

Perhaps the clearest manifestation of the Mbarika pilot's ambivalent legacy is found in the fundamental tension between narratives of productivity and narratives of exploitation. There is no question that the project introduced new technological tools and, for a brief instance, increased yield like never before. One elderly interviewee recalled that cotton grew as high as the stature of a person and produced four times as much as what people were hitherto familiar with. Onions, too, were a success, and Mbarika's onions briefly became a familiar brand name across the area – and even across East Africa, according to one interviewee. Recalling such success, several interlocutors spoke of the project with great pride and in almost idyllic terms. Benjamin, the abovementioned former court clerk, recalls that, at the time, ‘people could build good houses, there was food security, and sometimes you found there was no place to store the food, because of high yield. For three to four years, there was no more hunger’ (Dotto family 2015 Int.). Paul, an elderly man who participated in the scheme as a young farmer, similarly said that ‘the yield was so high it took us a long time to complete harvesting … and when they cultivated groundnuts it was also very productive. We didn't even finish harvesting’ (Mabula and Paul 2015 Int.). It was also mentioned that, throughout the project, the local school was provided with fresh food such as wheat flour and maize, allowing pupils to take their lunch on site. With the money earned, farmers constructed new houses – at the time, ‘a clear sign that one has advanced in life’ (Kamat 2008: 372) – that stand to this day.

But these achievements did not come cheap, and interviewees emphasised the intense labour that was demanded from workers. Respectfully referred to by some interviewees as ‘teachers’, the Israelis themselves are remembered as hard-working and diligent, and they would start working before dawn and only finish well into the afternoon. Noah, a life-long resident of Mbarika, who was in his twenties during the time of the scheme, summed up his impressions from the Israelis thus: ‘They used to wake up early in the morning, they were only wearing shorts, and they did not worry about the mosquitos in the area’ (Noah 2020 Int.). Interviewees told us that the development workers regarded work time as so precious that, instead of letting farmers walk home for lunch, they would drive them the short distance home and then back to the fields. This hard work ethic won them the locals’ admiration, but the expectation for total commitment – accompanied by sanctions against those who did not meet the high bar – made some participants feel exploited. We were told that Agridev's experts were keeping a strict registrar of attendance, keeping away community members who were not formally part of the scheme and deducting monthly earnings from those who failed to tend to their fields. This rigid code stirred discontent, allegedly even bringing some participants to complain to the government representative.

Perhaps the harshest critic that we encountered in Mbarika was 68-year-old George, who was born in the area and witnessed the scheme as a young teenager. In a 2020 interview, he told us how the Israelis came as neo-colonials and grabbed the land from the people by force, reducing their status so much that ‘the members of the community were like slaves’ (George 2020 Int.). He explained that, ‘they came and started enslaving the people to work in their farm as labourers. And at that time money was very difficult to get, and they [the foreigners] had money so people had to work hard on their farm to get money’ (George 2020 Int.). The fact that only little that is tangible and usable survived to this day, was for George an indication that the foreigners did not have the population's best interests at heart: ‘It was capitalism politics, they didn't want to build good infrastructure, there is nothing perhaps we could have inherited, maybe their palaces, but they didn't build such houses, only stores and small houses which they left and were very small. Even the machines that they left were destroyed immediately after they have left’ (George 2020 Int.).

In the next sections, we will try to locate George's harsh words and resentment within a historical context, notably the rise of Ujamaa socialism soon after the pilot's termination and its ideological emphasis on self-reliance. For now, it suffices to note how such a grim vision, which digs into Tanganyika's slavery and colonial past, was probably exacerbated by the emotional and physical distance between the development workers and the farmers. Development scholars note the centrality of cross-cultural ‘contact zones’ (Pratt 1991) of asymmetrical power relations and (mis)understandings that guide ‘interfaces’ (Long 2001) between local communities and foreign development workers and their visions of progress both abstract and concrete (Delgado Luchner 2018; Footitt 2019). While Israeli development initiatives in general had a reputation for informality, this seems to have not been the case in the three Mwanza schemes. Not speaking a common language, Agridev's small team ‘came with their own interpreters’ and largely kept to themselves (Noah 2020 Int.).¹⁴ They lived with their families in Mwanza – a fact that, as one Israeli expert regretfully observed towards the end of the project, ‘rendered organic contact with the [local Tanzanian] settlers non-existent’ (Israel National Archives, our translation). The older villagers to whom we spoke were divided as to whether the Israeli experts ever joined the community for a game of football or otherwise bonded with them outside working hours. While partaking in such leisure activities may appear trivial, the establishment of such bonds can prove highly consequential: as Monson (2009) shows in her study of the Chinese-led construction of the ‘Freedom Railway’ in Tanzania in the 1960s and 1970s, informal interactions between foreign and local project workers are responsible for some of the project's most lasting legacies. As an illustration, we can contrast the eagerness with which, according to Monson (2009: 58), the Tanzanian railway workers attended the screening of Chinese films during their off-duty hours, with how the Israelis supposedly only used their projector to show an educational film about agriculture, ‘on how crops like cotton are cultivated, the whole process of cultivation, from planting to harvesting. They showed different countries where they happened to do such kinds of project, including in their own country’ (Mabula and Paul 2015 Int.). The Tanzanian participants, on their side, also kept their distance. Mary, the wife of the abovementioned Benjamin, who was herself born in Mbarika, explained that, ‘in those days people were so fearful. After all, they [the Israelis] lived away from the community. They were living in their camp. So, the community was living at the centre. So, they didn't become familiar with the entire community. They were only familiar with the small group of people whom they were working with’ (Dotto family 2015 Int.).

One scene described in Agridev's own internal reports demonstrates this gap. In the scene, dated to mid-1966, the Tanzanian Minister of Lands, Settlement, and Water Development, Said Maswanya, is described visiting the site with his entourage and taking questions from the scheme's participants. The exchange, as translated to the Israelis, included a request by some farmers to ‘drive the tractors, operate all the machines, and control for pests ourselves’. The ministerial convoy rejected these requests outright, allegedly with a blunt retort, ‘you can't do those things because you will cause damage’ (Israel National Archives, our translation). While the foreign experts did train some local ‘understudies’ as they were called, the exchange reflects a hierarchical tone that is in keeping with the top-down development approaches of the 1960s in general and the attitude emerging from Agridev's internal correspondences in particular. Operating on a tight schedule, committed to showcasing the productive potential of state-of-the-art modern technology, and seeing themselves as ambassadors of their country's diplomacy in the East African region, the Israeli experts were extremely preoccupied with productivity and progress, possibly at the cost of overlooking actual experiences and interpretations of the project among the Tanzanian villagers. In their internal letters, references to the human dimension were highly technical, and local participants were largely reduced to classification as ‘good’ versus ‘troublemakers’. As scholars note, such neglect of the human component can compromise a project's desired outcomes in both the short and long term (Chambers 1997; Olivier de Sardan 2021). Indeed, when Agridev was wrapping up its work in Mbarika only a few months after the minister's visit, it became particularly concerned about insufficient training that would allow project participants to successfully take over the mechanical equipment. As we saw, Agridev's concerns were well founded, and the project proved quick to collapse.

In her study of Dar es Salaam through the turbulent 1970s and 1980s, Brownell (2020) dwells on the centrality of local resources and bottom-up improvisation as essential for complementing the shortcomings of the formal economy at a time of scarcity and as an assertion of residents’ ties to their city. Arguably, such bottom-up resourcefulness is especially important in lacustrine zones, where fast-changing ecological and social systems require extreme agility (Derbyshire 2019). That the Mbarika pilot was quickly discontinued tells us not only about daunting technical challenges – insufficient knowledge transfer and faulty handover, seizure of moveable equipment, challenges of maintenance, etc. – but also, we argue, is an expression of subtler feelings of exclusion, alienation and suspicion. As the aforementioned Masanja explained, ‘if people don't participate fully, they won't care about the project or even the machine or the equipment. Other equipment can also be stolen, such as fuel from the water pump; they [participants] should feel as part of the project’ (Masanja 2020 Int.). In the next section, we dig deeper into local residents’ sense of alienation from the project by reviewing the circulation of rumours and their relation to the project's vagaries, fluctuations and abrupt termination. In so doing, we draw an implicit link between feelings of inclusion and ownership on the one hand and the project's long-term viability on the other.

THE LUST FOR GOLD

Rumours in Mbarika, many of which persist to this day, are a powerful indication of the ambivalent legacies of the Mbarika pilot. As scholars note, rumours are a fluid narrative form best suited for capturing uncertainty and open-endedness (Ellis 1993; Osborn 2008; Sunstein 2009). Grassroot rumours thrive in environments of ‘information vacuums’ (Carlson et al. 2018), where they introduce their own explicative frameworks to compete with official (lack of) explanations. Seen as a response to knowledge gaps and anxiety – such as when a project makes unprecedented investment one day and disappears the next day, driven by occult logics that estrange people from their source of livelihood – they can be interpreted as a product of structural exclusion from the decision-making process. At the same time, rumours’ inherent dynamism and heterogeneity means that they are ever-subject to narrational re-emphasis and reinterpretation by individual storytellers. This diversity means that, even within a community, rumours may be controversial and subject to ongoing negotiation. In Mbarika, some of our interviewees venture deeply into the realm of rumours while others rejected them categorically. Our interviewee Michael, whose family has been established in the area for generations, asserted that, ‘truly I haven't heard any rumours different from what they [the Israelis] intended to do … they were only focusing on agriculture’ (Michael 2020 Int.).

One dominant story that circulates in Mbarika revolves around gold excavation. During our first visit to the village in 2015, we got to know Juma, a local teacher in his thirties. Well-educated and speaking excellent English, Juma had been swept in the gold rush of the last two decades. He was further engrossed in stories about the Israeli scheme, and was clearly suspicious that the enterprise had ulterior motives. Together with two Mbarika elders called Majaliwa and Kulwa, Juma took us to a small man-made hollow in the vicinity of the old farm, which they themselves dug some five years earlier. There, the group crushed some rocks and gave them to us to sniff: ‘it is like poison’, they explained the slight sulphuric-like smell.¹⁵ We were then told that the Israelis must have had knowledge that the Tanzanian government didn't regarding the rich availability of minerals in the area, and that that is why they had chosen to come to Mbarika of all places. As Kulwa explained, ‘what I know is that these people came under the shadow of agriculture, but they had two things in their mind: agriculture and at the same time looking for gold’ (Juma, Majaliwa and Kulwa 2015 Int.).

According to some, this core story takes a more sinister turn. Adding to the words of his colleagues, Majaliwa told us that, ‘when the Israelis entered into misunderstanding with the Tanzanian government, they decided to poison [the land], so that all of us could not benefit from it’ (Juma, Majaliwa and Kulwa 2015 Int.). In other words, greediness brought the foreigners to Mbarika, their presence was maintained through scheming, and vindictive retribution eventually led them to retaliate. The poison that the Israelis supposedly spread destroyed minerals across the village, a teleological narrative explaining why people have been able to find gold in the localities around Mbarika, but never in the village proper. Recognising the controversiality of the claim, the proponents of the poisoning narrative alluded to the supposed work of German scientists who, working with the Tanzania Minerals Audit Agency (TMAA), were said to have visited the site a few years earlier. Further corroboration allegedly came from a visiting South African geologist. Somewhat ironically considering the Mbarika pilot's proud harnessing of state-of-the-art agricultural science, Majaliwa explained that, ‘for us, we were not aware [of the poisoning]. We came to be aware when geologists started coming to explore the area, and discovered that the minerals are there but they have been poisoned. Can't you trust an expert? You must believe that’ (Juma, Majaliwa and Kulwa 2015 Int.). The younger Juma, on his part, refused to openly endorse the poisoning allegation, but explained that, ‘in 2010, people started looking for gold and they busted [rocks] here, and after busting they didn't find the gold – they found a kind of copper. Therefore, from their perception, maybe the Israelis did poison the area’ (Juma, Majaliwa and Kulwa 2015 Int.). For those like Juma, who hold a less conspiratorial stance, the Israelis may have involved themselves in mining only as a side activity alongside the agricultural project. Such ideas resonate with contemporary stories about other projects – from road construction to installation of latrines – that are quickly remodelled as artisanal mining initiatives once the workers stumble upon precious minerals. Reflecting on the afterlives of development, it is hard not to dwell on the symbolic significance of the debate between those who argue that, over the years, the effect of the alleged poisoning has diminished and those who suggest that, ‘as days go by, the situation becomes worse’ (Juma, Majaliwa and Kulwa 2015 Int.).

Accompanying this narrative, some villagers believe that the Israelis hid their extractive goods underground until they could come back and reclaim them. This narrative seems to imply that the experts’ exit was experienced as rushed – an experience that is in tension with the negotiation of departure over several months as emerging from the archival material. Among the villagers we spoke to, there is no consensus regarding the location of the treasure site, which in any event is said to be booby-trapped or guarded by snakes. One candidate site is a spot at the edge of the collective field, where the Israelis are said to have installed a tall antenna and other communication devices. Today, all man-made devices seem to have been cleared from the area – the remains gradually disappeared, Majaliwa explained, due to ‘people's carelessness and theft; they were stolen and used in people's house as poles; others [were sold for], like, scraps’ (Juma, Majaliwa and Kulwa 2015 Int.) – but rumours regarding the site still persist. In our visit to the spot, our guides pointed to a handful of concrete foundations that are still embedded deep within the soil. These rectangular blocks each contain a small metal hook – used for lifting and setting the blocks, but whose resemblance of a handle evokes the promise of unlocking. Pointing to one such rectangle, a local elder told us that the blocks can be lifted ‘like a suitcase’, implying that below them may lie the underground treasure trove.

To be clear, such line of rumours is not unique to the case of the Israel-led schemes. Across Africa, mistrust of foreigners and concern with exploitation has long manifested in circulation of rumours (White 2000). Many rumours tie outsiders, including development workers – or, for that matter, academic researchers – to the extraction of precious minerals as the true motivation (Onneweer 2014). In our work in Mwanza, rumours about the Israelis’ hunger for gold bore a remarkable resemblance to stories in the region about colonial exploitation of resources and the underground treasures that they left behind.¹⁶ Still, the portrait of mistrust that these stories paint needs to be put in its specific context. It appears that, in the minds of some villagers, it was not coincidental that the end of the project concurred with the promulgation of Tanzania's self-reliance policy. Even though the Arusha Declaration was only made a few months after the project ended, these interviewees pointed to it as the turning point for denouncing foreign interventions in general and Israeli involvement in particular (Bjerk 2011). Indeed, the strong anti-colonial sentiments associated with Ujamaa might help to explain why the Mbarika pilot came to be cast by some villagers as a colonial endeavour.

Furthermore, it did not help that the project ended during a time of heightened political tensions between Israel and its neighbours that resulted in the 1967 Arab–Israeli War, after which anti-Israeli rhetoric was toned up in Tanzania as in other African countries, with Israel being increasingly cast as a colonial aggressor (Oded 2018: 226). In this respect, Tanzania's termination of relations with Israel in 1973 following the Organisation of African Unity's binding decision was likely seen as an affirmation of such foul play. Today, many in Mbarika continue to believe that the answer to the Israelis’ sudden departure is found in international politics, for example by suggesting that, during a visit to the region, the President of Egypt Gamal Abdel Nasser convinced Nyerere to that effect.¹⁷ As Lilian, the elderly sister of local court clerk Benjamin, told us, ‘to hear from that time when we were young girls, it seems there was a misunderstanding between Israel and our nation, it was like a conflict between nation and nation. Because they left suddenly. I don't know if the contract ended, but it also reached at a time when our relationship with Israel was not good’ (Dotto family 2015 Int.) It is in this context that we can also understand a strand of rumours depicting Agridev's experts as spies, who may have been sent away after their ploy had been uncovered – a narrative built around the above-mentioned ‘communication centre’ at the edge of the farm. But to the extent that attitudes towards the Mbarika pilot have been shaped by the vagaries of international politics, one must also keep in mind the two countries’ gradual rapprochement since the 1990s. Indeed, our two rounds of fieldwork took place at a time of tightening bilateral relations, at least on the formal-diplomatic level.¹⁸

CONFLICTING NARRATIVES

Retrospective critiques of the Mbarika pilot can be set against the available data from the time including the narratives recounted by Agridev's experts in their internal exchange, shedding light on the project's many ironies. Probably the most explicit irony relates to the perception whereby the project was self-serving and that farmers were used as little more than cheap labour. Contemporaneous data shows that investment in the project's three sites was astonishingly high, with the Tanzanian government investing over 170,000 GBP in the project (equivalent of 3–3.5 million GBP in 2020) – costs that went towards the upkeep of the sites and part of the foreign staff's salaries. During the project, an Israeli expert assessed that the Tanzanian government was investing some 200 GBP per year per farmer (equivalent of about 3500–4000 GBP in 2020) (Israel National Archives). This is in addition to partial costs bore by Israel that paid part of the staff's salaries and donated equipment, machinery and seeds.

Another example of a narrative gap relates to the involvement of the TANU Youth in Mbarika. In their internal correspondences, Agridev's experts considered the Youth's arrival as an encouraging sign, and were even musing about the Youth eventually taking over the management of the scheme. However, some of our Mbarika interviewees remembered the integration of the youth organisation as suspicious. As local teacher Juma explained, the Youth's arrival early on in the project was interpreted locally as a sign that the government was never fully on board with the foreigners’ presence: ‘the government's leadership initiated the national service here because they had doubts about the Israelis’ presence’ (Juma, Majaliwa and Kulwa 2015 Int.).

Rumours about the mineral deposits reveal another conflict of perspectives. The Tanzanian government, drawing on the region's reputation for cotton and cooperatives as discussed above, allocated the land for the project, and in their internal correspondence, the Israelis often lamented not having a say in the selection of sites. On the backdrop of local concerns about the foreigners’ ulterior motives and their supposed superior knowledge about the availability of minerals in the area, there is special irony in the experts’ repeated complaints about soil quality and their regret that they did not conduct feasibility studies and failed to take soil samples prior to cultivation. To make up for it, they began taking soil samples half-way through the project – a newly introduced activity that might have raised locals’ suspicions regarding mineral extraction. As for the land poisoning narrative, it is conceivable that it might have been inspired by the spraying of pesticides by plane in the last year of the project. The timing of this activity as a last-ditch effort to optimise production may have been interpreted and remembered by some locals as a final act of spiteful revenge. Somewhat circularly, the very presence of the Israelis and the suggestion that they must have known something that the locals didn't may have indirectly contributed to the area's eventual small-scale gold mining boom.

In trying to understand such suspicion towards the Mbarika pilot, it is tempting to follow George's scathing words and interpret all such narratives in relation to the then-recent bitter colonial experience. For example, in Kelly Askew's work on local memories of villagisation in northern Tanzania, one interviewee notes that, ‘in 1968 some people believed that the purpose of getting people to live together was to make it easier for the colonialists to return and rule over us’ (Askew 2008: 113). This feeling can be understood in light of the continuities of a paternalistic and authoritarian tone from pre- to post-independence approaches to development in the country (Eckert 2007: 253–60). In this regard, the Israelis’ full control over the production chain – collecting the cotton in the temporary buying post, transporting it to Mwanza, and selling it in bulk – might have aggravated concern about the return of external domination, since it was not very different from practices used by late colonial cotton cooperatives (Maguire 1969; Schuknecht 2010). The feeling of workers’ estrangement from their produce was such that one interviewee even suggested that the cotton was helicoptered, not to Mwanza but to Nairobi, and from there to Israel. The experts being only interested in agriculture, and not investing in developing community services was further regarded by some as proof that they ‘were only after wealth and their interests’ (George 2020 Int.).

To some extent, this minority view is a familiar trope. As Olivier de Sardan (2021: 301–2) notes, development projects’ reliance on voluntary work among participants can evoke images of exploitation and bring to mind ideas about colonial-era forced labour (also see Monson 2009: 9). More specifically, engagement with long-term retrospectivity raises questions about oral narratives, and how they tend to be encumbered by cumulative layers of later – and perhaps also earlier – experiences and be presented in the service of our present positions (Portelli 1991; Tonkin 1995). While determining the effects of any given historical layer on personal narratives is necessarily speculative, it seems to us that, in the case at hand, we should consider the dramatic shift in national discourse following the adoption of Ujamaa and self-reliance, which increased suspicion of foreign development as potentially self-serving, neo-colonial exploitation. Frequent reference to the Arusha Declaration during our fieldwork emphasised that it was precisely such external capitalistic interventions that the socialist government was pushing against, casting capitalists, both in Tanzania and from abroad, as ‘parasites’ and ‘ticks’ sucking the blood of the country (Brennan 2006). Even though the country's socialist leaning greatly eroded after Nyerere, the fierce rhetoric of the time influenced the casting of pre-Ujamaa development aid as a remnant of foreign dependencies.¹⁹ Indeed, it may be precisely because Tanzania left behind its socialist experiment in the late 1980s to embrace economic liberalism that questions regarding the path towards development and ‘modernity’ have continued to haunt ‘remote’ parts of the country that have not, up to now, benefitted from the country's change of course.

CONCLUSION

In the five decades since the end of the Mbarika pilot, its tangible and intangible legacies continue to live on and to sprawl. In the 1980s or 1990s, an attempt was undertaken to revive the ‘Israeli farm’ – allegedly led by an Indian engineer using a simple downstream irrigation technique – but the attempt failed and was eventually abandoned. A few years later, the government donated Mbarika a new pump, which was placed in one of the two concrete barracks, with a plan – eventually unsuccessful – to reweld and reuse the Mbarika pilot's old pipes.²⁰ In their most recent visit to Mbarika, in the 2010s, the Tanzanian Ministry of Agriculture discovered that the rise of the lake's water levels – the same as for other lakes around the Rift Valley – had left the water pump almost submerged.

Other attempts at reviving the site were of a commercial nature. For example, we were told of a Greek man who, together with a Tanzanian from Arusha, asked to lease the land on a yearly basis in order to grow pepper, but negotiations fell through. A similar story involves a Canadian investor who allegedly asked to rent the land for 20 years, but his proposal was not seen favourably by the community. Yet another story tells of investors of unidentified origin – some say Chinese, some say ‘just African’ – who planted onions using tractors and cows, but gave up after a single season due to high running costs. In the 1990s, we were told, there was an attempt by a commercial company to start a ferry line from Mwanza to Mbarika's ‘Israeli port’, but the initiative was discontinued due to poor demand.

All these initiatives take us back to the notion of ‘afterlives’ with which we began. In the epigraph to this article, we quoted Agridev's report to the Tanzanian Ministry of Land on the eve of its departure, which argues that, while it is hard to assess the Mwanza project's long-term impact, there is no doubt about its ‘progressive influence’ including on farmers and their descendants. The years following the project's termination show that the unequivocal optimism has been overstated. In its Mbarika iteration, the project has indeed been moderately successful in terms of increasing production – certainly towards the last two years of its existence – yet it would be hard to argue that that increased productivity was, in itself, transformative. Still, while the experts’ departure led to a rapid collapse of the project proper, a more circuitous perspective on its long-term material effects may consider, for example, the family homes built with the project's short-lived revenues or the project's modest infrastructure that served as the basis of the village's growth in the following decades.

This question of subtler, indirect impact can also be read in Agridev's epigraph. While the organisation's claim for success seems overly optimistic, its admission that ‘there is no yard stick to measure the [project's] accumulated self confidence of the farmers and its influence on the new generation’ seems prescient. Indeed, it is difficult to assess just how wide the scheme's impact was, certainly beyond the tangible and the local. While we avoided, in this paper, tentative assertions regarding the project's resonance beyond the local level, it is noteworthy that Job Lusinde, who served as a minister under Nyerere, explained Israeli aid and its model of collective agriculture as one of the factors that inspired Nyerere to push for a stronger interventionist stance in the form of villagisation (Bjerk 2010: 296).

On a local level, the project's afterlives are found in lingering feelings of both nostalgia and resentment. For those who remember – or imagine – the project favourably, the question what would have happened had the scheme been allowed to continue leaves a nagging taste of what Piot (2010) had called ‘nostalgia for the future’. Stirring the mind and opening the door for alternative outcomes (see Brownell 2020: 185), the abovementioned Mabula, who participated in the scheme as a young farmer, mused that, ‘had (the Israelis) stayed here even ten years we could be very far in terms of development’ (Mabula and Paul 2015 Int.). In light of the project's collapse as a tangible scheme, its afterlives manifest most strongly in the evolution of (counter)narratives. Rumours about gold or espionage, we argued, are tied to a sense of alienation from the project and its goals. Over time, they become a force in their own right, and continue to orient people's views and aspirations. That such stories would be a key afterlife of a project aimed at sensitising participants towards cooperative self-management is an example of the ironies found in the gap between development's objectives and actual legacies. As Tatu, a Mbarika resident whose grandfather was a farmer in the pilot, summarised things, echoing the language of participatory development, ‘investors or community development practitioners should sit together with the entire community and agree on the terms on how the project should be undertaken, for example how many acres should be cultivated … this is where the project can be sustained … It should not be like our elders, who used to work in the [Israeli] farm as labourers; they didn't even know what was going on’ (Tatu 2020 Int.).