From people's politics to state politics: aspects of national liberation in South Africa 1984-1994


This article, written in 1994, the first year of the 'new South Africa', examines the shift from popular people's politics to state politics during the transition from apartheid.

Submitted by red jack on February 9, 2012

The 1980s in South Africa witnessed an explosion of popular-democratic struggles championed by a host of autonomous civil society organisations whose activities became central in the campaign against the apartheid system and the quest for the creation of a democratic state. By the 1990s, however, especially in the lead up to and after the election of the ANC into office, South African politics appears increasingly to be subjected to the same broad logic of statism that was experienced in other parts of Africa at the dawn of independence from colonial rule. At the heart of this statism is the defeat of the popular movements which, in South Africa, as in the rest of Africa, were so vital in the struggle for political change, being the oppositional forces with the popular democratic tradition and agenda that offered the best chance for the emergence of a democratic state. Increasingly depoliticised, the role of the popular movements has been emptied of the vitality that can ensure that `the people' are able to generate and make autonomous democratic prescriptions on the state. The politicisation of civil society and the democratisation of the state are projects in the South African transition which will have to be revived if the authoritarianism that inheres in statism is to be defeated. This is the challenge before the democratic forces of opposition and change in contemporary South Africa.

`Al die mamas en die papas, die boeties en die sussies, die oumas en die oupas, die hondjies en die katjies -- almaal is saam in die struggle'

(popular slogan/song from Cape Town, 1985). (All the mothers and the fathers, the brothers and the sisters, the grandpas and the grandmas, the dogs and the cats -- all are together in the struggle.)

`Generally ... I can say that the community is the main source of power, because the state has really lost the control over the people. He (sic) has no power over the people in terms of controlling them. This is why the people have formed these area committees, so that they can try to control themselves. What has been preached in the past about the Freedom Charter, even now we are trying to do that practically'

(An activist from the Eastern Cape, Isizwe, 1(2), March 1986).

`There is a remarkably strong corporatist current flowing in South Africa. The major actors -- labour, capital and the state -- are so caught up in it that they are hardly aware of the fact that they have become part of the current'

(J. Maree, 1993:24).

`South Africa's tradition of a strong and vibrant civil society needs to be reasserted. We must not replace apartheid statism, and top-down rule, with a new form of statism'

(Sam Shilowa, General Secretary of Cosatu 1995:27).


In early May 1994, less than a week after the first government of a liberated South Africa had been elected, the new Foreign Minister, Alfred Nzo, gave his first speech in his new capacity to a meeting of the Organisation of African Unity (OAU). In the speech he, inter alia, made the point that South Africa had finally won the battle for political freedom, and that it now had to address the battle for the development of its hitherto oppressed majority. In uttering these words, the Minister was truly expressing South Africa's final arrival into the community of African post-colonial states as, since Nkrumah, practically every African leader had reiterated the same slogan at independence. It was this slogan which largely encapsulated `developmentalism' and the nation-state project in Africa, namely, the end of popular politics and the beginning of state mobilisation for bureaucratic `development', with its attendant emphasis on 'nation building' and `Africanisation' (now known in South Africa as `affirmative action'). While 'nation building' in Africa was reduced to state building and the systematic suppression of regional and ethnic grievances under an obsessive concern to retain colonially inherited boundaries, `Africanisation' was the way in which the creation of a 'national' petty-bourgeoisie could be undertaken by providing access to jobs in the expanding `public' sector.

These jobs were ostensibly to be reserved for nationals, but often ended up being provided simply to members of the state-controlling nationality, thus undermining any 'national' pretensions of a newly forming class. In addition, of course, access to state jobs for this petty-bourgeoisie meant access to state resources, which could then be utilised for private appropriation rather than accumulation. Hence economic appropriation by the new ruling groups largely went hand in hand with the reproduction of `ethnic' or regional divisions as these largely constituted the main routes to such appropriation. This process of state formation was related, therefore, to the well-known absence of a clear separation between politics and economics in Africa -- between political power and accumulation -- which itself provided the conditions for authoritarianism and statism. It has now clearly been recognised (largely as a result of the crisis of legitimacy which has become more apparent in the SAP and post-SAP periods) that this state project in Africa has reached a point of profound crisis from which some even argue, the state seems incapable of extricating itself (Nzongola-Ntalaja 1993; Wamba-dia-Wamba 1993, 1994; Olukoshi & Laakso 1996).

Despite the currency of this debate, there is evidence that a consensus on some fundamental points is emerging, namely that this crisis is the result of the entrenched authoritarianism, more simply the overall statism of the African state since independence, that to a great extent this statism was built on the coercive practices taken over from the colonial period, and that the re-emergence of ethnicity as a central issue in Africa is intimately related to the failure of statism. At the same time, it is also reasonably apparent that statism prevailed relatively independently of whether state politics were based on a no-party, a one-party or a multi-party system, as in the words of Wamba-dia-Wamba (1993:103), the African state was, `in the `decolonization process', grafted onto a colonial, essentially undemocratic (variant of apartheid) state'. But statism was not an inevitable outcome of independence. Rather, its features, it is argued, have their origins in what Mamdani (1990) has called the `defeat of popular movements' in the period of transition from colony to independent state.

The more or less protracted struggles for independence in Africa had involved large numbers of people (overwhelmingly, but not exclusively, from the most oppressed and exploited strata) in a process of struggle over a more or less long period of time. These processes had been ones in which a whole variety of organisations (parties, trade unions, women's organisations, youth organisations, peasant cooperatives, small business organisations and so on) expressed the aspirations of the majority of the hitherto oppressed population. Although they were not always clearly distinguishable (nor did they always wish to distinguish themselves) from a broader 'national movement' during the anti-colonial struggle, these organisations were rarely simple component parts of the political party which was eventually to achieve state power. They arguably always had a significantly independent existence and often an independent support base as well as distinct, if not always explicitly articulated, demands. Neither were popular movements simply stages or steps on the road to the realisation of a nationalist ethos. The popular movements of which such varied organisations were the complex expression had often distinct forms of organisation and demands, which were not always simply reducible to the common denominator of political independence. They often involved (explicitly or implicitly) demands for social transformation, popular forms of democracy and equality, which went beyond the slogans of the political party which was to emerge victorious at independence (Kriger 1992; Mamdani 1990; Gibbon 1994).

In Africa, it has generally been the case that, as the independence process proceeded, these popular and politicised `civil society' organisations, were gradually replaced by state organisations in a number of ways (incorporation, cooption, disbanding (voluntary or enforced), bribing of leaders and general destruction). In brief, these organisations of civil society were usually eliminated or transformed and replaced by state (and/or party) institutions as part of a gradually developing nation-state project. Although these defeats were in no way inevitable, it is this general process which arguably lay at the inception of the statism which has been dominant in Africa since independence. The consolidation of this form of state power was largely secured in the run-up to independence, as the colonial state and its chosen successors collaborated in structuring the new state (Mamdani op cit; Gibbon op cit).

Part of this process was one in which such popular organisations, which during the struggle for independence, had entered the realm of political society (by addressing explicitly political questions, through interpellating their supporters as citizens) were gradually restricted to the realm of civil society (where they were to address the narrower concerns of `interest groups' defined by the division of labour) often before losing their independence altogether. The process of restricting popular organisations from a role in political society to one in civil society is, arguably, part of a general process of demobilisation to have been experienced by popular organisations in the period immediately preceding and following independence/liberation (Gibbon 1993). More precisely this process amounts to one of exclusion of popular organisations from the plane of territorial politics where 'national' issues are addressed exclusively by the colonial state and its chosen interlocutors, usually in the form of the incoming nationalist party itself already organised along state lines (for example with various departments corresponding to ministries, foreign representatives, etc) (Gibbon 1994).

The argument here is that the transition to liberation (the equivalent of independence) in South Africa can only be understood in this African historical context. Indeed, it may be suggested that despite many important differences, the process of transition in South Africa exhibits fundamental elements which parallel the history briefly outlined above. While it was thought by many -- including the present author -- that the popular movements which had developed in the 1980s in that country were strong enough to provide a counterweight to statist tendencies, the 1990s have witnessed a systematic (and astonishingly rapid) process of political demobilisation in the country. This has reached an extreme situation whereby the urbanised population, which had been massively involved in political activity in the 1980s as all aspects of life were politicised, was in 1995 showing signs of extreme apathy towards the local elections (Mail and Guardian, 27 October 1995).

In the 1980s South Africa had the unique experience of a mass popular movement, which was inaugurated in 1976, and which demarcated it sharply from the experience of the rest of Africa in the 1960s. This largely urban movement differed from the African norm, not simply because of its spatial uniqueness, but much more importantly in its involvement of large numbers of ordinary people in a mosaic of political organisations of an enormous variety and in independent and coherent nationwide political activity, which transformed people's lives and often set up democratic popular political structures independent of (and directed against) the state (Marx 1992; Murray, 1987; Lodge et al 1991). This process was arguably typified by practices (and not just slogans) of `People's Power', particularly associated with the United Democratic Front (UDF), along with the establishment of what has been called `Social Movement Unionism', that is the involvement of trade unions in mass national politics (Webster 1987). Along with these went practices and debates on the nature of 'non-racialism' and the future character of a non-racial and not just a `multi-racial' state. While such practices were undoubtedly contradictory and should not be idealised, they differed starkly from the present context (which has its origins in the early 1990s), in which 'nation building', `affirmative action' and elite-based `reconciliation', all organised and led by the state, are more typical of a now `official' practice and discourse of `liberation'.

The popular experience of liberation politics in the South Africa of the 1980s was a far cry from that of the largely elitist guerilla army in the bush, more often trampling on water rather than swimming like a fish in it, or for that matter from the equally elitist, teacher-dominated organisation having independence thrust upon it by a departing colonial power, which was so typical of the experience of other African countries. And yet, interestingly, the period 1990--94 was in South Africa largely also characterised by a process of popular demobilisation as an elitist deal was struck behind closed doors by an outgoing National Party and an incoming ANC in their interests (the story of which is told, with some accuracy, in Sparks 1995) and arguably in those of a 'new' South African bourgeoisie, while in the townships whatever remaining popular structures were largely destroyed through systematically organised state-inspired violence. The objective of this essay is to trace the main features of this process of transformation of the politics of liberation from a situation where dominant popular politics (with all their attendant contradictions) were substituted by state politics, that is a process whereby the centre of gravity of 'national politics' was more or less gradually, more or less rapidly, moved from the people to the state.

The argument will follow a structure whereby initially some of the main political and ideological features of the popular struggles of the 1980s and their alteration into different forms of politics in the 1990s will be outlined; this will be followed by a brief assessment of current interpretations of this period in the literature, and an outline of a theoretical perspective which encompasses recent debates on the African experience.


It is not intended in this section to provide a history, however brief, of popular struggles in the South Africa of the 1980s as such histories can be found elsewhere (Marx op cit; Lodge et al op cit; Murray op cit). Rather, after some brief comments on its origins, illustrations of the most characteristic aspects of popular politics in the 1980s in both township-based and trade union movements will be given and contrasted to the liberation politics of the 1990s. Throughout, the emphasis will be placed on urban movements, the dominant source of protest during the period.

Struggles in the townships

The origins of the mass urban protests in the 1980s are usually traced to the student upsurge of 1976 in Soweto. It was these youthful struggles which forged, in Mamdani's terms, `a new path of liberation' which was based on the lived experiences of ordinary people rather than on the failed sterility of the strategies of exiled movements cut off from the people on whose behalf they were supposed to be struggling (Mamdani 1994, chapter 6:33). The so-called `decade of peace' after Sharpville was a testament to the overall weakness of these exiled organisations. The 1976 Soweto uprising, along with the series of mass strikes in Durban three years earlier, shattered this `phoney peace'.

In fact, in structural terms, it was effectively this period of extreme repression which was to provide, through exceptional economic growth, the seeds of the destruction of the apartheid state. In the 1960s, South Africa's GNP grew at 6 per cent per annum and South Africa was at the time, along with Japan, the country with the highest growth rate in the world (Baskin 1991:17). This had a number of important consequences including a dramatic increase in the number of Africans working in manufacturing (from 308 332 in 1960 to 780 914 in 1980), denoting, an increase not only in the industrial working class but in the number of skilled African workers. At the same time, the South African economy became more dependent on consumption by blacks for its internal market (Marx op cit: 193). Another effect was an increase in black South Africans entering the education system to provide for this increased industrialisation. Between 1965 and 1975, the number of black students in secondary schools increased nearly fivefold, while between 1980 and 1984, their enrolment doubled from 577 000 to over a million; and the number of graduates tripled during the same period (Lodge et al op cit:30--31).

By 1982, however, not only were the effects of the world economic crisis being felt in South Africa, but a fall in the price of gold, which lasted until 1985, along with a balance of payments deficit created by the importation of capital equipment for this mini-import substitution industrialisation process, led to `an unprecedented level of foreign indebtedness' (Lodge et al ibid). As a result of IMF loan conditionalities, the government scrapped whatever subsidies to consumers were in place and increased sales tax, which shifted the fiscal burden to the poor. By 1985 the inflation rate was just below a record 17 per cent. All this led to a steep increase in the unemployment rate beginning in 1982 and accelerating thereafter:

By 1985, African unemployment represented about 25 per cent of the economically active population. Two-thirds of all unemployed Africans by the mid-1980s were under the age of thirty, and unemployment was often long term especially for school leavers (ibid).

A prolonged drought increased the price of food and also the level of rural-urban migration, while cutbacks in state expenditure affected townships in particular, so that local authorities administering them had accumulated a deficit of R32 million by 1982/83. From 1981, township residents were subjected to rent hikes which increased in frequency after municipal elections in 1983. At the same time, reviled township councillors were given more powers in 1982 (through the Black Local Authorities Act) increasing their control over the allocation of housing, trading licences, business sites, student bursaries and the collection of rents (ibid). They were obviously seen as benefiting from `the system', as the apartheid state structures came to be known among activists. The state attempted to manage the growing discontent by legal reforms which attempted to regulate union activity and restrict it to the workplace (primarily through the Industrial Relations Act of 1979) and to co-opt Indian and coloured South Africans into a tricameral parliament while power rested firmly with whites (the new constitution was inaugurated in 1984). It was these structural changes which formed the background to the mass upsurge of the second half of the 1980s.

The most important and truly original organisational expression of popular resistance in the 1980s was the United Democratic Front (UDF), which was formed in 1983 initially ostensibly to mobilise opposition to the state's constitutional proposals and other legislation (known collectively as the Koornhof Bills), including the Black Local Authorities Act. The UDF brought together under its umbrella a coalition of civic associations, student organisations and youth congresses, women's groups, trade unions, church societies, sports clubs and a multitude of organisations which retained, and often increased as a result of their affiliation to the UDF, their ability to organise independently. At its peak it claimed it had around 700 affiliates grouped in ten regional areas and amounting to a total of over two million people (ibid:34). With the upsurge of township unrest beginning in earnest in 1984, it was the young people of the townships who provided the main impetus behind the struggle, while the leadership passed over to the Trade Unions in 1988. In one important respect at least, the UDF managed to build on the experience of township-based organisations such as `civic associations', in that it successfully combined local and national grievances. In the words of one civic activist and intellectual:

From the late 1970s, civic associations not only opposed community councils, they challenged the very laws upon which such bodies were founded. For example, in 1979 the Port Elizabeth Black Civic Organisation (PEBCO), (now the PE People's Civic Organisation), called for a single municipality for the city of Port Elizabeth and rejected the community councils (in charge of African townships) and the white municipalities (in charge of white local affairs). PEBCO's aims included a commitment to fight discriminatory legislation, to seek participation in all decision-making processes, to fight for African freehold rights and to resist attempts to deprive Africans of their citizenship. Thus it can be seen that from their inception, civic associations tackled both local problems and issues with national political implications. In due course, local demands assumed a national dimension (Botha 1992:63).

It is not possible to undertake a detailed history of the UDF here due to considerations of space; fortunately this can be found elsewhere (for example Swilling 1988; Lodge 1991:23--141; Marx 1992). Nevertheless, it is important to point out some of the phases which this organisation went through, as they provide an accurate reflection of the changes in urban-based forms of struggle and popular involvement during the period of the UDF's existence until 1990 when it was disbanded. This analysis largely follows those of Swilling (op cit) and Lodge (op cit). The first phase of the UDF followed its activity to oppose elections to the tricameral parliament and the Koornhof Bills. But soon after August 1984, opposition political activity shifted to struggles initiated by local communities and became concerned with basic issues affecting township life. This inaugurated its second phase. The mass upsurge started in earnest in September 1984 and took the form of bus and rent boycotts, housing movements, squatter revolts, labour strikes, school protests and community stay-aways. This change in the focus of protest was not the result of any strategy by the leadership of the UDF or of a change in policy. It seems ultimately to have been forced on the leadership from below (Swilling 1988:101). Indeed, by mid-1985 it was becoming clear that the UDF leadership was unable to exert effective control over developments despite its popularity. In Lodge's words:

The momentum for action came from the bottom levels of the organisation and from its youngest members. It was children who built the roadblocks, children who led the crowds to the administrative buildings, children who delegated spokespersons, and children who in 1984 told the older folk that things would be different, that people would not run away as they had in 1960 (Lodge et al 1991:76).

According to Swilling, local organisations:

exploited the contradiction between the state's attempts to improve urban living conditions and the fiscal bankruptcy and political illegitimacy of local government. They managed to ride a wave of anger and protest that transformed political relations in the communities so rapidly that the UDF's local, regional and national leaders found themselves unable to build organisational structures to keep pace with these levels of mobilisation and politicisation (ibid:101--2).

He also stresses that mass actions mobilised unprecedented numbers of people. These succeeded in mobilising:

all sectors of the township population including both youth and older residents; they involved coordinated action between trade unions and political organisations; they were called in support of demands that challenged the coercive urban and education policies of the apartheid state; and they gave rise to ungovernable areas as state authority collapsed in many townships in the wake of the resignation of mayors and councillors who had been `elected' onto the new Black Local Authorities (ibid:102).

The third phase of UDF activity was inaugurated by the declaration of the first state of emergency in 1985 (and lasted until 1986) as the apartheid state attempted to control this mass upsurge and reassert control over `ungovernable areas'. Interestingly, both popular rebellion and political organisation grew during this period, which saw the setting up of `street committees' in particular. These took over the functions of local government, especially in ungovernable areas. One local activist in the Port Elizabeth area stated:

We said [to our people]: In the streets where you live you must decide what issues affect your lives and bring up issues you want your organisation to take up. We are not in a position to remove debris, remove buckets, clean the streets and so on. But the organisation must deal with these matters through street committees (cit Lodge et al op cit:82).

The ANC view as expressed by their spokesman, Tom Sebina, was that street committees `grow out of the need of the people to defend themselves against State repression ... and in response to ANC calls to make the country ungovernable and apartheid unworkable [so as to forge them into] contingents that will be part of the process towards a total people's war'. Contrary to this view, which saw street committees as tactical adjuncts to the development of a militaristic process and as simply `oppositional' to the apartheid state, local activists spelled out a different assessment:

The people in Lusaka can say what they like ... we know that the purpose is to enable people to take their lives in hand. Local government has collapsed. The state's version of local government was corrupt and inefficient in any case, but local government is necessary for people to channel their grievances. The street committees fill the vacuum. They give people an avenue to express views and come up with solutions (cit: Mathiane 1986:13).

In many townships, rudimentary services began to be provided by civics and youth congresses, while crime also began to be regulated through `people's courts'. These originally developed in some areas to regulate dispute between neighbours (as in Atteridgeville in Pretoria) and also as attempts to control the proliferation of brutal Kangaroo courts (for example in Uitenhage and Port Elizabeth). In Alexandra, outside Johannesburg, five members of the Alexandra Action Committee were nominated in February 1986 to sit in judgement over cases of assault and theft, while street committees were empowered to settle quarrels. In Mamelodi, one of Pretoria's townships, a number of `informal' systems of justice operated in the 1970s and 1980s and there were long-term struggles over the setting up of popularly accountable courts, which were also highly influenced by traditional African custom (for example the importance of elders, etc). Lodge concludes that:

Of all the manifestations of people's power ... the efforts of local groups to administer civil and criminal justice were the most challenging to the state's moral authority. More than any other feature of the insurrectionary movement, people's justice testified to the movement's ideological complexity and to the extent to which it was shaped from below by popular culture (ibid:135).

In addition to popular control of townships and popular justice, there was a complementary development of institutions geared towards the provision of `people's education'. These included, in particular, attempts to bring local schools under community control through the establishment of parent-teacher-student associations (PTSAs) and even attempts to develop a new curriculum in response to `Bantu Education', the central plank of the apartheid state in this sphere. The struggle for people's education was seen as intimately linked to establishing `people's power'. In the words of Zwelakhe Sisulu:

The struggle for People's Education is no longer a struggle of the students alone. It has become a struggle of the whole community with the involvement of all sections of the community. This is not something which has happened in the school sphere alone; it reflects a new level of development in the struggle as a whole ... The struggle for people's education can only finally be won when we have won the struggle for people's power ... We are no longer demanding the same education as Whites, since this is education for domination. People's education means education at the service of the people as a whole, education that liberates, education that puts the people in command of their lives. We are not prepared to accept any `alternative' to Bantu Education which is imposed on the people from above. This includes American or other imperialist alternatives designed to safeguard their selfish interests in the country ... To be acceptable, every initiative must come from the people themselves, must be accountable to the people and must advance the broad mass of students, not just a select few (Sisulu 1986:106, 110).

Or again:

I want to emphasise here that these advances were only possible because of the development of democratic organs, or committees, of people's power. Our people set up bodies which were controlled by, and accountable to, the masses of the people in each area. In such areas, the distinction between the people and their organisations disappeared. All the people young and old participated in committees from street level upwards (ibid:104).

However, at the same time as street committees were taking up local `grassroots' issues, they also functioned as vehicles for the direct challenge to apartheid state power by the people. A detailed assessment from 1986 made this point forcefully.

The street/area committees -- the structures of an embryonic People's Power -- are not only restricted to playing this kind of [local -- MN] role, but also have a far more directly or narrowly political dimension to them. At the same time as they are taking up the grassroots issues described above, they also form the units in and through which major political issues and strategies (for example the consumer boycott) are discussed and organised. Thus the street committee system is beginning to form not only the avenue through which people can begin to take greater and more democratic control of the immediate conditions of their existence, but they are also emerging as the form through which direct political action against the state and the ruling bloc can be decided on and implemented (White 1986:92).

Not surprisingly under such conditions, the apartheid state did not hesitate to intensify its repression. The fourth phase of the UDF lasted between 1986 and 1988 and was characterised by the massive repression of the second state of emergency which now covered the whole country. In the first six months of the emergency, around 25 000 people were arrested and isolated, the ability of the press (especially the vibrant `alternative press') to report objectively was systematically curtailed and the townships were placed under direct military rule while the state introduced a militarised bureaucracy (the National Security Management System) to run local government and to `win hearts and minds' (known as WHAM) following the classic counter-insurgency pattern which the Americans had perfected in Vietnam. In brief, this state offensive succeeded in undermining popular organisations considerably, and probably eliminating popular leadership altogether. This was not because the UDF ceased its activities; on the contrary, rent, bus and consumer boycotts continued unabated at least until 1987 (Lodge op cit:87--100). Rather, it was the popular aspect of the struggle which was fatally wounded as it depended for its democratic operation on consultative processes, relative freedom of movement, etc, and there was no army under popular control capable of defending popular gains and structures against military onslaught.

By 1986, a contradiction had emerged between those who wished to retain the broad front structure of the UDF with diverse affiliated organisations, and those who argued for a move to a more centralised party structure; in practice it seems that the latter position was becoming dominant (De Villiers 1986; UDF 1987:18--22). From late 1986 onwards, UDF campaigns more and more were initiated `from above', by the 'national leadership' operating exclusively at the territorial level. At the same time, more and more coercive measures were being applied to township residents to adhere to various boycotts (a fact which shows the weakening of popular control), `the struggle' was acquiring more of a militaristic character, and vigilante activities acquired increasing support from businessmen affected by youth-directed boycotts.

As a result, when resistance resurfaced in the final phase of the UDF, from 1988 to 1990, it became characterised by completely different practices from earlier periods. While the movement (now in alliance with Cosatu calling itself the Mass Democratic Movement and closely linked to the mainstream churches) was able to organise mass campaigns (for example the `defiance campaigns' of 1989) against segregated facilities such as hospitals, these became more and more reminiscent of the American Black Civil Rights Movement of the 1960s. These campaigns were now all organised on a territorial plane so that:

in contrast to the mid-1980s, when the insurrectionary movement was being pulled onto unchartered courses by cadres of youth in the streets of the townships, the popular protest in the late 1980s was choreographed and coordinated and seemed much more under the command of its leaders (Lodge et al op cit:114).

Under such circumstances, it would be relatively easy for leaders to disband the UDF in the wake of the unbanning of the ANC, as it was felt that the latter could now take over the organisation of popular political protest.

Before moving to a brief assessment of the 1990s it is pertinent to make some general points about the ideology and practice of this mass movement of the mid-1980s. What stands out is an attempt to develop genuinely popular forms of democracy in both ideology and practice. In particular, the general characterisation of the mass struggle as national and democratic combined both territorial and popular democratic aspects of the process. In fact, the two were regularly combined in attempts by leading activists to theorise the process of struggle. As Murphy Morobe, the `Acting Publicity Secretary' of the UDF, stated in 1987:

We in the United Democratic Front are engaged in a national democratic struggle. We say we are engaged in a national struggle for two reasons. Firstly, we are involved in political struggle on a national, as opposed to a regional or local level. The national struggle involves all sectors of our people -- workers (whether in the factories, unemployed, migrants or rural poor), youths, students, women and democratic-minded professionals. We also refer to our struggle as national in the sense of seeking to create a new nation out of the historical divisions of apartheid. We also explain the democratic aspect of our struggle in two ways ... Firstly, we say that a democratic South Africa is one of the aims or goals of our struggle. This can be summed up in the principal slogan of the Freedom Charter: `The People Shall Govern'. In the second place, democracy is the means by which we conduct our struggle ... The creation of democratic means is for us as important as having democratic goals as our objective ... When we say that the people shall govern, we mean at all levels and in all spheres, and we demand that there be a real, effective control on a daily basis ... The key to a democratic system lies in being able to say that the people in our country can not only vote for a representative of their choice, but also feel that they have some direct control over where and how they live, eat, sleep, work, how they get to work, how they and their children are educated, what the content of that education is; and that these things are not done for them by the government of the day, but [by] the people themselves ... The rudimentary organs of people's power that have begun to emerge in South Africa (street committees, defence committees, shop-steward structures, student representative councils, parent/teacher/student associations) represent in many ways the beginnings of the kind of democracy that we are striving for ... Without the fullest organisational democracy, we will never be able to achieve conscious, active and unified participation of the majority of the people, and in particular the working class, in our struggle (Morobe 1987:81--83, emphasis added).

I have cited this passage at length because it clearly sums up the systematisation of popular experiences and demands which some leaders were able to eloquently make. Clearly, this statement has more the character of an ideal to be struggled for rather than a simple description of reality; nevertheless it indicates the centrality of popular democracy within the ideology and practice of the movement. It is important to note first (I shall have occasion to return to this below) that the main slogan of the Freedom Charter (`The people shall govern') is given a specific interpretation by the UDF, namely to mean a popular form of democracy and not simply an electoral multiparty system or, for that matter, a one-party system (as its vagueness could also imply). In fact, the former is explicitly rejected as the exclusive form of representation, and as too limited a form of democracy. Thus an evidently vague and indeed `populist' slogan in the circumstances of the time, could be given an unambiguous popular-democratic content. It would be a fundamental error to confuse the content of such democracy with its own slogans and its self-presentation, as many did, who at the time dismissed the UDF as a `populist' organisation. In practice the social movement was giving rise to a form of mass democracy and a form of state unique in South Africa (and probably also in Africa as a whole); these forms of democracy and state have, arguably, gone largely unrecognised by most intellectuals, by the party of state nationalism, the ANC, and even by many of the movement's own leaders.

Two features of this democracy worth noting were a detailed system of controlling leaders to be accountable to the rank and file membership, and a different way of demarcating `the people' from `the oppressors'. Attempts at instituting internal democracy within organisations were strongly followed, although they obviously had various degrees of success. The important point, however, was that there was such a struggle for democracy in organisations. The various dimensions of this democracy were according to Morobe:

1. Elected Leadership. Leadership of our organisations must be elected (at all levels), and elections must be held at periodic intervals ... Elected leadership must also be recallable before the end of their (sic) term of office if there is indiscipline or misconduct.
2. Collective Leadership. We try and practice collective leadership at all levels. There must be continuous, ongoing consultation ...
3. Mandates and Accountability. Our leaders and delegates are not free-floating individuals. They always have to operate within the delegated mandates of their positions and delegated duties ...
4. Reporting. Reporting back to organisations, areas, units, etc. is an important dimension of democracy ... We feel very strongly that information is a form of power, and that if it is not shared, it undermines the democratic process. We therefore take care to ensure that language translations occur if necessary ...
5. Criticism and Self-criticism. We do not believe that any of our members are beyond criticism; neither are organisations and strategies beyond reproach ... (Morobe op cit:84--85).

However, by (February) 1989, it had become clear that some individuals were beyond criticism, as when an attempt was made by the UDF (and Cosatu) to publicly censure Winnie Mandela (by a committee including Murphy Morobe), it was blocked by the ANC in Lusaka. In fact, many were aware of the danger posed to popular democracy by the lack of control of the popular movement over a number of `charismatic' leaders who felt they had the authority to speak and act without being mandated. Thus, Isizwe, the main journal of the UDF made a rather prophetic statement in 1985:

One thing that we must be careful about ... is that our organisations do not become too closely associated with individuals, that we do not allow the development of personality cults. We need to understand why we regard people as leaders and to articulate these reasons. Where people do not measure up to these standards they must be brought to heel -- no matter how `charismatic' they may be. No person is a leader in a democratic struggle such as ours simply because he or she makes good speeches ... No individual may make proposals on the people's behalf -- unless mandated by them ... We need to say these things because there are some people and interests who are trying to project individuals as substitutes for political movements (United Democratic Front 1985:17, emphasis added).

The practices of `mandates and report backs', which had been adopted largely as a result of trade union influence, were taken particularly seriously in the mid-1980s, although there is evidence that they started to decline at the end of the decade. By 1991, the position had changed substantially so that Mayibuye, the journal of the ANC, now pompously proclaimed:

Accountability is the basis of democratic organisation. Accountability means that leadership must discuss decisions with the membership. Decisions must be explained so that members understand why they are made (Mayibuye December 1991:36).

We are a far cry here from `People's Power'. The manner in which the popular movement demarcated its members (`the people') from the oppressive state is also worthy of note. This largely surrounded the notion of 'non-racialism' as a way of characterising the ideology of the movement as well as the nature of the state which was being fought for. Originally inherited from Black Consciousness discourse, which used the term to refer to all oppressed racial groups in South Africa under the characterisation `black', 'non-racialism' was adapted by the UDF to include whites who supported the struggle. This struggle was visualised as uniting into a national opposition the disparate groups which the apartheid state divided, hence the main slogan of the UDF: `UDF unites, apartheid divides!'. One important aspect of non-racialism was that rather than distinguishing `the people' or `the oppressors' on racial grounds, it did so by demarcating on political grounds: popular-democrats from anti-democrats. The former were those who supported change `from below', the latter those who proposed some form of `tinkering from above', and who had by this period lost the confidence of the majority. Democrats were all those who opposed `minority rule' and supported `majority rule' through popular democracy. In the words of a UDF discussion document from 1986:

The essential dividing line that we should promote is between supporters of minority rule and majority rule. The common ground between the Botha (sic), the PFP (Popular Federal Party, the main White, big business-backed liberal opposition at the time -- MN) leadership and big business is that they all seek solutions within the framework of adapting minority rule. Although they differ fundamentally on who to involve in negotiation and how much adaptation is necessary, these elements all agree that the system must be changed from the top down, with the solutions being decided over the heads of the people. All those who accept the right of the people to determine the process of change are allies of the people and part of the NDS (National Democratic Struggle -- MN) (UDF Cape Town Area Committee 1986:10).

This meant that the conducting of the popular struggle should also be 'non-racial'. Terror Lekota, a senior UDF figure, put it this way:

In political struggle ... the means must always be the same as the ends ... How can one expect a racialistic movement to imbue our society with a non-racial character on the dawn of our freedom day? A political movement cannot bequeath to society a characteristic it does not itself possess. To expect it to do so is like asking a heathen to convert a person to Christianity. The principles of that religion are unknown to the heathen let alone the practice (cited Marx op cit:124).

Such a position was possible precisely because the social movement was not an elite movement and because white `progressives' (to use the jargon of the time) provided invaluable work both in the trade unions and in the UDF, thus becoming known and appreciated by the people of the townships. It served to divide a minority of white democrats from white racists (while forcing the uncommitted to commit themselves), in the same way as affiliation to popular organisations divided blacks between collaborators with the state (so-called `sell-outs') and the majority of the oppressed. This attempt to create the unity of a 'new nation' can be contrasted with the attempts, in the 1990s, to do so `from above' via `reconciliation', 'nation building', the Reconstruction and Development Programme (RDP) or indeed `affirmative action'.

In turning to a brief examination of the period after the unbanning of the ANC and other proscribed organisations up to the national elections of April 1994, my intention is simply to make the point that the politics of liberation have been conducted in a markedly different way, which I would describe as `state centred' rather than `people centred'. I do not wish here to attempt a detailed explanation for such a change as not only are considerations of space prohibitive, but this is intended to be the subject of future research. Two examples will help to illustrate these changes: the altered role of `civics' and some of the views surrounding this changed role; the changed role of mass mobilisation at least up to the end of 1992, after which it largely ceased to exist altogether.

It is very instructive to note the path taken by civic organisations. Perhaps the most important step on this path was the setting up of the South African National Civic Organisation (SANCO) in 1992. As its name indicates, this was set up explicitly as an organisation designed to operate at a territorial level while its member-organisations transformed themselves from being autonomous affiliates of an umbrella organisation (such as the UDF) basically into branches of a national body. In the words of one critical activist: `It requires that local civics surrender their autonomy and local accountability' (Jacobs 1992:24). The preamble to Sanco's constitution defines it as a body that will `act as a non-partisan democratic watchdog of the community on local government and community development'. While not all civics joined Sanco, the overwhelming majority have. Thus the leadership of the `civic movement' no longer see it as a political mass movement or a form of state (`people's power') but as a `watchdog', that is an `interest group' reflecting the aspirations of a narrow constituency defined by the division of labour (urban communities).

The first point to make in this context is that civics -- as indeed all other popular organisations -- systematically surrendered the plane of territorial politics to the ANC at its unbanning. This surrender was expressed ideolologically through the acceptance of all in `the National Liberation Movement' of the organisational `leadership' (that is, dominance) of the ANC in the `National Democratic Revolution'. In other words, it was overwhelmingly agreed by all `mass formations' that now that the ANC was unbanned, it alone was to concern itself with 'national politics'. All existing organisations which had taken a political role were now to drop such a role in favour of the ANC. While the trade unions and civics now relegated themselves to their sectional interests, the UDF was disbanded, and the South African Youth Congress and the Federation of South African Women disbanded and reconstituted themselves as the ANC Youth and Women's Leagues respectively. As a result, the youth have disappeared entirely from the political scene as an independent organised force, while women's organisations are now clearly elite controlled (for example the so-called `National Women's Coalition' made up of leading figures from the main political parties).

In so far as the civics were concerned, the question was: should they retain organisational independence or should they be collapsed into ANC branches? But to retain organisational independence, it was held, they were required to exit from politics. In other words, having conceded the monopoly of politics to the ANC, civics were forced (or forced themselves) into a position that if they were to retain independence, they should withdraw from politics altogether! There was never any question of them retaining a political role distinct from the ANC (that is popular politics). After a short debate in which some (a minority in the ANC/SACP, represented by N'imande and Sikhosana) argued that the political role of the civics should be maintained as they were the future bases of a people's state and thus should become ANC branches/state structures (that is `soviets' of some form!), the majority view prevailed that civics should not be collapsed into ANC branches, but should continue to represent residents irrespective of party-political affiliation. At the same time, the ANC was adamantly maintaining that it alone should be seen as the `leading organisation' (`vanguard') of the `broad liberation movement' and that all other organisations within this movement should recognise the primacy of the ANC in so far as political questions were concerned (Lanegran 1995:114). There is also evidence to suggest that the ANC feared losing popular support to the civics if no clear division of labour between them was agreed (Lodge 1992:61--2). The compromise which was eventually worked out, and which became the dominant viewpoint, was one in which the ANC would have the sole monopoly of politics while civics would restrict themselves to an independent (party-political) role. This compromise was made substantially easier by the fact that the majority of leaders in the civic movement were ANC supporters anyway. It is this dominant perspective, with all its contradictions, which is expressed by a civic leader as follows:

The basic role of the civics is not changed in my view. This role is building people's power and it is something that must play itself out in civil society ... Although the civics, within the UDF, were dominated mainly by the concerns of civil society the front's overall role was largely political. Pulling the civic movement clear of the political net is not easy -- and overlaps of personnel make that very clear (cited in Collinge 1991:8, emphasis added).

The question does not seem to have been asked as to how `people's power', a supremely political project, could be secured by civics if they were to be `pulled clear of the political net'! At the same time, it was clearly appreciated that the dangers of civics becoming bureaucratised or turning into the `conveyor belts (sic) of the ruling party' (ibid) could easily lead to statisation. In addition to arguing in favour of a distancing from the leading party, this dominant position also resolved that they should distance themselves from the local state and 'not attempt to take over local government'(ibid). This was justified in terms of the same arguments, but is probably more accurately explained by the fact that the nature of local government was the subject of negotiations at the territorial level, along with the nature of the central state. In fact, the comment by the civic leader cited above accurately expressed a real political contradiction between popular politics, which civics had incarnated, and the emerging dominance of state politics, which required the `depoliticisation' of civics if these were to remain independent of the ANC (as they rightly insisted on being). It is noteworthy that the same debates did not surface as forcefully with respect to the youth and women's movements. These were organisationally much weaker than the civics and allowed themselves to be `swallowed up' by the ANC.

Currently, a new contradiction has arisen between the `civil society role' and the `state role' of civics. This contradiction, which is a product of state-centred politics, is as yet unresolved because while civics are said to be `watchdogs' for the community on the one hand, their work is also being pushed more and more towards that of `development' on the other. While the idea behind their `development role' is to replace top-down planning with `community participation', the focus has been on `development through negotiation' as opposed to mass popular struggles for self-help (see Work in Progress 92, Sept/Oct 1993; Pieterse & Simone 1994). At the same time report-backs and other measures to ensure accountability have fallen into disuse, while donor funds for `training' and other ostensibly technical programmes have `depoliticised' the role of civics even further and have bolstered the spurious view of these organisations as politically neutral. This supposed neutrality, in fact, leaves them open to becoming state organisations through another route, whether as development institutions at the bottom of the ladder and/or as adjuncts of the state, incorporated into a corporatist structure. The events which unfolded on the unbanning of the ANC are also expressed more or less clearly by activists organising in some rural areas in a manner which is of wider significance, thus:

During the days of the UDF, it was easy for people to understand the struggle. Activity such as stay-aways, barricading etc involved people on the ground and made sense to them. When it came to [territorial -- MN] politics, people lost interest ... When we started, our struggle involved activity, but during the period of the unbanning, people had to go deeper into politics and they lost interest. They no longer wanted to participate ... After the unbanning, there was a lot of confusion amongst the civic organisations because the programme of the ANC and UDF was not the same ... The UDF had encouraged grassroots activity. Even football clubs had a voice in the UDF. The ANC structure and understanding was different as civic structures were not high on the ANC agenda and the civic momentum during the UDF period could not be taken forward. The disappearance of the UDF crippled civic organisations because the ANC was now looking strictly to political issues and not looking to civic related issues and this weakened them (Eastern Transvaal activist cited, Levin & Solomon 1994:256).

The consequences of this process were rather predictable. Because the link between local and territorial politics, which the UDF had successfully managed to enable, was broken, erstwhile organisations of `people's power' collapsed. A report in the ANC-aligned New Nation newspaper, which reflected the dominant view among state-nationalists of the role of `popular organisations' in the post-apartheid period, proclaimed soon after the April 1994 territorial elections:

Except for some of the more centrally located urban settlements, civic organisations are either poorly organised or completely non-existent. And even when they do exist, few have been able to revive street and block committees, which would serve as the ideal forums through which [the government could] consult communities about their [development] needs (New Nation, 3 June 1994).

The transformation of popular politics from 1990 was not only expressed in the debates surrounding the role of civics, it was also obvious in the changed role of `mass mobilisation'. It became more and more obvious that such mobilisation was now initiated and directed solely `from above' by leading members of the ANC hierarchy, and seen as a measure to `put pressure' on the negotiating partner, the National Party and the state. It was no longer part of a process of `empowerment'. So that when the ANC declared 1991 `The year of mass action for the transfer of power to the people', such action was designed to force the government to meet territorial demands such as `the immediate release of political prisoners ... and the unconditional return of exiles, the dismantling of bantustans, an end to violence against the people, the immediate repeal of all repressive legislation, the establishment of an interim government and Constituent Assembly'.

In fact, one of the main differences between the `left radicals' and the more `moderate nationalists' within the ANC seems to have been the role which such mobilisation was to play in the transition process. This is brought out clearly in a frank review by Jeremy Cronin, one of the main theorists of the South African Communist Party (SACP). He distinguishes between three different `strategic outlooks' characterising liberation politics: `the boat, the tap and the Leipzig way'. The first position was one where democratisation was seen as resulting from a negotiated pact between elites which deliver their constituencies. Mass action was perceived as `rocking the boat'. The second position was one where `mass action' was to be turned on and off like a tap. He comments, quite correctly, that `struggle in strategy two is not about the self-empowerment of the working masses. Instead, struggle is rather more narrowly seen as empowering the negotiators so that they can bestow upon the people their liberation' (1992:42--3). It is hinted, although never openly stated, that this was the dominant position within the ANC which was associated with Mandela himself.

The third `strategic outlook', according to Cronin, was `the Leipzig way' or the position according to which `the people transfer power to themselves in a revolutionary moment' of insurrection as in Eastern Europe in 1989. He argues -- also quite correctly -- that such a perception is also in essence elitist, apart from being impracticable in the South African case, so that it ultimately reduced itself to the `tap' position. The author frankly acknowledged that `all three have a tendency to fall into one or another version of statism' (ibid:53). These words proved prophetic. An attempt to `turn on the tap' on 7 September 1992 outside the Ciskei -- as part of the bid to remove Gqozo, the local bantustan leader, from office through the mobilisation of `mass action' -- resulted in 28 deaths and over 200 injured in what became known as the `Bisho massacre'. More long-term effects were the end of `mass action' as a tactic, as the ANC dropped it like a hot potato, and the final demise of the `left statists' as a meaningful force within the ANC. This not only showed that `mass action' had by that time little popular content other than providing canon fodder for the state's bullets, but it also allowed the right-wing in the ANC to present the choice facing the political movement as being between peaceful transition, which was equated with negotiations, and the escalation of violence, including mass action. This comes across very clearly in the ANC document Negotiations: a strategic perspective (ANC 1992), which provided, following on from Slovo's theorisation of the famous `Sunset clauses', the rationale for entering into a government of 'national unity' with the National Party.

It must be recalled that state and Inkatha violence was being systematically unleashed on township residents during the period leading to the elections, and that the ANC was powerless to stop it. Even though the formation of `self-defence units' (SDUs) was originally encouraged by the ANC, these were ordered to disband in 1994 as a quid pro quo in the disarming of Inkatha and all `private armies'. The following highly informative newspaper reports, which showed that not everyone concurred with the disbanding of SDUs, appeared in 1994, immediately after the territorial elections:

While self-protection is an inalienable right to any person, this has to be done within the existing laws of the country. We also accept that the SDUs came about under special circumstances -- the time when police were refusing to apprehend known vigilantes ... The SANCO (South African National Civics Organisation) activist said the time had come for the weapons of both the SDUs and other private armies (sic!) including the right-wing groups, to be handed to the authorities of the democratic government for safe-keeping (New Nation, No 4 13 May 1994). GUNMEN armed with AK47 rifles executed 12 residents in Thokoza on the East Rand on Friday night just three days after South Africa's first democratically elected State President, Nelson Mandela, was inaugurated. The massacre is the first since the elections (New Nation, 15 May 1994). Although the SDUs have been at the centre of controversy, residents described their role as vital -- especially at the height of unrest in the township. South African National Defence Force spokesperson Colonel Chris du Toit said initial inquiries indicated [that] `since the ANC ordered (sic!) the units to cease operating there has been evidence of in-fighting' (New Nation, 20 May 1994).

Trade union struggles

Interestingly, the trajectory of the modern trade union movement in South Africa was not all that different from other popular organisations. The main difference was that trade unions were able to organise a constituency which was able to effectively challenge more than the structure of apartheid local government. The challenge to South African business interests effectively undertaken by a relatively powerful and disciplined trade union movement was instrumental in pressurising big business in particular to push towards a negotiated transition to democracy. Trade unions were much less successful in organising workers in small businesses, women, rural labour, and the unemployed, while eventually they even lost support among migrant workers as they concentrated most of their work on the fully urbanised. Their broad historical trajectory was also one of `politicisation from below', followed by a process of depoliticisation which itself only acted as a prelude to a deeper process of `repoliticisation from above' -- of entering political society in the 1980s and then exiting from it only in order to re-enter politics on the side of the state in the 1990s. As is well known, the history of the modern trade union movement largely originates in 1973 when 100 000 workers went on strike in the Durban area. These largely spontaneous mass strikes revitalised trade union activity which had been dormant during the `decade of peace' after the banning of the ANC/SACP (and the PAC) along with SACTU (South African Congress of Trade Unions), which was largely the organ of the `Congress Alliance'. The unions which developed as a result of the Durban strikes saw it as crucially important to maintain their independence from nationalist organisations in order to avoid the same fate as SACTU. Rather they concentrated on developing strong shop-floor structures and a system of worker representation based around shop-stewards. Apart from being intrinsically democratic, it was argued that such a system would enable a small union organisation to better withstand state repression (Webster 1988, Lambert & Webster 1988).

This fiercely independent stance became the dominant position in Fosatu (Federation of South African Trade Unions), which was launched in 1979, and actually came to be adhered to rigidly like an article of faith (until the formation of Cosatu in November 1985) theorised by the intellectual high priests of the `white left' who had been instrumental in servicing the development of the new unions. Basically, the view was that `working class politics' should grow out of shop-floor struggles. Unions should not identify with any nationalist political organisation as union members belonged to different organisations, and also because it would mean accepting the dominance of a petty-bourgeoisie who supposedly dominated the township-based organisations, which, in any case, were said not to be as democratic as trade unions.

With the increasing development of popular struggles in the townships (in which trade unionists lived, after all), the question which was to occupy the centre of the intellectual stage on the Left in South Africa came to the fore, namely the question of the relationship, if any, between trade union struggles and township struggles or workers' organisations and national politics. This single question has given rise to a large volume of debate covering not only the above issues, but also ranging more broadly to include the question of class alliances, the road to socialism, the nature of the Freedom Charter, the question of `unity in the struggle', `liberation vs transformation' and so on.

More briefly, it is known as the debate between `workerists' and `populists' and was also conducted far beyond the confines of popular organisations, where it became transformed well beyond its original spheres of concern regarding the relationship between civil society and politics into an often acrimonious academic debate where arguments merely served to further entrench already rigidly adhered to positions. Shortage of space precludes an assessment of this debate here, although it is proposed to study it in detail as part of future work. Rather, a very brief account of the changes to the trade union movement which paralleled this debate is of greater importance to our immediate concerns. Briefly, the pressure for unions to become more involved in township nationalist politics came overwhelmingly from workers themselves, as they experienced not only oppression in the workplace but the same urban problems and coercion as all other residents at their homes. The main organisations which voiced these pressures were the Local Shop-Steward Councils (known simply as `locals'), which brought together shop stewards from a given urban area and which originated in the East Rand (Germiston, Wadeville, Katlehong). According to Webster (op cit:183):

Founded as a way of involving shop-stewards in the organisation of unorganised factories, these councils spread rapidly during the 1981--82 strike waves ... At the centre of this social movement in the East Rand hostels was the migrant worker.

Although the locals were originally founded as a way of spreading union organisation to other factories and to fight against scabbing, organised as they were in urban townships, they were bound to become involved in township issues. They started to become involved in questions of housing, unemployment benefits, adequate pensions and maternity rights, inter alia (ibid; Swilling 1984:118). In the words of Jeremy Baskin's study of a Shop-Steward Council in 1982:

The shop-stewards' council is characterised by its militancy, mutual support ... and strong grassroots organisation ... All this is made possible by strong local organisation. Workers in the area share many problems. They use the same buses and trains, they live in the same areas and they know other workers in neighbouring factories. The common conditions which workers face at local level becomes a major spur to militancy, once organisation gets started ... The fact that workers began presenting common demands generally strengthened their position in the area ... Workers are encouraged to see beyond their own union to the struggles of workers as a whole (Baskin 1982:47--8).

In addition, the locals became bases for democratic control over unions as more power lay in the hands of shop stewards and these structures were not bureaucratic. One shop steward explained:

We talk of unity ... what kind of unity and how far we should go as a local. What sort of help, what sort of things we should do, and the disciplinary procedures. Because if we are to be united, we have to have disciplinary procedures and some clear objectives ... As workers, then we are involved in political issues, so we have to be clear on how to react to such things ... Problems like rent have come up ... we have to do some things outside the factory (cit Baskin ibid:52).

As a result of these developments, the `Fosatu line' came more and more into conflict with its own shop stewards, especially after the formation of the UDF and the intensification of community struggles, and the greater and greater pressure from below for joint community-trade union action. What had been a very correct tactic in the early 1980s had become, by the middle of the decade, a sterile dogma, as the objective situation had fundamentally changed. One shop steward from the Metal and Allied Workers Union (Mawu) argued:

The situation of the worker in South Africa is that they are oppressed and exploited. The struggle goes beyond the factory gates. Workers must address themselves to the problems of rents, shacks, electricity tariffs, schools, recreation, etc. In Fosatu and Mawu, workers have been openly discouraged from taking up these issues and political organisations have been openly criticised. We recognise that the trade unions are not political organisations. But for them (Mawu) to say no politics in trade unions is nothing else but to keep their politics of reformism inside the trade unions (cit Swilling 1984:119).

It was this pressure from below which ultimately led to the formalisation of what Webster has called `social movement unionism', finally expressed in the formation of a new giant union federation, the Congress of South African Trade Unions (Cosatu) in 1985. Unlike its predecessor, Cosatu encouraged the politicisation of trade union activity and collaboration between unions and the UDF, even adopting the Freedom Charter as a guiding principle. Cosatu, therefore, became involved in building `worker control' (the equivalent of democratic `people's power' in the factories and unions) and insisted on contributing to the `working-class leadership' of the 'national democratic struggle' (although what precisely was meant by such leadership was not always clarified). Thus Jay Naidoo, the General Secretary of Cosatu:

Non-political unionism is not only undesirable but impossible in South Africa. Therefore, we believe that though COSATU is not a political party, COSATU has a responsibility to voice the political interests and aspirations of organised workers and also more broadly the working class. To do this, we have to look at how we build workers [power] and how do we locate workers as the leading force in our struggle for national liberations (sic) ... The key element in the building of the labour movement was, and still remains, the democratic principles of worker control ... In real terms, it means that the members of the trade union must have absolute control over all decision-making in the organisation ... COSATU has high regard for those communities and organisations that are building strong grassroots organisation in the form of area and street committees. We encourage this and see it as COSATU's policy for members and local structures of COSATU to play an active role in building such structures (Jay Naidoo 1986:3, 4, 8).

In this way, Cosatu entered political society and its national campaigns made a conscious attempt to address issues which were pertinent to the interests of the poor and unorganised in general and not simply to those of the organised workers (the most famous being the `Living Wage Campaign'). A survey on the state of the unions published in 1985 noted that in a sample of 23 of the largest industrial unions, there were 12 462 shop stewards, with 1 443 shop steward councils in place (Collins 1994:35). Not surprisingly then, Cosatu placed much emphasis on the role of `locals' which were seen as the foundation of the organisation:

In particular, the role of the shop steward councils was crucial. They assisted in organisational work and developed ordinary worker leadership. The local confronted the political issues of the day and developed resistance in practice (ibid:36).

However, by 1987 a number of weaknesses were revealed in an assessment conducted by national office-bearers. This noted that `local structures were weak and the Cosatu regions were not functioning' (ibid). Moreover, Cosatu was unsuccessful in organising the poorer sections of the population. An attempt to organise the unemployed failed, migrant workers were ignored as the fully urbanised workers came to dominate trade unions more and more (Mamdani 1994:chapter 6), while rural labour was left unorganised, a fact which cost the National Union of Mineworkers dearly during its 1987 strike (as scabs could be recruited from rural areas). It is not clear whether an attempt was made to explain and correct these weaknesses.

At the same time, for the overwhelming majority of leaders and commentators, this entering of trade unions into political society was explicitly or more often implicitly, taken to be a temporary situation. In the words of the sociologist Eddie Webster:

Where, as in South Africa, the majority does not have a meaningful voice within the political system, unions will inevitably begin to play a central role within the political system (Webster 1988:176).

There was, of course, nothing inevitable about it. If the Fosatu line had remained dominant, the entrance of trade unions into political society would have been limited. But the unstated assumption always seemed to have been that as soon as political parties with support among the oppressed majority were able to operate openly, unions would then `withdraw' to their 'natural domain' in civil society, ascribed to them by the division of labour.

This is indeed what happened, but it happened in a complex and contested way. Cosatu actually attempted, but failed, to get a place at the negotiating table, where only political parties and state agencies (for example bantustan governments) were represented. The National Party government absolutely refused and the ANC did not pursue the issue with much vigour anyway; neither probably did Cosatu, as it was agreed that its interests would be represented by the ANC and SACP delegations to the talks (and eventually a number of Cosatu figures joined state structures, ostensibly to represent `workers' interests' in the state). In actual fact, Cosatu in particular did not, therefore, so much desert the realm of politics as that of popular politics. In the 1990s, it has geared its main efforts towards having an input in all aspects of policy as they affect workers, and this process has extended more and more into a dominant corporatist trend.

In the words of one recent assessment of Cosatu's role in the 1990s, this rather astonishing remark appears:`[t]he opening up of political space after 1990 has meant that the unions are now in a position to directly extend their influence beyond industry to the national economy' (Collins 1994:36). This statement is astonishing because it has been regularly stated in different ways since 1990 (often precisely as an argument in favour of corporatism) and because Cosatu's role had never been restricted to `industry' alone! This remark systematically brushes aside the political role of Cosatu in the 1980s and is indicative of the political amnesia suffered by many leading trade unionists in the 1990s (in this case a person ostensibly opposed to corporatism!). Yet another major writer on trade unions (this time a major advocate of corporatism) openly remarks:

the trend is towards corporatism -- whether it is expressed through a social contract, reconstruction accord or socio-economic pact. While there remain different expectations of the scope, form and duration of such an arrangement, these are basically differences of emphasis. The end point is the same, and the NEF (National Economic Forum), NMC (National Manpower Commission) and NTB (National Training Board) are first steps in this direction ... for its proponents, bargained corporatism is nevertheless the best available option or, more pessimistically, the least worst alternative. For the union movement in South Africa, the corporatist path is, in effect unavoidable (Baskin 1993:2, 7 cited Shaw 1994:247)

Like the supposed inevitability of trade unions entering the political arena in the 1980s, there was nothing inevitable about corporatism in the 1990s. It was the result of a political choice by the Cosatu leadership as it decided to vacate the political space, surrender it to the ANC and then attempt to `influence' the policies of the ANC. This concentration on policy was undertaken in such a way that it emphasised the technical aspects of policy detached from explicit politics (let alone popular politics). Policy now became an issue for `experts' (who are overwhelmingly white, given South African conditions), and `consultation' with `interested parties' became one for `workshops', so that policy has been depoliticised, reduced to technical questions and removed from scrutiny by democratic structures. As one unionist put it: `policy has become ``received wisdom'' and the result is that the structures are simply ``transmission belts'' for discussion from above' (Collins op cit:38); or again:

Local agendas are dominated by the very many issues that come from `head office' which require mandates for national policy or national action ... the local has become a function of the national, the passive recipients (sic) of national directives (Marie 1992:22, 23).

At the same time, union representatives often go to 'negotiating forums':

without clear mandates either from the federation (COSATU -- MN) or their unions ... Where report-backs are given at the local level, they are often presented as top-down reports with little room for debate. They are often not discussed at all, in favour of dealing with more local issues (Collins op cit:38).

A survey of trade unionists conducted by the South African Labour Bulletin in 1992 (see Keet 1992 and also Pityana and Orkin 1992) found a general trend of the weakening of worker control with one national leader remarking that `workers are losing and losing workers control and it is in danger of becoming just a slogan' (Keet 1992:29). This was manifested inter alia in the fact that few shop stewards now turn up to (shop-steward council) local meetings (100 out of 500 possible members in Durban, 100 out of 1 000 in Johannesburg, Collins op cit:35); `Cosatu office bearers are less subject to the direct workers control that shop stewards can exercise within their own affiliates' (ibid); `bureaucratic tendencies have become evident both at COSATU and affiliate level. These tendencies are not restricted to officials, but extend to worker leaders as well' (Collins: op cit:39). The reviewer concludes:

To abandon worker control is to abandon union democracy, and to accept that ... formal democracy empty of any ongoing, direct control by members is the best that the trade union movement can do, given the conditions in South Africa in the 1990s ... Workers control of unions was seen as a means to worker control of production and society as a whole. It is a significant irony that in the 1990s, the unions are struggling to return to worker control of their own organisations, with control of production and society an ever receding possibility (Collins 1994:40).

It would indeed be tempting to causally link the dominance of corporatism among trade unions in South Africa with the weakening of control from below, but there is, as yet, little to suggest that the former is a function of the latter. It could be that they are both effects of much broader tendencies which, as we have seen, have involved other organisations of civil society -- such as civics -- as well. In any case, it is not a priori impossible that some form of social contract between the state and various organisations of civil society, including trade unions and civics, is compatible with popular politics and genuine popular participation. In fact, it is arguable that in the absence of such a combination, reversal to other forms of statism are practically inevitable. Corporatism, however, implies the `statisation' of popular organs of civil society, their `politicisation from above', which in no way requires a surrendering of organisational independence; it only requires the absence of political independence. Organisational independence is fully compatible with incorporation within state structures and, after all, this is typical of classical social democracy, where social democratic parties have `their' trade unions in the same way as communist parties had `theirs'. Rather than being about the democratisation of the state from below as in the 1980s, the dominant trend in South Africa which is in the process of becoming entrenched is about the statisation of civil society from above. It is not difficult to agree with Sam Shilowa, the general secretary of Cosatu (in one of the quotations heading this essay), that a new form of statism is beginning to dominate in post-apartheid South Africa. I have, so far, attempted to show that there has been in South Africa a fundamentally important process of political change between the 1980s and the 1990s, involving the major popular organisations of civil society (civics and trade unions in particular) and their relations to the state. This shift in the `mode of politics' has been one where a form of popular politics, with all its contradictions in the 1980s, has been replaced by `top down' politics, bureaucratic tendencies and, in brief, the greater centrality of state politics in the operation of these organisations which have either been, or are in the process of being, radically transformed. I have restricted myself to outlining some aspects of this change without attempting to discuss at length any explanations for it. Nevertheless, no explanation, however limited, is possible without some theoretical perspective. We must now turn to the brief elaboration of such a perspective.

Civil society and political society

In the 1990s, the process of democratisation in South Africa has been overwhelmingly state directed, not only because political parties and state agencies have taken the initiative and provide the fora in which decisions on such democratisation processes are made, but also largely because of the weakness of a culture of popular democracy and the absence of popular institutions through which that culture could be expressed. At the same time, many of the official documents emanating from the ANC (such as the Reconstruction and Development Programme (RDP) pay lip service to the `South African tradition of a strong and independent civil society' and often uncritically assume that such civil society, in particular what are now called CBOs (community-based organisations -- the term basically refers to civics), can provide the vehicle for popular direction of the democratisation process. More sober assessments, however, as we have seen, paint a different picture of a moribund popular movement, so that Sam Shilowa, the general secretary of COSATU recently had to emphasise that: ` ``the mass-driven character'' of the RDP is at this stage a total myth' (Shilowa 1995:36).

Central to the new process of transformation in the 1990s is what the ANC, following on American jargon, calls an `affirmative action programme', which fundamentally corresponds to what, in post-independence Africa, was known as `Africanisation', namely the appointment of personnel in the civil service and public sector more generally from the ranks of the hitherto colonised population. In Africa, this policy arguably had three major effects. First, it created a middle class and especially one which was tied to the state (as colonial officials left their posts and had to be replaced; the lower levels of the civil service were by and large Africanised anyway). Second, it transformed a demand for democratisation (ie inter alia the opening up of job opportunities to as many people as possible and the democratisation of access to resources and thereby the democratisation of state structures) into a replacement of `white faces by black faces' in the state apparatuses; in other words, democratisation was reduced to the formation of a middle class `from above'. Third, it created important new divisions (or at least intensified old ones) over access to resources as it was only some Africans who were given access to jobs (for example `kikuyuisation' was a more accurate description of this process in Kenya under Kenyatta). This last point became extremely dangerous as only some nationalities (or political parties), through their exclusive access to state resources, monopolised possibilities of accumulation, as the petty bourgeoisies of other nationalities (and parties) were left out. In sum, in post-colonial Africa, democratisation was reduced to `Africanisation', popular democracy `from below' was replaced by state nationalism `from above' (Mamdani 1990).

In South Africa `affirmative action' has been exhibiting some similarities as well as some differences from the pattern above. On the one hand, it is difficult to refer to `Africanisation' in South Africa, as everyone (including white South Africans) is African; on the other hand the term `African' tends to be restricted to the apartheid category (in relation to `European', `Indian' or `coloured') so that such a term would privilege only one section of the hitherto oppressed population. So much for terminology. Turning to more fundamental aspects, affirmative action is described as:

a type of `positive discrimination' as a measure to correct imbalances created by centuries and generations of ... oppression. It cannot be a permanent feature of society or an organisation, but rather an interim measure and a means towards full equality and an end to discrimination (Mayibuye, September 1991:32).

The above definition is in a sense contradictory since if affirmative action is supposed to correct `centuries and generations of oppression', it can hardly be temporary (unless its temporary nature is measured in decades, if not longer), and if it is to be temporary, it can hardly hope to correct such `imbalances'. This is, in a sense, revealing as the contradiction is only reconcilable if we are talking of the creation of a middle class. In this case, racial `imbalances' can be overcome pretty rapidly as blacks can be appointed to jobs without regard for democratic participation. In actual fact, this is precisely what has been happening. Rather than contracting, as the ANC had always argued, the state bureaucracy in the 1980s is expanding as the ANC acceded to a demand to keep apartheid civil servants in their jobs (including in the ex-bantustans) and is attempting to bypass their lack of cooperation by creating new appointments.

In addition, in South Africa, businesses in the private sector are cooperating in recruiting blacks for managerial positions. Whatever the extent of progress on this score, the point really remains that the process is one where a middle class is being created among `black Africans' primarily (the lower ranks of the civil service, particularly in the ex-bantustans, as well as the overwhelming majority of employees in all sectors are black anyway). In neither of these cases is it a question of democratising appointments or indeed state structures themselves. The popular struggle for democracy has been replaced by a scramble for state posts. This has been but an effect of the ANC's obsessive concern with `attaining state power' as a pre-condition for democratisation, as opposed to the reverse position -- the democratisation of politics and social relations as a prelude to the establishment of a democratic state -- which was gradually being developed by the popular movements.

The last aspect of this question worth mentioning is that such a procedure is creating contradictions between `racial groups', most obviously expressed as white (and indeed coloured in the Western Cape) resentment against `unqualified' blacks taking over highly paid jobs and sometimes fearing loss of their own. This was a factor in the Western Cape voting for the National Party at the general elections in April 1994, for example. At the same time, the top echelons of the civil service remain largely under the control of appointees of the apartheid regime (as the ANC guaranteed that all civil servants would retain their jobs) and are thus able to retard (and perhaps even sabotage) government policy. This leads to frustrations and resentment by impatient politicians who can also use this as an excuse for the lack of progress in `delivering' promises to the majority.

While this is happening, the state is directing a process of `reconciliation' between races which arguably has little `public support'. The choice of language is deliberate, because in a state-directed endeavour like `affirmative action' or `reconciliation', it is no longer possible to speak of popular democracy, but only of `public support', where the people are reduced to passive spectators. In the period of the UDF, as we have seen, the construction of a 'new nation' was attempted on the basis of combining South Africans of all races under the banner of popular democracy and opposing them to anti-democrats, but now the issue is the construction of a nation by the state itself and thereby the reproduction and even possible deepening of racial divisions. This is especially obvious in conditions where the majority have yet to see any benefits from liberation or the RDP (despite its exclusive accent on top-down `delivery'), and where official `reconciliation' appears to mollify extreme right-wing racists and to permit the `perpetrators of apartheid' to escape retribution. In this manner, popular `anti-racism' has been replaced by state directed `reconciliation'. As a result, nationalism has now acquired a new meaning. A media report commented recently:

the mood of nonracialism ... is fading. The views of the proponents of Black Consciousness are increasingly being heard again and people are listening to them. It is giving rise to the development of an Africanist ideology ... (Southern Africa Report, 8 September 1995).

In brief, these examples point to a process similar to that of the `statisation' of nationalist politics that was earlier noted with reference to both popular civic organisations and trade unions. In so far as a dialogue is taking place regarding the nature of democracy, it is more and more being dominated by technical-legal questions regarding the constitution, a bill of rights, regional versus central powers, individual freedoms, the `truth commission' and so on, with the result that ordinary people are largely bypassed by debates (although they are asked in the press to participate by writing to the constituent assembly with their suggestions!) and `consultation' is supposed to take place through `workshops'. Indeed `workshops' have proliferated in all spheres of political activity and have been justified ostensibly as a way of `consulting' and involving all `interested parties' in the debating, advising on and even the formulation of policy. This has largely amounted to an extensive form of corporatism where `experts' as well as leaderships of various community organisations and trade unions have gradually become part of a complex state structure. In the sphere of `liberation politics', `workshop politics' has gradually come to replace `mass politics'. Having no real legal, let alone constitutional authority, `workshops' have no powers other than `consultative' ones. They are more often than not representative of no-one other than themselves, as members have not usually been mandated by any political body or organisation. Not only are workshops largely unrepresentative, and popular representatives when they exist are rarely mandated, but as a result their resolutions and recommendations can be ignored as and when necessary. De facto therefore, the `dialogue on democracy' is not being conducted between the ANC and the people organised under distinct, independent and clearly defined politicised organisations. Rather, it is being conducted between the ANC and the representatives of white capital and the white fraction of the bourgeoisie overwhelmingly represented by the National Party, (along with right-wing Afrikaners and the Inkatha Freedom Party in so far as `regionalism' or `traditional leaders' are concerned) above the heads of the majority of the population. The role of multipartyism and universal suffrage in this process requires some elaboration, as both have been uncritically celebrated as major achievements by the media and in supposedly more academic writing (for example Friedman & Atkinson 1994).

It is perhaps important to recall first of all that multipartyism was established in the immediate post-independence period in Africa as the departing British colonial power in particular imposed a `Westminster system' of government on the newly independent states. Political parties, however, generally mobilised ethnic and regional interests, as the African petty bourgeoisie had not by then been able to constitute itself as a unity and was fundamentally divided along such lines. Multipartyism was quickly scrapped as the nation-state project could not countenance ethnic resistance and purported threats to territorial unity. Multipartyism imposed `from above', therefore, quickly proved to be a failure. In recent years, African authoritarian statism has generally survived the (re)introduction of multipartyism under the pressure of `political conditionalities' imposed by the West as a requisite for continued financial and political support in the `New World Order'. In South Africa, there have been some similarities with the African experience as neither the incoming ANC nor the outgoing NP had ever been committed to multipartyism and universal suffrage. Indeed, the ANC only started mentioning its adherence to multiparty democracy in the 1980s as the dialogue was initiated between it and the apartheid regime. The Freedom Charter only made a commitment to `government by the people', as we have seen. This could have easily been interpreted as a one-party system where the party is said to represent `the people' (as had been done on numerous occasions in Africa and elsewhere), and was provided with a popular democratic meaning by the UDF in the 1980s as we have also seen. The National Party was, after all, the main architect of the refinement of colonial disenfranchisement of the majority known as `apartheid'. Neither therefore had a history of commitment to multipartyism and universal franchise. The commitment to both of these features of the state in South Africa by both of these parties was the result of a number of factors, external and internal to the country.

The external factor was clearly the changed international conjuncture and the end of the Cold War, which generated pressure from the Western powers for democratisation in Africa (as well as in the Third World and in the ex-Soviet Union more generally). Internally there were arguably two major factors. The first was the obvious fact that both the National Party and the ANC, both white capital and the emergent black petty bourgeoisie, realised that they had to compromise over control of the state and that neither could win an outright victory over the other. A power-sharing arrangement through universal suffrage had to be worked out and, in fact, they have gone so far as to share the `governance' of the country through the establishment of a `Government of National Unity' until the next territorial elections in 1999.

The second factor was the pressure `from below', which could not have countenanced a `one-party state', as in the rest of Africa, because popular organisations (especially the trade unions, but also the civics) were arguably jealous of their independence and as we have seen argued strongly against statisation, a danger of which they were aware. Multipartyism and universal suffrage were, therefore, in South Africa, the result of an internal compromise between class forces and not the result of an external imposition. Therefore, it arguably has more of a secure basis in the country than it had in the rest of Africa, although the two main parties are still representative of racial groups and there is no indication of a change in this regard as yet. At the same time, the evident cynicism of the major parties vis-a-vis the first territorial elections of April 1994, whereby honest voting was apparently less important than the clinching of a deal in which all main parties were `given something' and were thus `satisfied', bodes ill for the future as it shows contempt for the `popular will' (Southall 1994).

Through the formation of a `government of national unity', all radical opposition has largely been delegitimised, and in so far as such opposition is being expressed from the Left, it is emanating from marginalised groups influenced by the `Trotskyist tradition' in Cape Town, which restrict themselves to denouncing the `politics of class compromise' and the dominance of the petty bourgeois nationalists, and call for the formation of a `workers' party'. There is little in these statements concerning democracy of the popular or any other variety, and they largely consist in re-stressing the vulgar fundamentals of some kind of millenarian `marxism' divorced from popular experience. This perspective is still overwhelmingly statist (as can be seen in any of the political writings of Neville Alexander for example). At the same time, the Left within the `Congress Alliance' is hamstrung by its support of the state (and particularly of statism) to such an extent that even though the need for independent popular organisations is often reiterated, there is a complete failure to understand how this could be accomplished other than by initiatives `from above'. Any genuine popular initiative would, of course, have to start by re-politicising `civil society' and would thus come into (some sort of) conflict with the ANC-dominated state, something which would be too much of a risk for the territorial leadership. Thus, calls for greater democracy from these quarters remain, unfortunately (and predictably), rather empty.

There is, as yet, as far as I am aware, no serious analytical study of the transition period in South Africa. There are, however, some important studies which attempt to assess the period of the 1980s. Most prominent among these are Marx (1992), Murray (1987) and Lodge (1991). Marx's book provides a wealth of detail, but concentrates exclusively on a discussion of the changing ideology of the anti-apartheid movements through the words of their leading personnel. Because of this, it restricts itself to the territorial level and absolutely fails to recognise not only the differences between local and territorial ideologies and struggles, but also possible contradictions between the leaders of such movements and the led. This modus operandi is particularly unacceptable in South Africa, as much of the ideological struggle in the 1980s surrounded precisely the issue of internal democracy and the accountability of leaders to the rank and file. Moreover, it is not particularly useful to concentrate on a discussion of the ideology of the leadership if, as we have argued, it was in fact the rank and file who were the main motive force behind the opposition to the apartheid state for a considerable time . Marx is thus forced into accepting the leadership view at face value such as the apparent `pragmatic' conception of the all-inclusive ideology of nationalism contained in the `Freedom Charter'. The point was that ordinary people, mobilised as they were in their own organisations, were actually giving the populist slogans of the `Freedom Charter' a popular content most obviously apparent in `people's power' which was anything but `all inclusive' and `populist', but which had an evident bias towards working people. Marx's own perspective ultimately leads him to adopt a linear conception of ideological change in South Africa, whereby the opposition's ideology is seen as following the chronological and somewhat teleological path of the organising principles of `race' (Black Consciousness), 'nation' (UDF) and `class' (Cosatu).

Murray (1987) is more critical of the `populism' of the UDF than Marx, preferring the supposedly more `working class' orientation of the `National Forum' (a small association of Cape Town-based Trotskyist and Black Consciousness groups with little popular support). This is determined by the former's adherence to a supposed `multi-racial radical populism' dominated by the `petty bourgeoisie' rather than the more frank anti-capitalism of the `National Forum' which understands that the `working-class struggle against capitalist exploitation and the national struggle have become one struggle under the general control and direction of the black working class' (ibid:230). Again, there is little attempt to understand the complexities of ideology here. Not only is this simply the slogan of an intellectual leadership with scant relations to the working people, but there is no attempt to investigate the possible relations or contradictions between the two, or even the possible inventiveness of popular struggle. In fact, there is little attempt to listen to popular perspectives which may not fit within the parameters of the author's rigid dogma. In many ways, Lodge et al's (1991:140) account is the most sophisticated of the three main analyses of the 1980s, as the authors are sensitive to the complexities of ideological formation and identity, and indeed recognise the importance of popular initiatives in the construction of a struggle from below. Thus:

Notwithstanding the ANC's popularity and the universal authority of the Freedom Charter, the UDF was an intellectually intricate organization, perhaps more so than even its leaders were aware. In its public rhetoric and printed polemics, different political persuasions were evident. And if ideology is taken to mean more than the self-conscious expression of doctrine and principle, and if the term is understood to embrace an organisation's repertoire of activities, then the picture becomes quite elaborate. For, just as the UDF's rhetoric animated its huge army of supporters, the organization became infused with their ideas and beliefs. These drew on folk morality and local interpretations of tradition as much as on externally derived conceptions of capitalism or socialist democracy. People's power was the crystallization of the rich and volatile mixture of ingredients within the UDF


And yet, despite this excellent assessment, Lodge insists on demarcating the ideology of the UDF along three strands: 'nationalist', 'national democratic', and `socialist', which at the time were clearly `vertical' divisions (op cit:129--134): `vertical' because they corresponded to divisions around which leaders would attempt to gather support and interpellate audiences. Thus, whether or not one agrees with Lodge that these currents were the dominant ones, the fact remains that it could be argued that a far more important division in the movement, especially with some historical hindsight, was that between those who attempted to stress popular democracy and control, consultation in decision-making, independence from political parties and leadership accountability, and those who were on the side of authoritarianism, statism and later the exclusive adherence to multiparty democracy. In other words, it could be argued that a more important ideological division, which cut across those identified by Lodge, was a `horizontal' one between popular democracy and statism (`state democracy' or authoritarianism). In the absence of an attempt to at least examine this distinction -- to detach popular-democratic nationalism (or socialism) from state (petty bourgeois) nationalism -- it seems to me difficult to do justice to the popular movement in South Africa in the 1980s, or to adequately explain the transition between the popular politics of this period and the state politics of the 1990s.

African scholars have recently began to reassess this kind of question from the perspective of the recent crisis of the African state, which is, in large part, a crisis of legitimacy. Earlier conceptions of the `betrayal' of `the revolution' by a petty bourgeoisie have been rightly rejected as explanations which not only rely on conspiracy theories but often idealise class subjects (the working class and/or petty bourgeoisie) and endow them with innate powers to make history on their own independently of social relations and struggle. Like most such analyses, they thereby end up ignoring issues of democracy, along with struggles between different forms of democracy and nationalism, conceding such notions in argument and in political practice to the dominance of the hated `petty bourgeoisie'. They therefore have little to offer as an alternative to nationalist statism but more statism, albeit tainted with a different ideological brush.

Central to this new work has been the writings on popular movements, particularly the theorisation developed by Mamdani (1990) and Mamdani and Wamba-dia-Wamba (1995). The former in particular analyses the development of state nationalism in the Africa of the 1960s, along with the defeat of the popular movements which had been a major component of the struggle for a new democratic independent state. He makes two crucial distinctions: first between `popular democracy' or 'national democracy' on the one hand, and `state nationalism' (the 'nation-state') on the other; second between 'national' as an equivalent of `democratic', and 'national' as the opposite of `local' (that is, as the equivalent of 'nationwide' or `territorial'). Through these distinctions, he is able to argue that popular demands and organisations, although often `local', were also often more democratic than `territorial' demands and organisations which reduced all such demands to the single one of `independence'. Anti-colonialism was not necessarily the same as anti-imperialism; for this perspective the latter implies democracy, the former does not necessarily do so. In addition, for Mamdani (1990:54), state nationalism was legitimised by the nationalist historian who:

tried to play down whatever features may detract from the national character of a social movement so as to emphasize its nationalist credentials, to remove the notes which could not easily be harmonized within a single national chorus, s/he also ended up obscuring local issues so as to cast in bold the one single national demand: self-government or independence! To use a somewhat modern metaphor, what was really a `rainbow coalition' was painted in a single grey!

Mamdani argues that the social movements which together made up the nationalist movement in the 1940s were gradually defeated by a reform strategy initiated by the colonial state and continued after independence. This reform had two consequences, first a splitting of the anti-colonial forces as concessions were made to bourgeois aspirants within the broad front, second the legalization of `the most important political organisations (trade unions, cooperatives, friendly societies), to bring them under the scrutiny of the state and step-by-step to undermine their autonomy and any element of popular accountability they may have developed' (ibid:49). He also draws a distinction between the `social movement' and the `political movement', the latter transforming itself into the state at independence.

Gibbon (1993) develops this perspective in his work on civil society and `developmentalist' states. He argues that a democratisation process, which in the interests of the working people must fundamentally always involve not a `deepening' or `strengthening' of civil society, but two mutually reinforcing aspects: the politicisation (not statisation) of civil society, and the democratisation of the state. Politicisation implies the entering of civil society organisations into the political sphere whereby their members are addressed as citizens and not simply as members of a(n) (`interest') group with particular interests determined by the division of labour. Democratisation suggests the transformation of the state so that it reflects the `sovereignty of popular institutions'. This argument is elaborated in his work on late colonial Tanganyika (Gibbon 1994), where he distinguishes a realm of `territorial politics' ('national politics' without the popular democratic content). He shows that local movements were often raising issues and democratic questions which went far beyond what Tanu was prepared to countenance and that Tanu and the colonial state actually combined to suppress and otherwise undermine the autonomy of such movements (similar points are made by Kriger (1992) in her work on the liberation war in Zimbabwe). The determining factor in the defeat of popular movements, he argues, was Tanu's monopolisation of territorial politics.

It is interesting to reflect on the South African case in the light of this perspective. Evidently, it is clear that the multiple popular organisations which formed the basis of the opposition in the urban areas entered political society systematically in the 1980s, and very much as a result of pressure from below. Through the `umbrella' of the UDF, they also occupied the terrain of territorial politics which was at the time not distinct from democratic politics. An important characteristic of this movement was the communication between the national leadership and the rank and file, especially while the former were reasonably accountable. Nevertheless, in this connection, a number of different and opposite trends and practices -- democratic and authoritarian, consultative and coercive, analytical and celebratory, accountability of the leadership and independence of dominant (often charismatic) figures -- were all evident simultaneously within the movement.

At the same time, there was no distinction between the `political movement' and the `social movement' as the latter put itself squarely within the ideological ambit of the `Congress Tradition'. Clearly, while the ANC was physically removed from the people and the struggle, divisions between the two were not evident. Distinctions between the two movements probably started developing in an obvious way as the `internal leadership' would regularly visit the capitals of the frontline states for `consultations' with the ANC. At this stage the latter shared the territorial space with the internal organisations and the South African state. More and more, however, the ANC and the apartheid state started to monopolise this political space and internal leaders began to deal with the regime as part of ANC delegations. This process took off after 1990, although the journalist Allistair Sparks shows quite clearly that the ANC and the apartheid state had been engaged in secret talks long before that (Sparks 1995).

In general, the politicised popular movement of the 1980s was not incorporated into state politics against its will. What seems to have been the case is first that the popular control over this movement was severely weakened by the second state of emergency. By the time the organisations were emerging from this extreme repression in the late 1980s, they had much more a character of being led `from above'. It was, therefore, easier for them to disband altogether, to transform themselves into adjuncts of the ANC (a procedure which severely weakened the women's movement and killed off the youth movement altogether -- `the young lions were tamed', to use the jargon) or to retreat from politics into `interest groups'. The voluntary nature of the ceding of the territorial political space to the party of state nationalism -- the ANC, in South Africa is therefore interesting to note. It is this distancing from politics which then makes it possible for these organisations, civics and especially trade unions, to re-enter politics within the realm of the state as corporatism becomes more entrenched, and politics is reduced to state politics.


In this essay, I have argued, following recent writings by African intellectuals on popular movements, that it is impossible to provide an adequate explanation of the transition period in South Africa, and to do justice to the experience of the social movements in that country without distinguishing popular-democratic nationalism from state nationalism. The former, I have suggested, was the dominant trend in the liberation politics of the 1980s, while the latter is in the position of consolidating itself in the 1990s. In the 1980s, the dominant trend of popular nationalism corresponded to the politicisation of civil society and the democratisation of the state from below, while the 1990s are witnessing the statisation of civil society or its politicisation from above. Central to the distinction between the two political and ideological forces of popular and state nationalism were different conceptions of the relationship between leaders and led. While the former stressed popular democracy and control, accountability and direct mandating of leaders, the latter stressed the independence of leadership, top-down prescriptions and statist arguments of various kinds. The latter conception of politics in fact reduced politics to state politics so that, in the words of Wamba-dia-Wamba (1994:250), in his discussion of post-colonial Africa: `political consciousness and state consciousness tended to be identical [so that -- MN] the state tended, then, to be internalised by those fighting or resisting it. This facilitated the emergence of `territorial nationalism' as the `foundation of post-colonial politics'. South Africa has been no exception to this general African trend. The best the state-nationalist position has been able to produce has been a negotiated compromise between different statist forces expressed in multipartyism and various other features of state-controlled democracy, which defeated the experiment of emancipative politics begun under `people's power'. None of the popular experiments of democracy seems to have found its way, so far, into the constitution of the country, and seem unlikely to do so.

I have stressed the importance, therefore, of distinguishing between horizontal divisions between democratic and authoritarian tendencies and ideologies within the nationalist movement, rather than the usual vertical divisions between socialists, nationalists or whatever. This distinction might also help to make sense of what often seem to be the contradictory positions espoused by activists, such as those pointed out in the survey of shop stewards in 1991, which discovered, as we have seen, that a massive majority supported the ANC and `workers' control' (including socialisation of the means of production and state control over industry). This combination seemed a priori contradictory as it suggested support for nationalism (the ANC never claimed to be socialist) and some form of `socialism' simultaneously; it seemed to suggest a combination of 'nationalist' and `socialist' `consciousness', which are often considered mutually exclusive (or at least distinct). In addition, it cannot be argued that such a combination implied support for an SACP position which argued for short-term support for the ANC, and long-term support for socialism (even though the `length of the term' was never elucidated); it seems rather that shop stewards supported `workers' control' as an immediate concern (in any case only a minority mentioned their support for the SACP).

What seems contradictory, if we remain at the level of vertical ideological distinctions, becomes somewhat clarified if we move to considering horizontal distinctions. There is nothing particularly contradictory in holding that `liberation' (`The people shall govern') is equivalent to `people's power' or `workers' control'. After all, this was the actual experience of a large number of people in both townships and unions in the 1980s, and it was the dominant view of activists within civics and trade unions during the same period. Liberation was equated for many with national popular democracy (while this popular democracy was referred to as `socialism' by some, and in the unions could more accurately be referred to as `syndicalism'), and there is nothing contradictory in doing so; in actual fact, the rigid opposition between 'nationalism' and `socialism' has arguably been the overwhelming concern of South African intellectuals. Such an ideological distinction (especially but not exclusively) in Africa has generally contrasted two marginally different forms of statism in which what Wamba-dia-Wamba (1993;1994) has called an `emancipative mode of politics' has been precluded. In other words, whether the post-colonial state in Africa called itself 'nationalist' or `socialist' was of marginal significance to the majority of the people who were, in either case, equally the subjects of intense oppression. The arguments of South African intellectuals, as I shall show at length in future work, were no exception to this general statist trend. On the other hand, the kinds of politics which were developing within the mass movement in South Africa were stressing, (although obviously not exclusively) a distinction between popular and state politics, thus providing some of the fundamental elements of a genuinely emancipatory politics. Thus both 'nationalist' and `socialist' oppositional discourses comprised statist as well as popular/emancipatory political trends. The latter form of politics can be understood as conforming in all essentials to democratic politics as formulated by Wamba-dia-Wamba (1995:16; 1994:259--260):

Democracy is society instituting itself against its own past traditions [of politics -- MN] for better ones. It is a break with the submissive consciousness or culture sustaining past traditions. This means that emancipatory politics emerging in civil society (grass-roots based), should be the originating site of the political prescriptions on the state leading to real state reforms consistent with the rising social demands. ... This [emancipatory -- MN] consciousness emerges and develops through the active participation in the development and treatment of matters political. When this participation is stopped, the consciousness is replaced by the internalization of the state perspective and the ensuing self-censorship. The party in its present form, cannot enhance this development; nor can a multi-partyism which reduces politics to a matter of numbers.

It was precisely such a break with submissive consciousness that South African working people were attempting to institute in urban townships in particular during the 1980s, and precisely its replacement by self-censorship which seems to have permeated popular consciousness in the 1990s. In the experience of many ordinary people, national liberation in the 1980s, came to be equated with such a popular democratic project of emancipation as, in the words of Murphy Morobe (1987:84):

the key to a democratic system lies in being able to say that the people in our country can not only vote for a representative of their choice, but also feel that they have some direct control over where and how they live, eat, sleep and work, how they get to work, how they and their children are educated, what the content of that education is; and that these things are not done for them by the government of the day, but [by] the people themselves. In other words, we are talking about ... mass participation rather than passive docility and ignorance, a momentum where ordinary people feel that they can do the job themselves, rather than waiting for their local MP to intercede on their behalf.

How can the defeat of this popular project for emancipatory politics and its replacement by state politics then be explained? Clearly, many factors were responsible and a full history can only be the outcome of future research. Yet four main factors seem to be thrown up by the above discussion. State inspired violence was a first crucial factor. This was successful in weakening popular structures, not only because of the lack of a developed underground (including military) with structures subjected to the popular movement, but also because of often bureaucratic practices within the movement itself. Although such authoritarian practices had been contested, especially under the `people's power' process, they were never successfully supplanted by popular democratic practices and control. Second, such undemocratic practices became more prevalent and dominant as popular structures disintegrated, while the ANC and its leadership, who were to negotiate `liberation' in particular, were completely unaccountable to any popular structures within the country. Third, popular organisations in both townships and trade unions vacated the realm of popular politics in the 1990s (although there were signs that this had been vacated de facto by 1988--89) and withdrew to a supposed `apolitical' civil society where they became `interest groups' with concerns limited by the division of labour. Fourth, the monopoly of territorial politics was thus acquired by the unaccountable party of state nationalism, initially as it became the only legitimate interlocutor of the moribund apartheid state and international forces, then as state structures were left untransformed and untouched by popular culture, and finally as popular forces, accepting its organisational `leadership', ceded the plane of territorial politics to it.

The outcome of the process was a negotiated series of compromises between the outgoing political forces representing overwhelmingly white capital, and the incoming new black petty bourgeoisie eager to appropriate and accumulate and/or to fill state positions with minimal control from below. Because the struggle over the content and form of the new state has, in the 1990s, only very marginally involved democratic popular forces and has largely been conducted over their heads, the prospects of statism dominating in South Africa, albeit within a multiparty context are overwhelming.


Adelzadeh, A & Padayashee, V 1994. The RDP White Paper: reconstruction of a development vision. Transformation, 25.

African National Congress 1992. Negotiations: a strategic perspective. The African Communist, 131 (special issue), 4th quarter.

Astrow, A 1983. Zimbabwe, a revolution that lost its way? London: Zed.

Atkinson, D (ed) 1992. The state and civil society, Theoria 79, (special issue) May.

Barrell, H 1990. MK, the ANC's Armed Struggle. London: Penguin.

Baskin, J 1982. Growth of a new worker organ: the Germiston shop stewards' Council. South African Labour Bulletin, 7(6) April.

Baskin, J 1991. Striking back: a history of COSATU. Johannesburg: Ravan.

Baskin, J 1993. Corporatism: Some obstacles facing the South African Labour Movement, Centre for Policy Studies, Johannesburg, Research Report 30, April.

Bird, A & Shreiner, G 1992. COSATU at the Crossroads: towards tripartite corporatism or democratic socialism. South African Labour Bulletin, 16(6).

Botha, T 1992. Civic associations as autonomous organs of grassroots' participation. In Atkinson, D (ed) (1992).

Bundy, C 1987. Street sociology and pavement politics: aspects of youth and student resistance in Cape Town, 1985. Journal of Southern African Studies, 13(3), April.

Carter, C 1991. `We are the Progressives: Alexandra Youth Congress Activists and the Freedom Charter, 1983--1985. Journal of Southern African Studies, 17(2), June.

Carter, C 1992. Community and conflict: the Alexandra Rebellion of 1986. Journal of Southern African Studies, 18(1).

Chaskalson, M, Jochelson, K & Seekings, J 1987. Rent boycotts, the state and the transformation of the urban political economy in South Africa. Review of African Political Economy, 40, December.

Collinge, J--A 1991. Civics: Local Government from Below. Work in Progress, 74, May.

Collins, D 1994. Worker Control. South African Labour Bulletin, 18(3), July.

Cronin, J 1992. The boat, the tap and the Leipzig way. African Communist, 3rd Quarter.

De Villiers, R 1986. UDF: front or political party. Work in Progress, 40.

Fine, B 1993. Civil society theory and the politics of transition in South Africa. Review of African Political Economy, 56.

Foster, J 1982. The Worker's Struggle: Where does FOSATU Stand? Review of African Political Economy, 24, May--August.

Friedman, S 1991. An unlikely Utopia: state and civil society in South Africa. Politikon, 19(1) December.

Friedman, S 1992. Bonaparte at the barricades: the colonisation of Civil Society. In Atkinson, D (ed) 1992.

Friedman, S & Atkinson, D (eds) (1994).The small miracle: South Africa's negotiated settlement. Johannesburg: Ravan.

Gibbon, P 1993. `Civil Society' and political change, with special reference to developmentalist states. Mimeo: Scandinavian Institute of African Studies.

Gibbon, P 1994. Some reflections on state, civil society and the division of labour in late colonial Tanganyika (some notes on the origins of the one-party state). Paper presented at the conference on `Dimensions of Economic and Political Reform in Contemporary Africa', Fairway Hotel, Kampala, April.

Haysom, N 1986. Mabangalala: the rise of right wing vigilantes in South Africa. London: CIIR.

Hughes, H 1987. Violence in Inanda, August 1985. Journal of Southern African Studies 13(3), April.

Jacobs, B 1992. Heading for disaster. Work in Progress 86, December.

Jochelson, K 1990. Reform, repression and resistance in South Africa: a case study of Alexandra Township, 1979--1989. Journal of Southern African Studies. 16(1).

Kasrils, R & Khuswayo, M 1991. Mass struggle is the key. Work in Progress 72, January/February.

Keet, D 1992. Shop stewards and worker control. South African Labour Bulletin 16(5), May/June.

Kriger, N 1992. Zimbabwe's guerrilla war: peasant voices. Cambridge: CUP.

Lambert, R & Webster, E 1988. The re-emergence of political unionism in contemporary South Africa. In Cobbett, W & R Cohen (eds) (1988).

Lanegran, K 1995. South Africa's Civic Association Movement: ANC's ally or society's watchdog? Shifting social movement-political party relations. African Studies Review, 38(2), September.

Maree, J 1993. Trade unions and corporatism in South Africa. Transformation, 21.

Levin, R & Solomon, I 1994. National liberation and village level organisation and resistance: a Central Lowveld case study. In R Levin, & D Weiner, (eds) Community perspectives on land and agrarian reform in South Africa. Project Report for MacArthur Foundation.

Lodge, T 1992. The African National Congress in the 1990s. In G Moss & I Obery (eds) South African Review 6. Johannesburg: Ravan.

Lodge, T et al 1991. All here and now: black politics in South Africa in the 1980s. Cape Town: Ford Foundation, David Phillip.

Mamdani, M 1990. State and civil society in contemporary Africa: reconceptualising the birth of state nationalism and the defeat of popular movements. Africa Development, 15(3/4).

Mamdani, M 1994. Non-racial apartheid: the making of citizen and subject in Contemporary Africa, Mimeo.

Mamdani, M & Wamba-dia-Wamba, E (eds) 1995. African studies in social movements and democracy. Dakar: CODESRIA.

Marie, B 1992. COSATU faces crisis: quick fix methods and organisational contradictions. South African Labour Bulletin, 16(5), May/June.

Marx, A W 1992. Lessons of struggle: South African internal opposition, 1960--1990. Cape Town: Oxford.

Mathiane, N 1986. The strange feeling of taking control. Frontline, 6(7), Xmas.

Matiwana, M et al 1989. The struggle for democracy: a study of community organisations in Greater Cape Town from the 1960s to 1988. Centre for Adult and Continuing Education, UWC.

Mayekiso, M 1992a. Working class civil society: why we need it and how we get it. African Communist, 2nd quarter.

Mayekiso, M 1992b. Hands off the civics and civil society! A response to Blade Nzimande. Work in Progress 21, April.

Mdluli, D 1991. Uniting civics across the Eastern Transvaal. Work in Progress, 74, May.

Molamu, L & Fako, T 1993. Violence in the political culture of contemporary South Africa: lessons for SADC. Paper presented to the SADRA workshop on `SADC After Ten Years'. NUL, Roma, October.

Morobe, M 1987. Towards a people's democracy: the UDF View. Review of African Political Economy, 40, December.

Murray, M J 1987. South Africa: time of agony, time of destiny, the upsurge of popular protest. London: Verso.

Naidoo, J 1986. Building people's power: a working class perspective, Speech at Grassroots Conference, 5 April 1986.

Naidoo, K 1992. The politics of youth resistance in the 1980s: the dilemmas of a differentiated Durban. Journal of Southern African Studies, 18(1).

Ngcobo, E 1993. Local Government Negotiating Forum off to a false start. Work in Progress Reconstruct Supplement, 93, November.

Nzimande, B & Sikhosana, M 1991. Civics are part of the national democratic revolution. Mayibuye, 2(5), June.

Nzimande, B & Sikhosana, M 1992a. Civil society and democracy. African Communist, First Quarter.

Nzimande, B & Sikhosana, M 1992b. Civil society and democracy: a rejoinder. African Communist, Third Quarter.

Nzimande, B & Sikhosana, M 1992c. `Civil society' does not equal democracy. Work in Progress, 84, September.

Nzongola-Ntalaja, G 1993. Nation Building and State Building in Africa. SAPES Occasional Paper No 3 , Harare.

Olukoshi, A O & Laakso, L (eds) (1996). Challenges to the nation state in Africa. Nordic African Institute and IDS, University of Helsinki.

Pieterse, E A & Simone, A M 1994. Governance and development: a critical analysis of community-based organisations in the Western Cape. FCR.

Pityana, S M & Orkin, M 1992. Beyond the factory floor: a survey of COSATU shop-stewards. Johannesburg: Ravan.

Sapire, H 1992. Politics and protest in shack settlements of the Pretoria-Witwatersrand-Vereeniging Region, South Africa, 1980--1990. Journal of Southern African Studies, 18(3) September.

Schreiner, G 1994. Beyond corporatism: towards new forms of public policy formulation in South Africa. Transformation, 23.

Seekings, J 1988. The origins of political mobilisation in PWV Townships, 1980--84. In W Cobbett, & R Cohen (eds) 1988.

Seekings, J 1989. People's courts and popular politics. In G Moss, & I Obery (eds) South African Review 5, Johannesburg: Ravan.

Seekings, J 1992a. `Trailing behind the masses': the United Democratic Front and township politics in the Pretoria-Witwatersrand-Vaal Region, 1983--84. Journal of Southern African Studies, 18(1), March.

Seekings, J 1992b. From quiescence to people's power: township politics in Kagiso, 1985--1986. Social Dynamics, 18(1).

Seekings, J 1992c. The United Democratic Front and the changing politics of opposition in Natal, 1983--85. Paper presented at the Ethnicity Conference, University of Natal Pietermaritzburg, September.

Seekings, J 1993. Heroes or villains: youth politics in the 1980s. Johannesburg: Ravan.

Shaw, T 1994. South Africa: the corporatist/regionalist conjuncture. Third World Quarterly, 15(2).

Shilowa, S 1995. Challenges of the transition period: a COSATU perspective. African Communist, 139/140, 1st Quarter.

Sisulu, Z 1986. People's education for people's power. Transformation, 1.

Sitas, A 1992. The making of the Comrades movement in Natal, 1985--91. Journal of Southern African Studies, 18(3), September.

Slovo, J 1992. Negotiations: what room for compromise? African Communist, 130, 3rd Quarter.

South African Institute of Race Relations 1989. Race Relations Survey, 1988/89, Johannesburg: SAIRR.

Southall, R 1994. The South African Elections of 1994: the remaking of a dominant-party state, Journal of Modern African Studies, 32(4), December.

Sparks, A 1995. Tomorrow is another country: the inside story of South Africa's negotiated revolution. Sandton: Struik.

Swilling, M 1984. Workers divided: a critical assessment of the split in MAWU on the East Rand. South African Labour Bulletin, 10(1), August/September.

Swilling, M 1988. The United Democratic Front and township revolt. In W Cobbett, & R Cohen, (eds) Popular struggles in South Africa today, London: James Currey.

Swilling, M 1990. Political transition, development and the role of civil society. Africa Insight, 20(3).

Swilling, M 1991. Socialism, democracy and civil society: the case for associational socialism. Work in Progress, 76, August. Also published in Atkinson, D (ed) 1992.United Democratic Front (1985, 1986, 1987)

United Democratic Front, Cape Town Area Committee (1986) `Broadening the Base', Discussion Document.

Wamb-dia-Wamba, E 1993. Democracy, multipartyism and emancipative politics in Africa: the case of Zaire. Africa Development, 18(4).

Wamba-dia-Wamba, E 1994. Africa in search of a new mode of politics. In H Himmelstrand (ed) African perspectives on development. London: James Currey.

Wamba-dia-Wamba, E 1995. The state of all Rwandese: political prescriptions and disasters. Paper presented at the Arusha conference on the Great Lakes Region, September.

Webster, E 1987. The rise of social movement unionism: the two faces of the black trade union movement in South Africa. In P Frankel et al. State, resistance and change in South Africa. Johannesburg: Southern.

White, R 1986. A tide has risen. A breach has occurred: towards an assessment of the strategic value of the consumer boycotts. South African Labour Bulletin, 11(5), April-May.

Newspapers and Periodicals
South African Labour Bulletin
New Nation
Sunday Times
New Era
Weekly Mail/Mail and Guardian