Human Ecology

, Volume 40, Issue 1, pp 59–68 | Cite as

“The Trust is Over! We Want to Plough!”: Social Differentiation and the Reversal of Resettlement in South Africa



In the early 1980s residents of Hobeni, in South Africa’s Eastern Cape, were subjected to forced resettlement, under “betterment” policy ostensibly aimed at soil conservation. They were moved into a spatially contiguous but socially differentiated village. South Africa’s political transition ended this policy, and in the early 1990s, some people, mainly from part of the resettlement area (Kunene) characterized by dense kinship networks who had faced pressure to leave, and began to return voluntarily to their former sites, opting to live in dispersed, flexible settlements. Few people resettled in Mhlanganisweni, a part of the village more diverse in its social composition, returned to their former sites. This research highlights the ways exclusion within “socially-embedded” land tenure systems, together with the layout of resettlement areas and other forms of social and economic differentiation, caused patterns of resettlement to diverge from planners’ intentions.


Involuntary resettlement South Africa Land tenure Planned villages Villagization 


This paper examines the long-term consequences of a resettlement program in the community of Hobeni, in the Eastern Cape province of South Africa, into which people were moved from scattered homesteads into a nucleated residential settlement (de Wet 1995: 26), by focussing on their return to their original sites. The longer-term consequences of resettlement may depend upon the durability of the generosity of host communities: in this case, in part of the resettlement area (Kunene) characterized by dense kinship networks, the forcibly resettled population faced pressure to leave, while those in an area more diverse in its composition (Mhlanganisweni) were better integrated, and few returned to their former locations. By examining a case where resettlement was not permanent for everyone who was moved, I draw attention to the ways in which relations between resettled people and host communities intersect with other forms of social and economic differentiation, shaping the longer-term experience and vulnerabilities of the resettled population.

The paper builds on Chris de Wet’s reflections (1988; 1995) on the Scudder and Colson (1982) framework in the light of villagization elsewhere in South Africa. Scudder and Colson’s well-known four-phase model of responses to resettlement suggests that in the “transition” period immediately after the move, removed communities “[cling] to the familiar” pre-resettlement forms of territorial organization (Scudder and Colson 1982: 273). However, as de Wet points out, this framework does not explain the longer-term variations that may emerge. He points to the ways the spatial layout of resettled communities may hinder or facilitate the preservation of pre-resettlement social ties, illustrating different long-term outcomes in several South African betterment cases (de Wet 1995: 10–15). My argument complements de Wet’s approach through explicit attention to the effects of different forms of social affiliation in what was a single resettlement area, at least in the planners’ abstract mapping of space. The spatial layout and physical configuration of resettled communities may be an incomplete guide to the longer-term outcomes of resettlement in situations where resettled populations and hosts are socially differentiated.

Cernea draws attention to the potential deleterious effects of resettlement on host communities, increasing their vulnerability to impoverishment and “social disarticulation” (1997: 1575). Both the long-term effects of resettlement and the degree and forms of social disarticulation—the “destabilization of community life” (1997: 1575)—are shaped by the relationships between those subjected to resettlement and the existing residents of the resettlement area. However, the case described here reveals ways in which over time people challenge the anomie and insecurity associated with the social disarticulation created by forced resettlement (Cernea 1997: 1575).

Finally, I consider resettlement in relation to the argument that land tenure in rural Africa is “embedded” in society, and thus negotiable and fluid (Berry 1993; Juul and Lund 2002). Peters has countered that contemporary African land tenure practices reveal “social divisions” more often than ambiguity and indeterminacy, and has called attention to “those situations and processes…that limit or end negotiation and flexibility for certain social groups or categories” (Peters 2004: 269). The case of resettlement described here reveals how social divisions may have spatially uneven effects on claims to land. This point was implicit in Downs and Reyna’s early formulation of access to land in terms of “the ease or difficulty of acquiring rights in particular areas” (1988: 12), but has remained underemphasized in the tenure literature, which tends to focus on change over time more than variation across space. In practice, the distribution of people on the landscape may make access uneven, with some settlement areas more willing to provide land to outsiders than others.

“Betterment” in South Africa and Hobeni

Apartheid South Africa was infamous for its spatial interventions through policies aimed at securing land for whites, segregating racial and ethnic groups, and controlling the movement of labor. Less well known than these overtly racist policies is the resettlement of more than three million Africans from scattered settlements into villages, beginning in the 1930s and implemented gradually for decades thereafter (de Wet 1995a: 28). Worried that soil erosion and declining rural production might lead Africans to abandon rural areas for the cities, the South African government insisted that rural Africans should live in “villages”–whether they wanted to or not–and divide their land into residential, agricultural and grazing zones, to be administered by the Department of Agriculture. The government called the policy “betterment,” but across the country, people resisted with mixed success.

The term “betterment” has its origins in a discourse dating from the 1930s that represented Africans as responsible for their own poverty and in need of intervention to promote development. From the late 1940s the policy was renamed “rehabilitation,” likewise connoting a degraded or damaged state in need of repair (cf. Ashforth 1990; Wylie 2001). Neither of these terms entered the lexicon in rural Xhora where, although officials and educated isiXhosa-speakers use the term “betterment,” Xhora residents generally refer to “iTrasti,” “the Trust,” a term apparently having its origins in the 1936 Natives Land and Trust Act, which provided the legal basis for the application of “betterment” in the reserves (de Wet 1995; Jacobs 2003). I generally use the Trust as a translation of the Xhosa term.

Xhora District, where Hobeni is situated, came late to the Trust. The district is part of the Transkei region, an area of South Africa’s southeastern coast inhabited by isiXhosa-speaking peoples and never subject to large-scale white settlement. Since its annexation in the nineteenth century, it was treated primarily as a labor reserve, ruled through village headmen under district magistrates and later chiefs (after the Bantu Authorities Act of 1957). Prior to “betterment,” homesteads were clustered in neighborhoods1 affiliated to subheadmen, situated on the ridges above the streams and rivers that incise the landscape, with gardens adjoining homesteads and fields and grazing land more distant. Since the first decades of the twentieth century, migrant labor has been central to local livelihoods, with state welfare grants becoming more prominent since the 1990s. Cultivation of maize and a range of secondary crops, almost exclusively for home consumption, remains important, however, and livestock provide a primary means of savings and a vital resource for ritual and ceremonial events.

The policy of forced settlement met little success in Xhora district through the 1960s, but the government of the Transkei “homeland” in the 1970s finally overcame this resistance.2 Xhora’s traditional leaders abandoned their resistance to villagization in the 1970s, joining the political camp of Transkei President K.D. Matanzima (Lodge 1983: 281–286). Reports circulated that Matanzima had given instructions that a strip along the Transkei coast should be cleared of people in order to preserve forests and marine life and to create commercial cattle ranches (McAllister 1989: 357). Chiefs and headmen were recorded as giving their assent to planning, and in early 1978, a formal request was made for Hobeni to be planned. In 1979, the Transkei Department of Agriculture and Forestry (TDAF) prepared a hand-drawn planning map for Hobeni,3 with land use zones shaded in colored pencil. Apart from this map, there is little administrative documentation available concerning the Trust in Hobeni (cf. McAllister 1989: 349–350).

The spatial layout envisioned for Hobeni was clearly related to the administration’s longstanding concerns about people living and cultivating in proximity to Cwebe Forest, which was designated a Nature Reserve in 1977: one neighborhood adjacent to the forest, Velelo, would be cleared of residences and fields; three others, MaVundleni, KwaDingata and KuBhula; would be converted entirely into agricultural land; Mhlanganisweni and Kunene,4 along the main road, would be consolidated into one dense residential settlement. In effect, the homesteads far from the road were to be moved closer to the road, while those near Cwebe Forest would be moved further away (Fig.1, Based on Transkei Department of Agriculture and Forestry Map AF6203-79).
Fig. 1

Planners’ Vision for Hobeni after Villagization (1979)

For several years after the maps were prepared no changes appeared on the ground. In aerial photos from 1982, there is no evidence of villagization; homesteads were still scattered around the designated grazing area, and homesteads in the proposed village still had large gardens and open spaces between them. Sometime between 1982 and 1984, removals began in earnest.5 The timing compounded the hardship of “betterment,” falling in drought years in which nearly 30% of local cattle died, according to records at local dipping tanks. Moreover, the removals came shortly after the fencing and closure of Cwebe Nature Reserve, which deprived people of an important source of building materials, as well as marine resources and grazing in the grasslands within the reserves (Palmer et al.2002).

The administration focused on the neighborhoods along the main road closest to the entrance to the Cwebe Nature Reserve, MaVundleni, Velelo, and Mhlanganisweni, effectively creating a buffer zone of roughly two kilometers between the reserve gate and the residential settlement. According to local residents, the Department of Agriculture staff came first to the headman and his councilors, then to each neighborhood, pointing out the borders of the residential areas and plots, and allowing the headman’s councilors and subheadmen to allocate sites within the areas surveyed. People who lived near the forest insist that they had no option but to move; located along the main road, and largely visible from reserve gate, these areas could be easily monitored by the staff of the Nature Reserve. All of the homesteads in the areas by the forest (Velelo and MaVundleni) and some in the southern and eastern portion of Mhlanganisweni were moved and settled among the homesteads of existing residents of the neighborhoods of Kunene and Mhlanganisweni.

“The Time of Politics”: Reversing Forced Removals

A military coup under General Bantu Holomisa in 1987 inaugurated a period that appeared to bring the Transkei to the verge of anarchy (Southall et al.1992: 278–9). However, many residents remember this period as a time of opportunity and freedom, “the time of politics.” The Holomisa government won popularity by canceling a range of taxes, and—most relevant here—by canceling enforcement of villagization. People in Xhora District who had long resisted “betterment” began to return to their former homestead sites beginning in early 1989, while the district administration, plagued for decades by inadequate resources and lack of control over land allocation and use, stopped enforcing villagization.6 State involvement in land tenure in the district effectively collapsed. The Transkei legislation that had mandated “betterment” was repealed in 1996 as part of the process of reincorporating the homeland into South Africa.7

People in Hobeni did not wait for (and probably did not hear about) this legal change. The first people began moving out of the resettlement area in Hobeni in 1993, and by 1999, just under half of those removed had returned. Concurrently, community leaders began to organize a land claim for Cwebe Nature Reserve under the new government’s formal land restitution policies (Palmer et al.2002). From 1995 to 1999 this facilitated the process of returning from the resettlement area by making available a supply of forest products for building.

Patterns of Post-Resettlement Movement

Distinctive spatial and social patterns are evident in the way people responded to resettlement and to the possibility of moving back to their pre-resettlement sites. The households that were affected by villagization in Hobeni can be broken down into three categories: 1) outsiders (from Velelo and MaVundleni) who were moved into the Kunene half of the resettlement area; 2) outsiders (from Velelo and MaVundleni) who were moved into the Mhlanganisweni half of the resettlement area; 3) insiders who were moved from the southern and eastern portions of Mhlanganisweni into the Mhlanganisweni portion of the resettlement area.

Outsiders moved into Kunene tended to move back early on, and nearly all had moved back by 2001. Outsiders moved into Mhlanganisweni, in contrast, were slower to move back; more than half remained in 2001. Moreover, fewer than a third of people moved within Mhlanganisweni have returned to their pre-resettlement sites (Figs. 2, 3, and 4).
Fig. 2

illustrates when outsiders moved into Kunene (n = 32) moved back to their pre-resettlement sites, and those who had not returned as of September 2009 (n = 3)

Fig. 3

illustrates when outsiders moved into Mhlanganisweni (n = 10) moved back to their pre-resettlement sites, and those who had not returned as of September 2009 (n = 4)

Fig. 4

illustrates when insiders moved within Mhlanganisweni (n = 21) moved back to their pre-resettlement sites, and those who had not returned as of September 2009 (n = 15)

This pattern demonstrates that resettlement has had differential effects related not only to the material-spatial location of resettled households, but also to their social position and neighborhood affiliation. As de Wet argues (1988; 1995), attention to the spatial layout of resettlement areas is necessary to understand the longer-term outcomes of resettlement, but is in itself not sufficient. The risks of impoverishment associated with resettlement (Cernea 1997), particularly loss of land, are unevenly shared. Likewise social disarticulation (Cernea 1997) is uneven as is evident in the differing relations between resettled people and the original residents. The sections that follow elaborate these points.

Socially Embedded Land Tenure in the Rural Eastern Cape

The kinship composition of neighborhoods of Xhosa-speakers in the rural Eastern Cape broadly falls into two patterns, with different implications for land tenure in Hobeni: in areas where a limited number of families are predominant, agnatic kinship is the primary means for access to land. In areas of diverse kin composition, other relationships (for example, friendship, church membership, common employment, etc.) are used to gain access to land (Fay 2005: 182). In Hobeni, polygyny was common among some families in the years 1880–1920, a relatively prosperous time for agriculture and cattle-keeping. Polygyny required extensive cattle holdings for bridewealth payments, but it also brought in women’s labor and a claim for additional arable land, enabling increased production. The impact of polygyny on land tenure is evident in patterns of allocation of new land and subdivision.

The contrast between the two portions of the resettlement area is an instance of these differences. The “village” that was demarcated by planners straddled the boundary between Kunene and Mhlanganisweni. In Kunene nearly all residents trace their descent to a few wealthy polygynists of the Bamba and Dingata clans who settled in the area in the late nineteenth century, and few people from outside these clans have settled there since; instead land has only been available to the descendants of the early settlers. In Mhlanganisweni, in contrast, polygyny had been uncommon; no recorded ancestors had more than two wives, and very few had two. Over the course of the twentieth century the neighborhood came to accommodate people from a number of unrelated families. These differences are reflected in the experiences and vulnerabilities of people moved into the two areas.

Returning from the Resettlement Area in the Early 1990s

The first people to return to their pre-resettlement sites in 1993 moved voluntarily and enthusiastically, were relatively prosperous, and employed outside the residential area. In the first case, a newly-married young man returned in order to have a site and garden which could be expanded as his family and livestock holdings grew; in the second, an older man returned because his homestead had already outgrown the small residential site which he had received at the time of the Trust. While their moves might be explained in terms of narrow economic incentives, they spoke of their decisions in political and moral terms that reflect the political climate and ceremonial importance of the natal homestead. As one put it, “we were not afraid to return even though we heard rumors that our homes would be bulldozed.” The ANC was sure to take power after the upcoming elections, and its campaign slogans had called for land to be restored to the victims of forced removals. At the same time, this informant’s decision reflected his homestead’s position in the developmental cycle. He was a first-born son whose father had died. His mother was still living, and in theory, he would have been entitled to inherit her homestead in the resettlement area in Kunene. Instead, he chose to establish a new homestead at his natal home in Velelo.

Returning to this site offered him practical advantages over staying in Kunene: a larger site with potential for expansion and an hour shorter walk to his job at the Cwebe Nature Reserve. However, his new homestead was surrounded by agricultural land and his livestock were likely to trespass in the fields. He solved this problem by staggering the move: when he first returned, he kept his animals at his mother’s site among the fenced homesteads and garden in the resettlement area. About three years later, after other removed people had reestablished their homesteads on their old sites and had begun fencing their gardens and fields, he brought the livestock to the new site. Moving also meant that his youngest brother would be circumcised at an appropriate site. Because most of the ceremonies that accompany circumcision need to be conducted in the cattle kraal, they had to have a kraal in place before the initiation could begin.

The second person to return in 1993–94 was the head of a household in a later phase of the domestic developmental cycle. He had been removed to a site that could not be expanded as his household grew, situated on a small, steeply sloped plot adjacent to a stream. In 1998, five years after he moved, he had one of the largest households in the area, with 13 resident members, three absent members, and substantial livestock holdings. The site to which he returned was one of the largest plots in the area, and the garden he established there was four or five times larger than his entire resettlement site. Nevertheless, he described his decision to move in terms of political and moral claims: he spoke of forced removals under the Trust as part of a single process of dispossession, beginning when his ancestors were moved out of the Cwebe Forest in the 1920s–1930s. Both moving back to his former site, and working on the land claim, he argued, constituted steps to undo the injustices of the previous governments.

As with the earlier informant, he had a personal interest in being close to the Cwebe Nature Reserve after it was reopened to local harvesting in 1994. He was able to retire early from migrant labor in part because of his success as a diviner (igqirha). His new homestead was now only a 15–20 min walk from the entrance to the forest, giving him a plentiful local supply of a range of medicinal plants.

For both of these households, their relative wealth and access to reliable incomes made it possible for them to take advantage of the changed political situation. The other five households to return from the resettlement area in Kunene in 1995 or earlier were similarly well off; all had family members employed as migrant laborers or at the Haven Hotel (situated inside the Cwebe Nature Reserve), with the exception of a recently widowed woman who had insurance money from her husband’s employer. These families were able to begin recreating their old neighborhood rather than staying in the resettlement area. Comparing their experience with that of people who moved later, especially from Kunene, highlights the way relative wealth and poverty affected people’s ability to move back, and conversely, the duration of the stresses and anxieties associated with living in the resettlement area.

Pressure to Move from Kunene

If “pull” factors were important to the first people to return, “push” factors became more important as the newcomers in Kunene came to feel less and less comfortable living among a few closely-related families. Tensions on both sides were expressed in terms of “building the homestead.” As McAllister explains, this phrase has material, religious and social connotations:

Going out to work as a migrant, and successfully practicing agriculture at home, provides the homestead with material assets that enable it to enter into a mutually satisfying relationship with the ancestors, through ritual, and with the wider society. Performing a ritual for the ancestors will bring forth blessings that are manifested in material and social terms, and the ritual itself involves a series of social, economic and religious actions. Using the homestead for the benefit of others will elicit reciprocal benefits and political support, and is pleasing to the ancestors not only because it is morally good, but also because it ensures the economic well-being of the homestead, and thus that future ritual obligations will be met. (McAllister 2001: 178)

While others have emphasized the relation of “building the homestead” to notions of order (Kuckertz 1990) and sociability (McAllister 2001), I suggest here that this idiom plays into a “politics of harmony” similar to that described by Rose in land disputes in Swaziland (1991). Rose notes that researchers examining land disputes in rural southern Africa seldom find that few people are willing to discuss conflicts directly; instead they invoke a “harmony ideology” in order to further their interests while maintaining the appearance of social unity (1991: 192). People in southern Hobeni were reticent about conflict and threats related to the Trust. Social disarticulation in the form of tensions between newcomers and the original residents of Kunene was expressed indirectly through appeals to socially positive values. The neighbors of resettled homesteads would appeal to their need to build and expand their own homesteads and to provide sites for their sons, arguments which are considered legitimate and difficult to challenge. In interviews and daily conversations, they were reluctant to admit that they would threaten anyone, while removed people, particularly those still living in the resettlement area, were reluctant to accuse their neighbors of threatening them.

The changing connotations of the phrase “iTrust iphelile,” “the Trust is over,” likewise show how a statement of shared values can nevertheless hold very different strategic meanings. In the early 1990s, this was a celebratory statement indicating freedom from government control, and a reassertion of local autonomy. For people who had been forcibly removed, it meant that they could return to their former sites without fear of punishment. However, by the end of the decade, people who had not been removed would use the phrase as an indirect threat. Rather than invoking freedom, it meant that those had been removed by the state should now return to their original sites or face unspecified consequences.

People who had returned in the late 1990s described a change in the tenor of their relationships with their neighbors in the last few years, after most of the people removed to Kunene had already returned to their former sites. A man who had recently returned to MaVundleni explained that the first people to move back “went on their own, they left on their own, they weren’t chased.” He continued, “we were staying together well up there. [But] it started getting bad when people said ‘the Trust is over;’ they began to intimidate us. They said they wanted to plough.” At this point, my research assistant pointed out, “but they’re not ploughing.” He continued, “no, [but] they’re chasing us away now. They say, go, get out of here–I want to plough here!”

Others questioned the statement that people had been “staying well” in the past. As one woman who had returned the year before explained, “people there would kill your goat if it got into their garden–they were jealous, because they wanted to expand their gardens.” Such forms of retribution, though, still avoided the need for direct verbal communication of threats. She continued, “they would say nothing, but they would speak with their actions.”

With nearly two-thirds of the removed households returned as of 1998–99, many people who had not yet returned said that they hoped to move because of their isolation from kin who had already relocated. One family with a successful shop and an expensive cement brick house in Kunene moved despite the loss of these investments; as one observer put it, “all her neighbors from Velelo had already returned. She was feeling that she was the only person from Velelo left and was concerned about her neighbors.”8

The original residents would almost never speak aggressively about the need for the outsiders to leave, but often portrayed their going as a process that would restore land that they lost through villagization. A woman at one Kunene homestead which lost a productive garden to the newcomers said of the process, “we’re not sure what will happen, but we see a need to get our land back now that those who were removed are [moving back to] their rightful places.” This quotation is typical of the way people discussed the process in that the people moving back are not referred to by name. When discussing potentially sensitive situations, people would generally speak with indirect references or descriptions of attributes of the person (“that young woman we were speaking of,” “that guy who has the ferry on the Mbhashe,” etc.).

Spatial messages also provide a way for original residents to express their intentions without resorting to verbal confrontation. As other authors have noted (e.g., Perry 1981; Rose 1991: 116), land disputes in southern African societies often begin with one party ploughing a piece of land as a means of testing another’s resolve, rather than direct confrontation or a formal dispute. The usual activities involved in the physical aspects of building the homestead–ploughing or fencing a garden, building or maintaining a house or livestock enclosure–could imply or be interpreted as a threat. One Kunene resident made an indirect claim to a still-occupied “betterment” site by expanding and fencing his garden until it was nearly adjacent to his neighbor’s houses. At the same time, his neighbor was cutting grass and hiring thatchers to maintain her homestead in Kunene, rather than returning to her former site. Such an approach to pressuring people to move enables people to avoid an accusation that they are intimidating their neighbors; they justify their actions by stating that they are simply working on their own homesteads, even as they convey an intention to expand their land further.

The most elaborate form of indirect communication was the revival in late 1999 of an agricultural ceremony which had not been performed in roughly 25 years. The residents of Kunene organized the inqoloqo ceremony, in which a chain of young girls with maize cobs on their heads travelled through the neighborhood singing but visiting only the homesteads of the original residents of the neighborhood. It was immediately deemed effective when on the first day it was performed a soft but steady rain fell for most of the afternoon. But beyond its ostensible purpose, by visiting only specific homesteads, inqoloqo was in effect publicly demarcating membership in the neighborhood.

By the end of 1999, the combination of “push” and “pull” factors had led most people moved into Kunene to return (see Fig. 2). Moreover, they had returned despite the fact that this meant that they would be losing agricultural land: because the state had not reallocated agricultural land after moving people, removed people had generally converted their former residential sites to fields, effectively increasing the arable land at their disposal. This gain, however, was not worth the discomfort of living in Kunene.

Those who had not returned from Kunene generally described themselves as too old or too poor to move. Wealth-ranking card sorts and a household survey confirmed this assessment: their livestock holdings were below average, they held fewer than average work parties (an essential measure for building a new house), and they were twice as likely as returned households to have a household member who had been sick for more than six months. Forced to move by the government a decade before, they were now vulnerable to pressure from their neighbors as well, forcing them to devote resources to trying to prepare to return even as they were struggling with poverty and weakness.

Justifying Decisions to Return: Moving as a Moral Project

While the notion of building the homestead provided an indirect means for original residents to pressure removed people, it also allowed some of those who were pressured to move to reframe their actions in a positive light. Given that the cattle kraal at the center of the homestead is the site of sacrifices and often the site of ancestral graves, ritual frequently provided a means to combine returning with events that restate the importance of commitment to building the homestead. The most important event which people integrated with their return is the circumcision of young men. The natal homestead is the point of departure for the “separation” phase of this rite of passage; kin of each initiate and a dozen or so guests gather at the initiate’s homestead for the shaving of the initiate and the slaughter of a goat before the initiate moves to the homestead of the father who has sponsored the event. After four to six weeks in an isolation hut (itonto), a ritual of reintegration (umnyatheliso, lit. “making him take a step”) marks the first time they return to their natal homesteads since their circumcision. Finally, their natal homestead forms each initiate’s final destination at the end of the ceremonies on their departure from itonto.

The speeches that accompany circumcision make explicit the connections between returning and building the homestead, and publicly express a positive interpretation of the move. The following case illustrates the way in which an impoverished family used their sons’ circumcision ceremony as a way of both representing their return to MaVundleni in a positive light, and in the longer term, making their sons into responsible adults.

I first spoke with a woman in December 1998, at which point her household seemed to be too poor to try to move out of Kunene. The houses at her villagization site had collapsed, and all eight of the family members were sharing a single hut, borrowed from someone who had already moved back. Without cattle or sufficient labor, they had left their field fallow since the time they were removed in the 1980s. They had begun building a house at their old site, but it was unthatched, and she wasn’t sure when they were likely to move. Their prospects were not good: one of their sons had just returned from Gauteng after a few months unsuccessfully searching for work. Her husband complained of isithukuthezi, translatable as a mixture of anxiety and melancholy. He was unable to work after losing most of his right hand when a mine shaft collapsed on him. His disability pension and occasional work with fishermen at the holiday cottages in the Cwebe Nature Reserve were their only source of income.

A few months later, I attended a work party at their new house; it was still not complete, but they had managed to organize workers to make some progress on construction. At the time, two of their sons were reaching the age where they were seen as dangerous and in need of the subduing influence of circumcision. A month later, the family had moved into the single completed hut at their new site. He made a point of explaining that he was preparing his new homestead to be ready by the time his sons were circumcised, showing me the poles he had gathered to construct the kraal for the preparatory sacrifices. “Ndiyafa,” he said, “I am dying,” referring to his injury in the mines and a series of subsequent ailments, “but I want to see my sons circumcised before my death.” Returning and hosting the preparations for circumcision at their old site provided a way of transforming this necessity into action consistent with locally shared values.

By subdividing existing homesteads in order to allocate land to people being forcibly removed, villagization in Hobeni sowed the seeds for later conflict. By the late 1990s, those who were removed into Kunene faced pressure to return, making the best of difficult situations as they moved back. People who could afford to move out did so fairly soon, while others took longer, struggling with poverty and illness; by 2009, all but three of those forcibly moved into Kunene had returned to their pre-resettlement sites (Fig. 2). Rather than live among unfriendly neighbors, they sought to restore their social and spatial neighborhoods in their pre-resettlement sites.

Remaining in Mhlanganisweni

The contrast between the later actions of people forcibly relocated in Kunene and those moved to Mhlanganisweni is pronounced (Figs. 2, 3 and 4). Fewer people have returned from Mhlanganisweni to their pre-resettlement sites, and those that have, have done so at a slower pace. Despite being subject to the same centrally-planned project of forced resettlement into what appeared on the map as a single resettlement area, people in Mhlanganisweni have had substantively different experience since resettlement.

In the late 1990s, people who were moved to Mhlanganisweni by the Trust, would talk about returning to their prior homesteads as an option, but without anxiety or urgency, and they did not experience identifiable pressure to return. Even relatively wealthy people who could afford the expenses of moving often chose not to return and instead invested in their resettlement sites. Rather than living in an area dominated by a few agnatic extended families, as in Kunene, they were in an area of diverse kin composition which had historically been more welcoming to incomers.

Notably, despite the patrilineal emphasis in rural isiXhosa-speaking society, the land rights of women in Mhlanganisweni appeared secure. Four households headed by widows who had married into local families were among those staying put: despite being outsiders both in terms of their pre-marital and pre-villagization origins, they did not face particular pressure to return, and none of them were planning to do so. One explained in 1999 that she had no reason to: she got along well with her neighbors, and she had no sons to inherit her pre-“betterment” site, so she was happy to stay in her site in the resettlement area; moreover, she was adding a new flat to her homestead to accommodate her unmarried daughter and her daughter’s children.

The scale of internal removals in the two neighborhoods also differed substantially. In Kunene, all but two people were moved in from other neighborhoods. In Mhlanganisweni, however, roughly one half (18 of 37) of the homesteads already within the boundaries of Mhlanganisweni, comprising the southern half of the neighborhood, were forced to move under “betterment.” As a result, the people removed within Mhlanganisweni were in a different position from all others who were removed. In Cernea’s terms, they experienced less “social disarticulation” (1997: 1525): they were moving to an area occupied by people from the same neighborhood, with whom they already had a history of social, ceremonial and economic relationships, including kinship ties, collaboration in work parties and ploughing companies, and organization of beer drinks and larger ceremonial events.

The net result is a situation where the original residents of Mhlanganisweni are more sympathetic and less antagonistic towards removed people in general. Neither people moved within the neighborhood, nor those removed from outside have faced pressure to return to their pre-resettlement sites (see Figs. 3 and 4). While the outsiders removed from the areas by the forest had no agnatic and few affinal ties to their new neighbors, Mhlanganisweni residents were nevertheless accommodating of them.

The patterns of land use that resulted from the Trust also gave people removed within Mhlanganisweni another important reason not to return. Unlike in Kunene, where local residents simply lost land to outsiders, in many cases the incoming residents of Mhlanganisweni actually increased their arable landholdings as they were able to convert old house sites to fields.9 This gain in arable land was viewed as a beneficial effect of the Trust. The name Mhlanganisweni refers to “a meeting place,” and residents would interpret this name in characterizing its territorial limits: “we are surrounded here where everyone meets; we are in the middle, with no place to expand.” Prior to “betterment,” Mhlanganisweni residents explained, the neighborhood had been suffering from a shortage of arable land.

By converting sites to fields, people created both an incentive not to relocate and an obstacle for those who would do so. Around 1995, rather than move back, the owners of about 12 of the former Mhlanganisweni sites (converted to fields) collaborated to fence the area, led by a polygynist who had the wealth and labor to cultivate all four of his former homesteads. Anyone attempting to establish a new homestead here would risk having their livestock damage other people’s crops, with the quarrels and fines that would entail.


The case of Hobeni illustrates how social differentiation may lead to diverse outcomes in a single resettlement area. The spatial layout of the resettlement area relative to pre-existing settlement pattern is important (de Wet 1995), but different social ties and disarticulation between resettled and host populations are also critical to the long-term experience of resettlement (Cernea 1997). Socially embedded land tenure does not mean that access to land was necessarily equally flexible or negotiable across the resettlement area (cf. Peters 2004). Differences in the social composition of neighborhoods can lead to differences in how outsiders are accepted and hardship shared.

By drawing attention to intra-community variation, this case also highlights the limitations of a focus on the abstracting and simplifying logics involved in the planning of villagization schemes in southern and eastern Africa (e.g., Drinkwater 1991; Scott 1998; Moore 2005), and the racialized application of these policies (e.g., Delius and Schirmer 2000, Jacobs 2003; Moore 2005). While these certainly shaped the conditions that led to “betterment” policy and implementation in Hobeni, to restrict the analysis to “the relationship between the dominant and the subordinate” would involve a kind of “ethnographic refusal” to engage local political complexity (Ortner 1995: 176). Without explicit attention to the relationships between the spatial layout of resettlement areas and the social networks that may bind and/or differentiate resettled people, the analysis of resettlement may miss the eminently local but forceful factors that shape longer-term outcomes.

The policy lessons from this case mostly illustrate practices to avoid. As Cernea has argued, planners “tend to overlook” issues of social disarticulation and community reconstruction, “and are rarely concerned to facilitate reintegration within host populations” (1997: 1581). “Betterment” in South Africa’s Transkei was no exception, with little evidence for anything approaching the level of planning that would be required for a contemporary World Bank project, or even the detailed planning found in other places and times in apartheid South Africa (de Wet 1995). Instead, by subdividing existing sites and placing removed people among existing homesteads, planners created antagonisms that would only add to the hardships of resettlement. This case highlights the importance of planning—where resettlement is absolutely unavoidable—in ways that minimize potential conflict between newcomers and host populations; to do so requires attention to the details of local social organization and landholding, and both spatial and social placement of resettled people in ways that facilitate maintenance of existing ties and avoid creating new antagonisms.


  1. 1.

    These groups, isixeko or isiphaluka, are variously referred to in the literature as “neighborhoods,” “sublocalities,” “village sections,” and “subwards.” I prefer the term “neighborhood” because it suggests cooperation and frequent interaction among the members of these groups.

  2. 2.

    Faced with opposition to apartheid, the South African government aimed from the 1960s to create the appearance of reform by creating “homelands” for Africans.

  3. 3.

    Correspondence in Department of Agriculture offices, Mthatha; file A5 1-005 (27) “Planning and Reclamation - Hobeni A.A.- Elliotdale - Vol. 1,” and Transkei Department of Agriculture and Forestry Map AF6203-79.

  4. 4.

    For the sake of clarity, I refer to this area as Kunene, the short form of Kunene bukaNoMama, rather than its more common name, MaBambeni.

  5. 5.

    I use the terms “removal” and “remove” to describe the state’s action in forcing people to move from their scattered homesteads because they capture isiXhosa usage. The verb ukususa (“wasisus’ uRhulumeni–“the Government moved us”), and the passive form ukususwa (“sisuswa ngeTrasti”–“we were removed by the Trust”), have connotations of disposing of something unwanted. Similarly, I translate the term ukubuya, used to refer to a move back to pre-betterment sites, as “return,” although it has some additional connotations that give it a particular moral force; it is used e.g., to describe the return of migrant laborers to their homes or in the slogans of the liberation movement–“mayibuye,” “let it [Africa] return,” with implications of restoration and righting injustice.

  6. 6.

    JZ Mbude, Magistrate, Elliotdale letter of 26 March 1989 to Agriculture Officer in Charge, Office of Agriculture & Forestry, Elliotdale; file housed in Botha Sigcau Building, Mthatha.

  7. 7.

    Michael Coleman, pers. comm.

  8. 8.

    Contrary to what one might initially expect, witchcraft was surprisingly absent in peoples’ accounts of pressure to return. This becomes more understandable when one considers that witchcraft accusations tend to focus on behavior that violates social norms. This is not to say that removed people did not fear witchcraft; but they found it unnecessary or inappropriate to speak of clandestine or anti-social measures like witchcraft, given the broadly-based opinion in Kunene that removed people should return. In general, the accounts of witchcraft I heard in Hobeni were related to intra-household disputes. Unlike wives secretly killing allegedly unfaithful husbands or brothers killing their brothers in inheritance disputes, people reclaiming land that was “rightfully theirs” had the moral high ground, and the opinions of most of their neighbors in Kunene in their favor by 1998–99. The absence of accusations of witchcraft in this context appears related to the degree of consensus on the rights of people in Kunene to reclaim their land. Both people from Kunene and those who are leaving the area for their former sites speak of the situation in terms of correction of injustice.

  9. 9.

    These fields are within a kilometer or so of the residential area, and are situated downhill so that most people with fields can watch for trespassing livestock from their homesteads. A few people here found themselves with more land than they could cultivate, and one donated land for use as a site for a new school.


  1. Ashforth, A. (1990). The politics of official discourse in twentieth-century South Africa. Clarendon, Oxford.Google Scholar
  2. Berry, S. (1993). No Condition is Permanent: the Social Dynamics of Agrarian Change in Sub-Saharan Africa. University of Wisconsin Press, Madison.Google Scholar
  3. Cernea, M. (1997). The risks and reconstruction model for resettling displaced populations. World Development 25(10): 1569–1587.CrossRefGoogle Scholar
  4. Downs, R. E., and Reyna, S. P. (1988). Introduction. In Reyna, S. P., and Downs, R. E. (eds.), Land and Society in Contemporary Africa. University Press of New England, Hanover.Google Scholar
  5. Drinkwater, M. (1991). The State and Agrarian Change in Zimbabwe’s Communal Areas. Macmillan, New York.Google Scholar
  6. Fay, D. (2005). Kinship and Access to Land in the Eastern Cape: Implications for Land Reform. Social Dynamics 31(1): 182–207.CrossRefGoogle Scholar
  7. Jacobs, N. J. (2003). Environment, Power, and Injustice: A South African History. Cambridge University Press, Cambridge.CrossRefGoogle Scholar
  8. Juul, K., and Lund, C. (2002). Negotiating property in Africa. Heinemann, Portsmouth.Google Scholar
  9. Kuckertz, H. (1990). Creating Order: The Image of the Homestead in Mpondo Social Life. University Press, Johannesburg.Google Scholar
  10. Lodge, T. (1983). Black Politics in South Africa since 1945. Longman, New York.Google Scholar
  11. McAllister, P. A. (2001). Building the homestead: agriculture, labour and beer in South Africa's Transkei. Aldershot, Hampshire, England, Ashgate.Google Scholar
  12. McAllister, P. (1989). Resistance to ‘Betterment’ in the Transkei: A Case Study from Willowvale District. Journal of Southern African Studies 15(2): 346–368.CrossRefGoogle Scholar
  13. Moore, D. S. (2005). Suffering for Territory: Race, Place, and Power in Zimbabwe. Duke University Press, Durham.Google Scholar
  14. Ortner, S. B. (1995). Resistance and the Problem of Ethnographic Refusal. Comparative Studies in Society and History 37(1): 173–193.CrossRefGoogle Scholar
  15. Palmer, R., Timmermans, H., and Fay, D. (2002). From conflict to negotiation: nature-based development on South Africa’s Wild Coast. Human Sciences Research Council, Pretoria.Google Scholar
  16. Perry, J. (1981). Land, Power and the Lie. Man 16(2): 235–250.CrossRefGoogle Scholar
  17. Peters, P. E. (2004). Inequality and Social Conflict Over Land in Africa. Journal of Agrarian Change 4(3): 269–314.CrossRefGoogle Scholar
  18. Rose, L. (1991). The Politics of Harmony: Land Dispute Strategies in Swaziland. Cambridge University Press, Cambridge.Google Scholar
  19. Scott, J. (1998). Seeing Like a State: How Certain Schemes to Improve the Human Condition Have Failed. Yale University Press, New Haven.Google Scholar
  20. Scudder, T., and Colson, E. (1982). From Welfare to Development: A Conceptual Framework for the Analysis of Dislocated People. In Hansen, A., and Oliver-Smith, A. (eds.), Involuntary Migration and Resettlement: The Problems and Responses of Dislocated People. Westview Press, Boulder.Google Scholar
  21. Southall, R., Segar, J., and Donaldson, A. (1992). Transkei beyond the transition: Towards good government or back to the frontier? Journal of Contemporary African Studies 11(2): 270–279.CrossRefGoogle Scholar
  22. de Wet, C. (1988). Stress and Environmental Change in The Analysis of Community Relocation. Human Organization 47(2): 180–187.Google Scholar
  23. de Wet, C. (1995). Moving together, drifting apart: betterment planning and villagisation in a South African homeland. Witwatersrand University Press, Johannesburg.Google Scholar
  24. Wylie, D. (2001). Starving on a Full Stomach: hunger and the triumph of cultural racism in modern South Africa. University of Virginia Press, Charlottesville.Google Scholar

Copyright information

© Springer Science+Business Media, LLC 2012

Authors and Affiliations

  1. 1.Department of AnthropologyUniversity of California, RiversideRiversideUSA

Personalised recommendations