Are concepts such as ‘margins’ and ‘marginalisation’ useful for analysing rural life in the Eastern Cape Province, South Africa?

Abstract

This paper draws on empirical material gathered in the Eastern Cape Province, South Africa, an area that includes what used to be known as the Transkei, to consider the nature and usefulness of the concept ‘marginalisation’. It shows that structural factors alone cannot account for the marginalisation of individuals or groups within a community, and that to understand why some may be marginalised, and others not, the nature of the community's social and moral relationships and how these are linked to economic production must be considered. The morality that governs economic relations is evident in beer-drinking rituals and inter-homestead cooperation (work parties). The findings are theorised in terms of Bourdieu's analysis of various forms of capital and the exchanges between them, and this is linked to notions of ‘moral economy’ and the moral import of ritual as argued by Rappaport. The paper concludes by discussing some implications for rural development policy.

1. INTRODUCTION

This paper is concerned with the question of living on the margins, but it argues that marginalisation is best viewed as a relative concept rather than an absolute concept. There are regions and communities in South Africa, both urban and rural, which might be considered economically ‘marginalised’. However, within these regions and communities there are usually economic survival strategies that some people pursue more successfully than others, and from which some are excluded or exclude themselves, and there are some people who are ‘marginalised’ in relation to the larger community or region. So the answer to the question posed in the title of this article is that the concepts ‘margins’ and ‘marginalisation’ are useful only if they are qualified in this way. The danger of not doing so is to homogenise and essentialise regions or communities. Marginalisation is also best seen as a relative concept in the sense that people may be economically well off but marginal in some other way – socially or culturally or geographically, for example. Sometimes, of course, different kinds of marginality will coincide, as demonstrated below in Section 2.

But there is more to it than this. We need to understand why, in a particular community, some are socially and economically marginalised and others not, and why some are more marginalised than others, or less, and how this might not be a permanent state of affairs but something that may change over time. One way of making sense of marginalisation, or of degrees of inclusion and exclusion (terms that may be preferable), within territorial communities is to examine the structure and organisation of social relations in a community and the values on which these are based. Since there may be structural determinants of marginalisation and poverty that affect communities or regions, one can analyse inclusion and exclusion within communities in terms of structural factors (e.g. age, gender, kinship) and their associated value principles. However, these structural and value factors by themselves do not necessarily determine inclusion, exclusion or marginalisation. Practical (organisational) and contextual factors have to be taken into account, and detailed observation over time is required to illustrate how these are relevant and to monitor and record changes in them. To demonstrate these points, this paper draws on ethnographic evidence gathered in one rural community over a long period of time.

This was a case study of the residents of a sub-ward in the Shixini administrative area in the Willowvale district of the former Transkei, now part of the Eastern Cape Province. Within this community, inclusion and exclusion (marginalisation) are usually context dependent, variable over time, and related to the economic strategies people pursue and the values that underpin them. In other words, it is necessary to consider the community as a moral community of a territorial kind, and the nature of the local economy as a moral economy in the sense used by Thompson (1968) in his study of the English working class, and as applied by Scott (1976) in his work on peasants in Southeast Asia. To grapple with these questions as regards Shixini, the paper firstly uses an anthropological and ethnographic framework: a social analysis of aspects of the community based on the author's experience of living in it and observing its activities over a period of 25 years, and on repeated observations and the identification of patterns and regularities. The analysis avoids the homogenising and essentialising tendencies referred to above that are sometimes associated with concepts such as ‘marginalisation’ and ‘social capital’. Secondly, the paper presents some ideas on how one might theorise these observations, and finally it concludes with some remarks on the policy implications of the findings.

The findings have important implications because they signal a need to identify and address differences within communities that may be seen as marginalised (and probably also within communities that may seem not to be marginalised, making one question again the usefulness of the concept ‘marginalisation’), and because this paper stresses not just the economic and social factors but also the moral factors that help to account for such differences. This in turn makes it possible to suggest ways of dealing with marginalisation and conceptualising development – because the regularities and the patterns that are observed include those that enable marginalised communities both to survive and to exclude those members who threaten their survival strategies. Such exclusion is linked to the moral evaluations that are made of people's actions in particular contexts or circumstances.

2. ECONOMIC LIFE AND SOCIAL RELATIONSHIPS IN SHIXINI

The Shixini sub-ward community of around 80 homesteads with which the author is familiar is not, like much of the Eastern Cape, characterised by de-agrarianisation (defined as a move away from agricultural activities and into alternative means of making a living in rural areas). Instead, most homesteads, while dependent for many things on cash from remittances and/or social welfare, also actively work the land every year, and most keep livestock of one kind or another. Maize yields, as has been demonstrated elsewhere (McAllister, 2001), are reasonably good, and many other crops are grown as well, although usually only in small quantities for household use. Extensive use is made of natural resources such as thatching grass, wood for fuel and construction, clay for building bricks, and so on. The maintenance of homestead production implies a worldview and lifestyle very different from, say, people from the more overcrowded and modernised parts of the Eastern Cape, or those historically subjected to agricultural ‘betterment’ schemes, which seriously disrupted their social institutions and sense of community and made it virtually impossible to continue such production.

This does not mean that Shixini is highly unusual. Although it is a conservative and relatively isolated coastal community, the findings presented in this paper can be generalised to other similar communities throughout the coastal parts of the former Transkei, particularly those that have escaped betterment or other forms of social disruption. Detailed data from the adjoining district of Elliotdale (Bomvana) bear this out (Fay, 2003).

On the one hand, it is possible to identify a structure of social relationships in Shixini based on factors such as gender and age hierarchies, political status, clan affiliation and kinship relations. Other forms of hierarchy include patron–client relations and relations between employers and employees, in the case of the few small business enterprises that exist. On the other hand, these relationships are overlaid with an egalitarian ethic and are organised into territorial groups of up to 30 neighbouring homesteads, each being a section of a sub-ward. Much of the economic life of individual homesteads is organised on a hierarchical basis, with homestead heads (male or female) as primary decision-makers. But economic relations between homesteads within a section and sub-ward are largely egalitarian, at least in principle, and these relations are crucial for maintaining homestead production.

As a territorial and political unit, the sub-ward is economically important in that its homestead heads, in conjunction with a sub-headman, play a vital role in many aspects of social and economic life. This includes the allocation of arable land and building sites, the control of grazing and other natural resources, and the organisation of larger cooperative work groups. But most everyday cooperation and mutual assistance – both informal or ad hoc and of a formal, organised kind (in the form of cooperative work parties) – takes place within sections of a sub-ward, closely knit neighbourhood groups that often also contain clusters of closely related kin. Often there are also close ties of this kind between certain sections within a sub-ward, or between sections of adjoining sub-wards, and (on a larger scale) between territorially contiguous sub-wards. However, normative arrangements of collaborative work between sub-wards are invoked fairly seldom, and basically the principle of inter-homestead cooperation is one where the further apart homesteads are, the less they collaborate economically. This is, of course, relevant to the question of inclusion and exclusion but, theoretically at least, each homestead is at the centre of a network of similar interdependent entities, and excluded from those that are too far away to be of practical importance.

In the literature on the sociology of development these aspects of social structure and organisation are referred to as ‘social capital’, a widely discussed notion that ‘captures the idea that social bonds and social norms are an important part of the basis for sustainable livelihoods’ (Perry & Ward, 2001:210). Social capital is seen as a resource actors can use in pursuing their individual productive interests, consisting of the social structure and organisation that connects them, and constituted by the networks of social relationships between them, backed by notions of reciprocity, trust, and an ethic of mutual assistance (Coleman, 1988). Generally speaking, the literature on social capital has not been widely adopted by anthropologists, probably partly because it appears to condense in a relatively simplistic way the complexities and interconnections of social life with which anthropologists are primarily concerned, and partly because there is a tendency to reify social organisation and its underlying values as ‘social capital’. In Putnam's well-known views on social capital as a property of whole communities and even nations (Putnam, 1993, 1995) there is, as Portes has pointed out, a tautological argument, a logical circularity, with social capital being both cause and effect (Portes, 1998:19). On the other hand, the notion of social capital has received sophisticated theoretical treatment in the work of Bourdieu (1977, 1988), whose ideas are referred to in Section 3 below. Bourdieu's notion of capital (which includes social as well as other kinds of capital, and the exchanges between them) is useful in making sense of inclusion and exclusion in Shixini, particularly when used in conjunction with the idea of a ‘moral economy’.

One of the most striking features of life in Shixini, both despite and because of a heavy reliance on labour migration, is the dense network of relationships between homesteads based on both neighbourhood and kinship, involving collaboration and interaction in a variety of agricultural and other activities. It is largely due to this that homestead production is possible. Kin and neighbours pool their animals to facilitate herding arrangements, form various kinds of work parties to overcome shortages of labour, implements and oxen, and assist each other in a variety of small but significant ways on a day-to-day basis. Combining labour is also a means of redistribution and, to many households, providing labour for others is an important (although always partial) means of securing subsistence, as well as reciprocal labour at some time in the future. This has been documented in detail elsewhere (McAllister, 2001), so the various types of formal and informal labour combinations are not described here.

An important topic that has also been addressed, in other publications, is the question of the political and economic efficiency of cooperative work groups. Some of the literature on cooperative work in Africa claims that it is inefficient or exploitative, and that those who host and benefit from work parties tend to be the better-off homesteads, while those who do the work are largely the poorer ones. (For a review of such literature, see McAllister, 2004.) Such a tendency is present in Shixini, as indicated in this section, but it is only a tendency, and on balance the evidence suggests that the relationship is not usually exploitative. This is also the case in other parts of Africa, where, it has been argued, poorer homesteads would simply not be able to produce at all without these cooperative labour arrangements (McAllister, 2001, 2004).

What does cooperative labour mean in terms of inclusion and exclusion within sub-wards and within sections of these? In Shixini the members of each homestead have an obligation to develop the homestead, an ethic that is summed up in the oft repeated phrase, ukwakha umzi (‘to build the homestead’). Except in the case of the very wealthy, and then at the risk of social and moral exclusion, this individual imperative can only be achieved through cooperation between a number of homesteads. Each homestead's well-being is of concern to the wider community, because other homesteads depend on it. Each homestead's assets, in the form of labour, equipment, livestock, and so on, are in a sense both its own assets and communal assets – they are combined with the assets of others in order to achieve each individual homestead's economic goals.

Morality, the judgements made about the actions of others, turns on this interdependence, which is basic to the survival of homestead production. Scott (1976) claimed that at the basis of peasant morality is a subsistence ethic, a rational principle that prioritises economic survival. Peasants judged the actions of others (such as the government) according to whether they helped or hindered the struggle to survive, and Scott sought to explain peasant revolt in this way. Beinart and Bundy (1987) provided a similar explanation of the Pondo revolt in the early 1960s. Anthropologists and others have at times romanticised tribal and peasant communities as altruistically seeking to ensure the welfare of each homestead, and this apparently un-self-interested approach has been seen as the basis of their morality, contrasting with an alternative model of peasant behaviour as economically rational and self-interested. A similar contrast within sociology between rational choice theory and sociological approaches that saw the actor as shaped by the social context inspired Coleman's development of the idea of ‘social capital’ as an attempt to resolve this problem (Coleman, 1988). Bourdieu puts it somewhat more succinctly – his overall theoretical approach is an attempt to resolve this divergence between agency and structure.

The morality of Shixini people clearly combines elements of both individual and communal interest. It effectively seeks to ensure the survival and well-being of every individual homestead in the community, but is at the same time a matter of self-interest. This is because each homestead that dies out is a threat to the economic survival of the others in a situation where this survival depends on cooperation between homesteads. However, this cooperation is an ideal, and does not translate into unqualified support for every homestead. People who cut themselves off from the community or who act contrary to local morality and do not contribute to the economic survival of others by helping with labour and other resources cannot expect reciprocal assistance, and may be marginalised and excluded. Ties of kinship and common residence are not enough. Reciprocity is basic to the Shixini people's moral economy, but it does not exist as a resource that can be drawn on at will; it requires action that demonstrates commitment.

This combination of self-interested economic rationality and support for communal well-being can be seen, for example, in the organisation of what are called agricultural companies (iinkampani zokulima). Almost every Shixini homestead belongs to one. They consist of from three to a dozen or so homesteads that combine their mechanical, livestock and labour resources to cultivate each other's lands in rotation each year. The leading members of a company are invariably those with the most resources to offer, and they receive priority when it is time to cultivate. But most companies include resource-poor homesteads, often headed by widows, who can offer nothing more than their labour and their political support. They are marginal members, in a sense, but members nevertheless, and without this membership they would not be able to cultivate, which would jeopardise the viability of their homesteads. And it is in the interests of the resource-rich members to ensure this viability, not just because they need labour and allies, but because life is fragile and they may not remain well off indefinitely. All are aware of many cases of wealthy homesteads being impoverished virtually overnight by factors such as livestock diseases, illness, job loss and crop failure. On the other hand, a poor homestead headed by a widow today may turn into a relatively resource-rich homestead tomorrow if circumstances change; for example, if a long absent migrant returns home, a pension application is granted, or long awaited remittances start to arrive. Gender and marital status alone, which are structural factors, do not automatically lead to marginalisation. Some agricultural companies include, as leading members, homesteads headed by widows. It is the physical resources that the homestead can offer, and the willingness with which it offers them, that are important.

One other example of cooperative work will demonstrate how hierarchy, egalitarianism and a degree of marginalisation may coexist, be contextually defined, or be temporarily reversed. When labour is required for a large task, such as hoeing a field or cutting thatching grass, a homestead may organise a work party, putting out the word to members of the sub-ward section that labour is needed, and providing food and drink for the workers. A homestead has to be relatively well off to do this, and some (but by no means all) of the workers who turn up are people from poor homesteads who regularly respond to such calls. For this they receive food and drink as well as the patronage of these larger, wealthier homesteads that often host work parties. However, on such occasions the workers have high status. If beer has been brewed they are in charge of it, allocating it to those who have come from other sub-wards simply to drink. The principle of inter-homestead cooperation and equality within the section is signalled by the fact that members of the section who have not responded to the call for assistance are not permitted to come along only to drink the beer. In addition, the workers are not ‘paid’ with the food and drink, or with the prestige of being able to allocate beer to others. While this is regarded as a just reward for their participation, it is generally accepted that their work will be reciprocated in kind by the receiving homestead when they in turn host work parties.

The following account shows how inclusion and exclusion are linked to individual and communal survival.

M is a relatively well-off man of about 55 years of age who was a labour migrant some 30 years ago (the author has known him all this time) and who depends for economic survival largely on his skill and hard work on the land, his interpersonal skills and his genial personality. Over the period of the author's association with M, his cooperative relations with other homesteads have varied according to their changing circumstances and his, and been affected by unavoidable factors such as ageing and death. During this period some homesteads have been drawn into closer relationship with him, while others have become distanced or disappeared altogether. In yet others, inclusion was followed by exclusion and marginalisation, followed by gradual reinclusion. The case of his brother K is an example. In the late 1970s, K was an oscillating labour migrant,1

working closely with M when he was at home between work spells, and depending on M to care for his homestead while he was away. At these times, M ensured that K's arable land was ploughed and planted, and generally kept an eye on the homestead, which was occupied by K's wife N and her children. M and K belonged to an agricultural company along with another reasonably well-off homestead headed by Q (no relation to the other members) and which also included the homestead headed by N's widowed mother, who lived nearby.

After a number of years of oscillation, K found a job in nearby Umtata (now Mthatha), which did not require him to renew his contract and spend periods at home every year. (Because the job was in an ‘independent’ Transkei, the law obliging migrants to return home at regular intervals did not apply.) His wife and children joined him there. He and his wife did not visit home for many years, and he did not send any money or goods back to M. His homestead fell into disrepair, and his lands became overgrown with weeds. M did not feel obliged to maintain K's homestead despite their close kinship, since K neither returned home periodically nor made any material contribution to it or to his rural kin, and M was unsure whether K would ever return. The small agricultural company became unviable without K and N and so broke up, both Q and M seeking membership of other companies. N's mother was forced to rely on assistance from her sister, also a widow, and her sister's son and his wife, who occupied a nearby homestead.

Some years later, K fell ill and died in Umtata, and his wife was forced to return home with her young children. At this point she and her children were destitute, depending on handouts from her mother, who was now receiving a pension, and occasional assistance from M. No longer in a position of voluntary exclusion, she was marginalised because people were unsure of her commitment to the local community. It was a number of years before she was able to recover economically and demonstrate that she was home for good and that her homestead would survive, and she gradually became more fully reincorporated into the sub-ward section. By 2005 she had re-established the homestead, and M had taken on the responsibility of ploughing and planting her land, in return for which she provided her labour (and that of her children) when M needed it for weeding and other tasks.

In contrast to this process of gradual exclusion followed by gradual reincorporation was the situation of M's father's brother's daughter, Nx, also a widow and the sole occupant of her homestead, which adjoined M's. As Nx's was a separate but related homestead, M took responsibility for cultivating Nx's fields, and for the overall care of her homestead. Nx was (and still is) a regular participant in the work parties of others in the section and sub-ward, and she helped with daily tasks in and around M's homestead. As Nx's children (both girls) grew up, however, and became labour migrants, and Nx herself grew older, she was incorporated more and more into M's homestead, leaving it only to sleep in her own hut every night. Her large garden, which immediately adjoined M's and was regularly cultivated by him, became part of his garden – the two gardens were joined up and in effect became one. The benefits to her were that she was now dependent on M and fully cared for by him as part of his homestead, while the benefits to him were that his arable land had effectively almost doubled in size and he now had, as a member of his expanded homestead, a kinswoman who could provide labour, who qualified for an old age pension, and whose daughters in town regularly remitted money to help support her.

People in Shixini do not discuss issues such as these in the terms used here. They talk of ‘building the homestead’ or of not doing so, of being a migrant who supports his or her rural home (a member of the community – umntu walapha or umntu waselali), or of being one who does not do so (an absconder of no moral worth – itshipha). So how does one demonstrate publicly that one is a member of the local community, a person who subscribes to the rural ethic of building the homestead and to the principles on which local economic endeavour is based – mutual assistance (ukuncedisana) and good neighbourliness (ubumelwane) – since people do not go about discussing themselves or each other in these terms? The answer to this lies in the ritual of beer drinking. It is through beer drinking that the above concerns, principles and moral precepts are symbolically constructed, practically constituted, and expressed.

It is not possible to go into this in any detail here (for a full account, see McAllister, 2005). It is sufficient to say that every Shixini homestead, rich or poor, has an obligation to brew maize beer ‘for people’ at regular intervals, and that this is regarded as a social, moral and religious act essential for ‘building’ the homestead. At beer-drinking rituals, the everyday interdependence of neighbours, and the importance of the section and the sub-ward as the context for economic cooperation, are symbolised and constituted by the structure of the event, primarily in the allocation of seating places and beer. At beer-drinking rituals, the important structural principles of the society, such as age, gender, political status and kinship, become manifest in a variety of ways. But context is important too, as judgements are made and allocations modified according to the prevailing social circumstances and current events. And in the actual consumption of beer, the principles of egalitarianism, sharing, and caring for all members of the community are acted out. To brew beer and to attend beer drinks is to act out one's membership of the moral community, to commit oneself to it, and to accept the principles on which the community constitutes itself.

This, according to Rappaport (1999), is a feature of all ritual. For Rappaport, ritual is metacommunicative and metaperformative, and through it moral obligations are both forged and enacted. In enacting a state of affairs by means of its liturgy, ritual brings about this state of affairs and invests it with moral (and supernatural) authority, establishing a commitment to this state of affairs and to the social order which it articulates among the participants. Xhosa beer drinks, it may be concluded, ‘bring into being … a state of mutual cooperation and interdependence, establishing the morality of such cooperation and inculcating acceptance of it’ (McAllister, 2005:307). To perform a ‘liturgical order’ such as a beer-drinking ritual is to accept its moral legitimacy and the obligation to conform to it. Beer drinks do not simply express or model certain norms and values; they establish their existence and the necessity of living by them. This is why some people who do not drink beer, such as members of the local Zionist church, nevertheless brew for beer-drinking rituals and attend those held by others.

It follows from this that people who do not brew exclude themselves from the community and are marginalised as a result, and there are a few homesteads like this in the sub-ward where the fieldwork for this paper was carried out. Most of them are homesteads that have effectively become de-agrarianised, sometimes in conjunction with conversion to some alcohol-forbidding church and/or more or less complete dependence on income from a small business, cash remittances from town, or social pensions. The number of such homesteads will probably grow in the future. One or two do not brew simply because they are headed by people who are antisocial and withdrawn, or whose social status (e.g. as a diviner, in one case) makes them exceptional. There are no homesteads that are unable to brew beer due to poverty, since it is always possible to beg or borrow the maize required for beer if necessary, and poor homesteads are acutely aware of the economic need to maintain reciprocal ties with those who are better off. Brewing is one of the important ways they are able to do so; it is an investment in social capital.

3. THEORISING INCLUSION AND EXCLUSION

Bourdieu's (1977, 1988) work on the nature of capital and the conversions between different kinds of capital is different from the bulk of the ‘social capital’ literature (Portes, 1998) and somewhat more useful for making sense of the above account of the beer-drinking ritual. Bourdieu's analysis also dovetails with Scott's observations on the nature of peasant morality, discussed above, as well as with Rappaport's understanding of ritual. In the ethnographic description presented above there is evidence of what Bourdieu calls a ‘field’, in this case a rural field (which of course intersects with an urban–industrial one). Within fields there are different forms of capital, which are also forms of power (Swartz, 1997). Material (or economic) capital in Shixini consists of things such as land and other material resources, as well as cash. Inter-homestead economic alliances and cooperation involves exchanges of labour for labour, and of labour for material goods (food and drink, etc.) aimed at enabling homesteads to produce the material means of subsistence. In so far as participation in cooperative social relationships of this kind is socially approved as a moral norm, involves social networks, and relies on reciprocity, it also involves the accumulation and exchange of social and symbolic capital. In this sense the notion of a moral economy is useful. Failure to make exchanges of capital of this kind, either material or social, indicates unwillingness or inability to engage with other members of the moral community, and results in a degree of marginalisation. Inclusion does not depend on whether a homestead is well off or relatively poor; as long as it has at least some resources with which to meet its social obligations it will be accepted into the community.

Capital is a resource but also a form of power, and those who accumulate gain power themselves – in this case economic and social power – which can be converted into political power. Thus a poor homestead that receives assistance from a wealthier one and cannot reciprocate economically can do so by lending political support to the giver, thus converting the giver's outlay of material capital into a more powerful form – symbolic capital. A significant aspect of such political support and the symbolic capital that accrues from it is the beer drink ‘for the harvest’ that the poorer homestead holds, and at which the contribution of the wealthier homestead (as leader of the ploughing company) is publicly recognised and lauded. The term ‘symbolic capital’ refers to the authority and legitimacy conferred on and secured by people through their relations with others within the moral community.

Beer drinking involves the conversion of material capital (including material things such as the homestead, and maize) into cultural and social assets, thus into social and cultural capital. In this conversion they are turned into symbolic capital, being used as authorising and legitimising agents. Since symbolic capital is more powerful than material capital, the legitimacy of the homestead's economic efforts and its continued place within the relations of production in the section comes to depend on such conversion being made to demonstrate its participation in and acceptance of the moral community – the section and sub-ward, and the kin group. Bourdieu's concept of legitimacy here is virtually identical to the notions of moral authority (as implicit in Scott, 1976) and moral legitimacy (in Rappaport's sense, 1999).

4. WHAT ARE THE IMPLICATIONS FOR POLICY?

These concluding thoughts are fairly speculative, but they follow from the analysis presented above. Firstly, developing or improving people's standards of living should not be seen in material terms alone. A material benefit that comes at the cost of cultural and social capital may not be acceptable to people if it endangers the networks of social relations in the moral community. The corollary to this is that enabling people to participate more fully in the moral community – for example, by enabling them to participate more fully in exchange networks, or facilitating increases in their social and cultural capital in other ways – may enable them to become materially better off. It is also possible, once one is aware of cooperative arrangements such as the agricultural companies discussed above, that such arrangements can be used to foster rural development, perhaps as an alternative to the ‘small farmer’ approach currently favoured. Community resistance to development initiatives taken without adequate knowledge of local social arrangements may also potentially be explained in terms of the risks to productive activities undertaken by groups such as agricultural companies.

Perry and Ward (2001) point out that rural and agricultural development projects rarely recognise local institutional arrangements and that the environmental costs of this have been significant. On the other hand, where local institutional capacity is harnessed, the long-term sustainability of development projects is enhanced. In so far as development and the eradication of poverty frequently involve people working together in groups, or collaborating with each other in constructive ways, it is important, in view of the above arguments, to identify the nature and sources of economic and other forms of capital, and the ways in which one kind of capital may be transformed into another. Such knowledge, which may be generalisable over large areas (as in parts of rural Transkei) or limited to small groups of a certain type (such as urban based rotating credit associations), can be of vital importance in ensuring that development or poverty eradication initiatives take into account established forms of capital exchange and do not disrupt such exchanges, leading to the possible marginalisation of individuals or groups within the community. Similarly, attempts to raise living standards that involve people working together need to take into account the structure of social relations within a community, including the extent and nature of marginalisation, if any. Here it is necessary to reiterate a point made at the start of this paper, namely that marginalisation needs to be studied empirically and not assumed to exist or to coincide with structural factors such as gender and age.

The author would like to thank an anonymous reviewer who made a number of useful suggestions that have been incorporated into this paper.

Notes

¹Apartheid legislation compelled migrant workers to return to their rural homes between labour contracts, and thus to ‘oscillate’ between country home and urban–industrial workplace.