The historiography of agrarian change has, in Southern Africa, been a relatively recent phenomenon dating from the early 1970's. The decade of the 1970's witnessed a veritable upsurge of very stimulating historical studies which took as their fundamental point of reference, the development of capitalism in rural relations under colonialism. The writings of Arrighi (1973), Arrighi and Saul (1973) and Van Onselen (1976) on colonial Rhodesia, those of Trapido (1971), Bundy (1972, 1979) and Morris (1976) on South Africa, those of Palmer and Parsons (1977) and Kowett (1978) on the Southern African region all dealt, in different though stimulating ways, with a process of capitalist development which directly affected rural producers. This body of historical literature, which was overwhelmingly of a radical `Left' persuasion, has been substantially added to in the decade of the 1980's especially in South Africa where it has come to be considered as a specific school of thought. The writings of Beinart (1982), Keegan (1986) and Beinart et al. (1986) spring particularly to mind in this context.
This historical literature was supplemented and underpinned by more theoretical (e.g. Wolpe, 1972, 1975) and political-economic studies (e.g. first, 1983) which in combination, have produced an overall dominant paradigm, which has come to constitute the conventional view of capitalist rural transformation on the Left in Southern Africa.2 This paradigm -which I outline below - I have called elsewhere the position of `radical political economy' (Neocosmos, 1987a; Levin and Neocosmos, 1989). It has coloured, to a greater or lesser extent, the overall view which anyone on the Left holds of the social formations in Southern Africa toady, including our view of the classes at play in such formations, and the nature of the state. As a consequence it has affected our political practice as intellectuals.
Theory is quite clearly not politically neutral. The contents of the present paper are submitted, not simply as an argument for a different kind of history, but also as a plea for a different political practice by left intellectuals in the region. One of the many positive attributes of African historiography (as opposed to the dominant historiography in the West) is its lack of fear of theory (as I experienced at first hand during a fruitful short stay at UDSM in 1981). I feel therefore justified in treating what is at heart a theoretical question in a conference of professional historians.
The overwhelming majority of historical writings on the region, some of which have been referred to above, operate within the dominant paradigm of `radical political economy'. This position maintained inter alia that the peasantry had, on the whole, been systematically proletrianised during the colonial period. This process had taken place primarily in order to service the mining industry in South africa and Rhodesia and was a direct effect of state intervention in the economy (e.g. in South Africa through the Land acts of 1913 and 1936). Overall proletarianisation, it was held, was the direct result of a number of repressive colonial measures such as land alienation and the creation of reserves, taxation, forced labour and so on. It was conceded that even if the peasantry had not been fully proletarianised, the trend towards proletarianisation was abundantly clear.
This process led, over time, to the countries of Southern Africa being `dependent' on the South African economy in particular. The principal aspect of this `dependency' was said to reside in the provision of labour power for South African capitalism. Thus, the countries in the region were said to be `labour reserves' or `labour producing economies', the South African bantustans were seen as compounds for a `reserve army of labour' and so on.3 This analysis was said to hold, with appropriate variations, for the then colonial countries which had achieved independence in the late 1960's as well as for the South African bantustans themselves.
Clearly, this is only a schematic outline. It does not and cannot do full justice to the sophisticated nature of the discussions in many of the texts noted above. My intention however is merely to outline the general orientation of these writings, which was overwhelmingly one of the destruction (full or partial) of peasant producers. These arguments are well known and need not be elaborated in detail. Rather I want, in the rest of this paper, to discuss some of the most important consequences of this argument. I wish to argue in particular:
(a) that the thesis of the proletarianisation of the Southern African Peasantry (the `linear proletarianisation thesis') fails to explain the differentiation of the peasantry into classes (or potential classes) and importantly fails to provide a correct account of the class physiognomy of the social formations in the region. At the same time it fails to explain the continued production of agricultural petty-commodity producers under imperialist conditions;
(b) that as a result,this argument fails to provide a correct account of the nature of the post-colonial state in the countries of the region;
(c) that this view tends to see the only two alternatives facing the oppressed people as capitalism and socialism and hence that it fails to address the question of democracy, in particular the importance of popular struggles for democracy for a socialist future;
(d) that it misunderstands the nature of imperialism, in particular what might be termed the social relations of imperialism in Southern Africa
(e) that it cannot offer a solution to the national question in South Africa in particular, as it fails to comprehend the division of the oppressed into nationalities which is founded on a basis of petty-commodity production under imperialist conditions.
I will deal with the first three questions (a, b and c) in this section of the paper. The next question (d) will be dealt with in section three; I shall then end with a brief discussion of the link between the national and agrarian questions in South Africa (e) in section four of this paper.
One of the effects of the position of `radical political economy' in the academic sphere of activity at least was a proliferation of migration studies in the Southern African region, particularly in so far as the mining industry was concerned. The political economy of peasant production was systematically ignored or, at best, relegated to a subsidiary position despite the obvious fact that it provided the source of the migration process and that this position itself talked in terms of a semi-proletariat. This was the result of what might be termed a kind of fetishism of the working class (what in South Africa is sometimes referred to as `workerism') whereby an ideal future of a preponderance of proletarians is seen as already in existence or at least `just around the corner'.
This wishful thinking had a number of consequences, two of which are worth mentioning in this context. The first was that, as the radicals deserted the field of agrarian relations in Southern Africa (apart from considering rural areas negatively in relation to capitalism as `traditional' or dominated by a largely unspecified `pre-capitalist' mode, as in Wolpe, 1975) it was left wide open for the ideologists of the state and imperialism to take over and monopolise with their superior resources. The second consequence was the fact that when an upsurge of peasant production occurred after independence, as it did in Zimbabwe for example, it was not possible to explain its origins. Thus what has been referred to as the `peasant miracle' in Zimbabwe where peasants have produced a regular crop surplus in most years since independence, dealt a heavy blow to the linear proletarianisation thesis (Phimister, 1986). Although it is becoming more and more evident that this surplus is being produced by a small stratum of the peasantry (both in regional and in class terms), it is obviously leading to a re-examination of the assumptions of the linear proletarianisation thesis, in Zimbabwe at least.4 Evidently,such a successful rich peasantry does not arise from nowhere and certainly not from a supposedly homogeneously immiserated semi-proletariat.
It is ironic that Zimbabwean history should provide the first clear example in practice in Southern Africa, of the fallacy of the linear proletarianisation thesis. After all it had been the radical studies of Rhodesia by Arrighi (1973) and Arrighi and Saul (1973) which had provided, through a systematic critique of colonialist derived ideology, the seminal studies of proletarianisation in Southern Africa which were to develop into the dominant trend we still know today. It must be recalled that they argued that:
unlike the situation elsewhere on the continent, the African peasantry here (South Africa and Rhodesia- M.N.) has been effectively proletarianised in the sense that the balance between means of production outside the capitalist sector (mainly land) and the subsistence requirements of the African population has been severely and irreversibly upset (Arrighi and Saul, op.cit.:64)
From this argument Arrighi and Saul derived two important political conclusions which I shall have occasion to refer to below. For the present it is only important to note the equation which they make between the proletarianised peasantry and the African population, thus obviously conflating class and racial divisions. It must be stressed that this view is not peculiar to Arrighi and Saul. Rather, it is a necessary consequence of seeing the overwhelming majority of the population of Southern Africa (whether in the colonial or post-colonial periods) as proletarians or proletarians in the making. It is adhered to by most writers from within the `radical political economy' perspective. Thus, a recent text on Swaziland maintained that:
The majority of the population living in the Swazi Nation areas are conventionally described as `peasants'. However... over 80% of such households are dependent for their existence on the regular sale of the labour power of one or more homestead members in the wage sector at any given moment. This very high incidence of continuous migratory labour outside of these areas... suggests that the process of proletarianisation of such rural producers is very far advanced. Thus, far from constituting a `peasantry', the large majority of the population of these areas are a semi-proletariat who make up a reserve army of labour for South african capital. Increasingly, they are coming to constitute part of the absolute surplus population which the development of capitalism in Southern Africa has scattered across the region (Davies et. al. 1985: 37)
In an even more recent text which deals with the Southern African region as a whole, Mafeje, for example, maintains that:
extensive labour migration is an index of poverty among rural Africans and further that the reserves were not meant to, and could not have been meant to guarantee the subsistence of the African labour whose services were needed elsewhere. They were only meant to supplement at a very low level the starving wages that the Africans were receiving from their exploitative employers. It is therefore logical nonsense to have expected Africans to develop or even sustain subsistence agriculture in the Reserves (Mafeje, 1988: 101-2).
And yet, the incredible thing is that some Africans have succeeded from their position in the reserves not only to sustain subsistence agriculture, but also to accumulate as the Zimbabwean `miracle' and other evidence from the region shows (as we shall see below). The masses of the people are obviously more resourceful than intellectuals sometimes allow.
The major problem with Mafeje's position, which his remark makes clear, is that `radical political economy', like dependency theory of which it is a variant, gives such an overwhelming power to the structures of imperialism, capital and the state, that the African people are powerless to affect their own destinies in any meaningful way. The result is that history lies beyond the control of the masses (Bernstein and Nicholas, 1987a; Kitching, 1985).
I have shown at length elsewhere (Neocosmos, 1987a; Levin and Neocosmos, 1989) how this view of an overwhelmingly proletarianised population implies a homogenous view of the population of the countries of Southern Africa. This is of course because no class differences are usually associated with a proletariat or proletariat in the making. Indeed the argument of proletarianisation is usually justified - as in one of the quotations above - with reference to aggregate data. One of the consequences of this purported homogeneity of the population is the fact that it becomes impossible to account for the existence of an indigenous ruling class and its state other than by reference to extra-social or extra-territorial forces.
This has led to considering the ruling class as a `clique' or `stooge' of imperialism,much as under colonialism itself, thus denying the importance of explaining the basis of the ruling class and of the reproduction of state power within the context of national social relations. Mafeje (op.cit.) is an admirable example of this kind of reasoning. He refers to the ruling classes of Southern Africa as a "petit-bourgeois elite" (pp.97, 106), a "governing elite" (p.106), or a "ruling elite" (p.121). The term `elite' is, of course, not ideologically neutral as Poulantzas noted long ago in his debate with Miliband on the state (Poulantzas, 1972). The concept refers to a psychological attribute of individuals or organisations theorised by bourgeois sociologists such as Weber and Michels. In other words it refers to an attribute which is given by phenomena beyond the social relations in which people live and exist. Thus these ruling classes are not accounted for by Mafeje in terms of the contradictory material social relations of the societies of Southern Africa - they cannot be as the population is theorised as overwhelmingly homogenous - rather they are parachuted from `outer space' so to speak on an unsuspecting population. Their `extra terrestrial' nature means that they cannot be removed by collective agency.
In addition, Mafeje does not use the term `petty-bourgeois' in its scientific sense - as a contradictory combination of bourgeois and proletarian class positions and practices5 -but rather, as is common amongst African intellectuals, as a term of abuse:
While colonialism is responsible for these distortions (in the agricultural economy - M.N.), African governments stand indicted for having pursued the same policies after independence and for having, thus, instituted neo-colonialism in their countries. During their tenure things got worse because, unlike the colonial administrators,they lacked self-discipline and a sense of organisation. Above all, they were too eager to accumulate material wealth before they had learnt how to produce it. Their petit-bourgeois greed or crass materialism meant not only wanton waste of resources but also super-exploitation of the direct producers. This had to be the case because, unlike the bourgeoisie in Europe, they had neither alternative sources of wealth nor a viable economic and political project. Theirs became a pathetic example of fishing by emptying the pond (ibid.: 96-7).
Rhetoric aside, Mafeje clearly uses the term `petty-bourgeois' to refer to a kind of pseudo-bourgeoisie which lacks the supposed attributes of a fully-fledged bourgeoisie. One has to be a fanatic believer in Weber's `protestant ethic' to hold that the European bourgeoisie, like bourgeoisies everywhere, did not exhibit "greed" or "wanton waste of resources". Anyhow, the important point is that bourgeoisies do not act as wilful subjects affected by psychological motives, but within the context of social relations imposed beyond their will. Mafeje is forced into employing such extra-social or extra-historical determinants in order to account for the existence and characteristics of African ruling classes because of his underlying assumption of an overall homogeneous population bereft of historically determined socio-economic differences.
Insofar as explanations of the post-colonial state are concerned, similar problems are encountered by `radical political economy'. I have noted elsewhere (Neocosmos, 1987b: 102-3) how in Swaziland for example, the view of Swazi society as essentially lacking in any visible social contradictions, implies that an account for the existence of the state has to be found outside the country itself. The state thus ends up being seen as an external imposition on Swazi society much as under colonialism itself, so that the fact of political independence is, to all intents and purposes denied, and neo-colonialism is equated with colonialism.6 The notion of a `comprador state' as used by Daniel (1985) inter alia, implies that it is the `dependent' international division of labour which, of necessity, leads to the identity of interests between the state and imperialist capital. Under such conditions it evidently becomes impossible for the masses of the oppressed to change their own situation. Their oppression is given by forces beyond their control.
Two main points need to be made in relation to a notion of the `comprador state' (or `comprador bourgeoisie'). The first, rather provocatively, is that there is no such thing as a `dependent economy', there is only dependent forms of capital accumulation. What do I mean by this? I mean principally that the fact of dependency must be understood not as a given state of affairs, but as a necessity for capital accumulation under the leadership of the bourgeoisie. It is only the different kinds of bourgeoisie which, when in power as ruling classes, are forced to compromise with reaction (imperialism), both in order to accumulate and to repress the masses of the people thus restricting democracy. We may say that the bourgeoisie when in power is forced to take the `dependent road' to capital accumulation. There is no strict overriding necessity for a people's state in a period of National Democratic Revolution7 to take such a road as was shown in China in the 1940's and 1950's for example.
Second, surely the extent and form of `compradorisation' by the state both depend on the class which rules. These will be different under the rule of a petty-bourgeoisie (e.g. as in Tanzania in the 1970's or in Mozambique), different under the rule of a national bourgeoisie8 (e.g. as in Zambia and possibly Zimbabwe), it will be different again under a clearly pro-imperialist ruling class (e.g. as in Swaziland, Lesotho or Malawi).
In any case, the important point is that both these remarks imply a necessity to investigate the class character of the state and the reproduction of state power through an analysis of the changing political economy and structure of class relations pertaining in the social formation in question. I have attempted to provide the beginnings of an explanation of the form which the Swazi state takes precisely in such terms in my (1987b). It should be evident that such an analysis is precluded by the position taken by `radical political economy'.
The next effect of the arguments of `radical political economy' which I wish to discuss is a political one. It was maintained that colonial countries, such as Rhodesia, Angola and Mozambique in the early 1970's, could and would move after liberation straight from colonialism to socialism, as the two main social forces were the proletarian (a semi-proletarian) masses on the one hand, and colonialism/reaction on the other. The importance of the `middle groups' (such as the petty-bourgeoisie which, it should be recalled, was later berated for taking power!) between these two extremes was substantially ignored in favour of a kind of mainchean dualism. While this argument obviously led to a conflation of racial divisions with class as I have already noted, it also led to a much more important political problem. The central issue in any liberation or transformation process, that of democracy (i.e. of the control of the state by the people) was noted addressed, because such a question did not arise for a viewpoint which saw democratic struggles as primarily necessitated by `backward' social relations - i.e. in the present context, relations in which a majority of peasants dominate the social formation.
In the absence of such a predominant peasantry and the presence of a seemingly ideal opposition between capitalists (albeit white and foreign) and an overwhelmingly proletarianised population, the issue of popular democratic struggles was seen as largely irrelevant for the overcoming of dependency. Only socialism and development (state-led of course!) could do that. Hence there developed at independence what Shivji (1985, 1988a) has called the "ideology of developmentalism", an ideological discourse whereby the masses of the people (who were the ones to have fought for independence in the first place) were systematically depoliticised, and the state itself was seen as the agent of `development' and of `socialist transformation'. In addition, as the masses of the people were seen as homogeneously proletarianised, there could be no detailed class analyses from the perspective of `radical political economy', as a prerequisite say to developing a debate on the kinds of democracy which could be pertinent or strategically advisable in such conditions.
This argument regarding the practical equation of liberation/independence and socialism was maintained by academics and also by the overwhelming majority of liberation movements (with the notable exception of the ANC). Thus Arrighi and Saul(op.cit.) derive two conclusions from their argument that the peasantry has been effectively proletarianised in South Africa and Rhodesia. First they argue that there is "little, if any, room for a neo-colonial solution" in these countries, and that in the then Portuguese territories "the neo-colonial solution has been blocked by the `ultra colonialism' of Portugal" (ibid.: 87). This particular statement can only raise a smile today given events in those countries since independence, especially in Zimbabwe. Their second point is that "the revolution in South Africa and Rhodesia, if it is to come, can only be a proletarian and socialist revolution" (ibid.: 65, emphasis added).
At the same time, liberation movements such as FRELIMO (although the MPLA and ZANU also argued in a similar vein) were,through their leadership, talking about `abolishing the exploitation of man by man' as the sacrifices of the people were not undertaken to replace Portuguese exploiters by their Mozambican counterparts, but to abolish exploitation as such. Bourgeois nationalists and other political representatives of rich and middle peasants,whose accumulation had been drastically restricted by colonialism (and who thus held an objective anti-colonial position), were either purged or left FRELIMO which at liberation took on a rather sectarian character and eschewed a broad-based front (Hanlon, 1984; Hermele, 1988: 22-6; Wuyts, 1989; deBraganca and Wallerstein, 1985: 121, 217). No detailed class analysis of the people seems to have emerged from the ranks of the FRELIMO leadership with disastrous consequences (see Hermele, op.cit.; Wuyts, op.cit.).9
After independence, it became more and more obvious (especially in Zimbabwe) that far from developing socialist relations, the newly independent countries were conforming, or were being forced by real material relations to conform, to the neo-colonial pattern which prevails in the rest of Africa obviously contrary to Arrighi and Saul's expectations. It has also become clearer that bourgeoisies and petty-bourgeoisies have emerged as ruling classes since independence; evidently such classes did not emerge from the semi-proletarianised masses.
It must be noted also that this demarcation of classes (and hence clarification of differences) between the new ruling classes and the people, as well as that amongst the people themselves (as evidenced, for example, by the rise of a rich peasantry in Zimbabwe), is one of the most important consequences of political independence which distinguishes neo-colonialism from colonial domination. At the same time, for example, the Mozambican state has had to give greater and greater emphasis in recent years to market relations, given the poor performance of state farms and of the economy in general under war conditions. It has thus had to retreat from its earlier `optimistic' line of immediate socialist transition, to one which takes the existence of a rich and middle peasantry and other sectors of the bourgeoisie and petty-bourgeoisie more seriously (Hermele, op.cit.). In sum, the reality of material social relations has been forcing the identification of classes more and more into the center of debate in the post-independence period.
Before ending this section of the paper, it is pertinent for me to stress what I believe to be the class content of the discourse of `radical political economy'. After all, I would be contradicting my own argument if, having criticised a position for failing to understand class differences, I did not at least make an attempt to elucidate the class content of this position (although this failure is a common one among academics who emphasize class analysis).
The position which I have been discussing tends to see capitalist development in the region in purely destructive terms. While there is no denying the extremely destructive content of capitalist development especially in colonial conditions,capitalist relations are much more contradictory than this position allows, thus enabling the development of a class (or classes) which has a direct stake in the system of oppression as well as many other intermediary classes. This takes place even under the most severe forms of oppression as in South Africa, as I hope to show below. The exclusive concentration on the destructive side of capitalism is linked in `radical political economy' with the view, characteristic of dependency theory as a whole, of the oppressed masses as unable (or incapable) of making their own history precisely because of the overwhelming power of capitalism in its imperialist phase.
The view that change is beyond the control of the masses is common to all bourgeois ideologies, as Marx brilliantly showed in Capital. Hence bourgeois theories always have recourse to extra-historical (or extra-social) phenomena (metaphysics) in order to explain socio-economic reality. Thus `radical political economy' combines a critique of capitalism/imperialism with an adherence to bourgeois forms of thought. This contradictory combination is also to be found in its socialist orientation and its fear of popular democracy; in its wanting to construct socialism under statist conditions; in its vacillation between `revolutionism and its advocation of statist `development'; in its conflation of socialism with national liberation.
These contradictory combinations, in conjunction with its anti-imperialism, clearly define this discourse as an example of petty-bourgeois nationalism. It can be described as petty-bourgeois because it combines, in ideology, the contradictory class positions of the bourgeoisie and the proletariat; or we may say that it combines in ideology the problems and solutions which correspond to the contradictory class location of petty-commodity correspond to the contradictory class location of petty-commodity producers in the economy (Gibbon and Neocosmos, 1985: 192). It is a form of nationalise, principally because of its adherence to an anti-imperialist stance. A bourgeois nationalist position would rarely equate nationalism with socialism for example.
The petty-bourgeois, rather than proletarian, side of this ideology is to be seen also in the fact that it largely homogenises population groups and fails to account for class differences (other than those introduced and imposed from the outside by imperialism itself). Concurrently, it has the tendency to see the ideal future opposition between the bourgeoisie and the proletariat in the present, to "make the wish the father to the thought" as Lenin used to say, rather than analysing objective reality. At the same time this ideology removes the petty-bourgeoisie, rather conveniently it may be added, from its class configuration of society, precisely by seeing the only important classes as imperialism (and its internal stooges) on the one hand, and the proletarianised masses on the other. If a more objective understanding of the political economy of the region is to be developed, this view should be superseded.
It must be noted first of all that it has not been my intention to argue that the proletarianisation of the peasantry under colonialism did not occur, or that it was not a dominant effect of the development of commodity relations under colonial conditions. Rather, my point is that it is theoretically and empirically illegitimate as well as politically disastrous to the historical development of capitalist relations in rural areas. "Facts are stubborn things" as the English are sometimes said to assert, a saying which Lenin was fond of citing. Before turning to an examination of some of this evidence, a few words are necessary regarding the theoretical conceptualisation which I employ.
Two related sets of questions are important here. First what I term the production of petty-commodity production in agriculture as part of the historical development of commoditisation (capitalist relations) in Southern Africa and second, the simultaneous differentiation of petty-commodity producers into classes, or incipient class formations.
It must be understood that it is not a question here of finding evidence for the supposed `persistence' of `traditional' peasant producers in Southern Africa. Such a view does not move us forward in any meaningful way, as it sees proletarians and peasants as polar opposites and the latter as essentially pre-capitalist. This dichotomous opposition provides, after all, the two poles of the linear proletarianisation thesis. I am aware that the view of the contemporary `peasantry' as essentially pre-capitalist is a common conception, but I would argue that such a view is one which restricts itself to a phenomenal understanding of the peasantry and does not ask questions as to its conditions of existence. Such an analysis is based on a Weberian ideal typical methodology, and not on a marxist one (Gibbon and Neocosmos, 1985; Bernstein, 1987).
It was this understanding of the `persistence' of peasant production which ariticulationist theory attempted to capture especially in the writings of Wolpe (e.g. 1975), with its notion of the `preservation-dissolution' of pre-capitalist modes of production derived from Meillassoux's work. As is well known, Wolpe argued in the case of South Africa, that the state slowed down the process of proletarianisation of the peasantry through the reproduction of so-called `pre-capitalist modes' (their partial `preservation' and partial `dissolution' in the interests of capital), in order to enable the super-exploitation of labour through the subsidisation of the latter by the pre-capitalist mode. It is although the proletarianisation process was operating `normally' until some crisis of South African capitalism `required' a political intervention by the state (conditioned by struggles of course) to put the breaks on the `normal' proletarianisation process - hence apartheid. Now, while this argument lends a certain sophistication to the linear proletarianisation thesis, it merely provides a delay in what is seen as a temporary interruption to the ineluctable process of proletarianisation, which presumably is seen as resuming its course once apartheid is finally removed.
It seems to me however that, apart from other criticisms (and then have been many), this argument errs in assuming a radical distinction between capitalism and apartheid (while maintaining that the latter is functional to the former) under an understandable and correct desire to avoid conflaring the two in a spurious notion of `racial capitalism' for example. Apartheid thereby appears as a historical accident, a kind of freak mutation on an otherwise predictable capitalism. The problem arises because of the failure to identify the correct features of capitalism in the present epoch. This seems to me to be the fundamental problem with the linear proletarianisation thesis in general. We are not living simply under capitalist conditions, but since the 1890's we have been living simply under capitalist conditions, but since the 1890's we have been living under monopoly capitalist conditions (imperialism). Monopoly capitalism in Africa especially in Southern Africa, has given rise to specific relations which substantially and systematically modify the phenomenal conditions of capitalism in general (although not its essential relations).
I shall return to a discussion of these relations below, for the present it is important to stress the point that articulationism fails in its attempt to rescue the linear proletarianisation thesis. Rather than talking about the persistence of `pre-capitalist' peasant production in Southern Africa, we should be referring to the production of agricultural petty-commodity producers as an aspect of the development of commodity relations in general. This is both more historically accurate and more theoretically valid. After all can we really describe the producers of the lineage and agro-pastoral societies which dominated in the pre-colonial period as peasants? It seems to me that we cannot without stretching the concept to a point where it becomes valueless. Even if we were to describe such producers as peasants, we would be forced to admit that they are essentially different (if not always phenomenally so) from those which exist today.
One of the important aspects of Bundy's work (1972, 1979) is precisely the fact that he clearly shows the development of peasant production in the eighteenth and nineteenth centuries (the period preceding imperialism) through the dissolution of lineage societies under colonial capitalist impact - a process which has been termed `peasantisation' (see Cohen, 1976). For Bundy, and most historians of Southern Africa, the proletarianisation process follows on the development of the period of imperialism, the development of the mining industry and that of `white' agriculture, principally as a result of state intervention in capitalist relations.
If agricultural petty-commodity production develops as a result of capitalist relations in the region then, though these relations may be somewhat modified by monopoly conditions, there is no theoretical reason why this form of production should cease to exist unless it can be shown in theory that imperialism undermines its conditions of existence. This has not so far been shown and, to my mind, cannot be shown. This is because, as I have argued at length elsewhere (Gibbon and Neocosmos, op.cit.), petty-commodity producers (whether agricultural or not) are founded on a contradictory combination of capital and labour in a single enterprise (individual or household). As long as the essential relations of capitalism - capital and wage-labour-exist (a fact which is not modified under monopoly conditions), then the likelihood is that their combination in petty-commodity production (p.c.p.) will also exist, although perhaps in a modified form as well shall see below.10
Because p.c.p. is an effect of an otherwise not directly experienced relation between capital and wage-labour, it also contains the seeds of its own differentiation. This differentiation need no longer be seen as solely the effect of forces external to such production as is the case in all theories of a supra-historical `peasantry'. In any given situation the `side of capital' can dominate in agriculture p.c.p thus creating `rich peasants', while if the `side of labour' predominates we have `poor' peasants' or semi-proletarians; if both `sides' are `equilibrium' we have a `middle peasantry'. The development of p.c.p. can, in this manner, be understood as paralleling its own differentiation as part of the overall process of commoditisation as Bundy inter alia shows in his work. Some producers always manage to gain and retain control over means of production while others (usually the majority) lose such control. This is a general law of all forms of capitalism.
If we now turn to an examination of the evidence for this dual process since the 1890's, it is perhaps appropriate to begin by referring to Zimbabwean history. In a recent historical reassessment of the linear proletarianisation thesis in Rhodesia Phimister notes of this perio
neither the state offensive nor eviction notice succeeded in transforming all Africans into labour migrants when they were not subsistence cultivators and into subsistence cultivators when they were not migrants. Although administrative measures oppressed all Blacks, their impact was mediated by the process of differentiation and stratification already under way in rural areas (Phimister, 1986:244).
During the early part of the twentieth century, he notes that:
... it was those cultivators covering the spectrum of independent household producers stretching from middle peasants whose family labour and land just sufficed to reproduce themselves, through to rich peasants on the brink of transforming themselves into capitalist farmers, who began to constitute themselves as a class. Although occasionally indistinguishable from their poorer neighbours in terms of standards of consumption, they increasingly came to occupy a quite different place in the social relations of production. Normally possessing large above-average size of herds of cattle and fields, they monopolised communal grazing land at certain times of the year they employed labourers to till their fields and harvest their crops (ibid.: 250, emphasis added).
He thus concludes that "all of this meant that the attacks unleashed by the state and capital after 1907", precisely to provide labour for the mines, "were made on increasingly uneven and broken terrain" (ibid.: 251). The result was, among other things, that:
each district followed its own more or less distinctive path towards proletarianisation not least because the very pace and extent of that process varied according to the local importance of classes and social categories whose collective capacity to absorb the demands of capital and the state differed enormously (ibid.: 252).
I have cited Phimister's recent work at some length, because it constitutes a practically unique account of the historical process of differentiation in imperialist dominated Southern Africa, from the alternative perspective which I have outline in this paper. He shows quite clearly, not only that proletarianisation was not uniform and that differentiation occurred even during the most intense period of state oppression, but also that the activities of the state were themselves conditioned by the social relations pertaining in rural areas. It follows therefore, that the state and capital did not have the overwhelming power ascribed to them by the voluntarism of `radical political economy', and also that the oppressed classes were indeed capable of determining history, even under the most vicious attacks by the state.
In this, Phimister's work on colonial Zimbabwe differs fundamentally from that of the South African agrarian historians, even from those who discuss the post 1948 period (e.g. Morris, 1976). Most of these historians seem preoccupied with debating the precise date of the completion of the proletarianisation process (e.g. Beinart, 1982), or with the `complexity' of social differences including primarily non-class, but also some class differences (e.g. chiefs, headmen, wealthy families,christian families, `traditionalists', `modernizers' and so on, as in Beinart, op.cit. or Beinart and Bundy, 1987). The result of this relativistic abandonment of theory has been the production of a wealth of historical detail but ultimately analytical confusion, as no way is provided to priorities the analytically more important processes from the relatively secondary ones.
As far as Swaziland is concerned, I have suggested in a preliminary way, that there are a number of indications especially from the 1890's onwards which suggest the emergence of petty-commodity producers from the confines of the lineage or `tributary' mode of production, to use the term preferred by Bonner (1983). These indications are founded on the increased importance of agricultural production to the Swazi economy between the 1850's and 1890's and include the introduction of the plough in 1890. Within a context of commoditisation, this had the effect not only of increasing productivity (and presumably differentiation), but also of systematically transforming the gender division of labour in homesteads from one where women dominated agricultural production (in the lineage system), to one where men controlled, through ploughing, the increasingly commoditised produce (Neocosmos, 1987b: 85-88).
This transformation of the division of labour on Swazi homesteads to a more directly `peasant' one and the increased productivity corresponding to it, had the effect of developing cash crop production as Kuper (1963: 43) makes clear. Thus when the attacks of the state designed to force labour to the mines eventually came through taxation and partition, only a minority (although a large minority) of between 25 and 40 percent of the male adult population migrated between 1911 and 1936 Booth, 198: 11). By this period, the lineage mode had been systematically undermined, as men could buy cattle through their access to cash. Moreover, the creation of the reserves through the land partition of 1907, did not have the sole consequence of proletarianisation. It also had the effect, through limiting the land available, of increasing sedentarisation, of undermining shifting agriculture and of restricting patterns of settlement which are all required by lineage forms of based on agro-pastoralism. It thus encouraged a process of individualisation which is the hallmark of petty-commodity production.
The creation of the reserves also had the crucially important effect of providing a new basis for the chieftaincy to control the population which it had previously controlled through bridewealth (cattle). Now control over land became the basis of the power of the chieftaincy and remains so to this day. This power could be exercised effectively only because land was so central to agricultural p.c.p.. Thus the kind of state which eventually emerged in Swaziland at independence, owes its origins to a process of production of capitalist relations on the land during the colonial period. This goes some way towards explaining the `traditional' form of the state in that country along with its capitalist content (Neocosmos, 1987b: 92ff). This `tradition' is in no way a simple left-over from the past, but a product of the social relations of colonial capitalism.11 The importance of land in this respect cannot be over-emphasised as its control enabled (and still enables) the state to exercise direct control over the majority of its population, the agricultural petty commodity producers. This is shown in particular during the colonial period (and the immediate post-colonial period) through the successful struggles waged by the chieftaincy against the liberalising efforts of colonial administrators and advisers which intended to water down the power of chiefs by opening up access to land (ibid.: 92-102). During the post-colonial period it is shown in particular during the colonial period (and the immediate post-colonial period) through the successful struggles waged by the chieftaincy against the liberalising efforts of colonial administrators and advisers which intended to water down the power of chiefs by opening up access to land (ibid.: 92-102). During the post-colonial period it is shown through the buying up of land by the state and its transfer to Swazi Nation Land.12
In sum, in Swaziland during the colonial period, agricultural p.c.p. was developing in tandem with proletarianisation, which is not surprising as they are both aspects of an overall process of commoditisation. Along with these processes there developed, during the same period, a specific type of state which came to be `traditional' in form although capitalist in content. The extent of differentiation during this period has yet to be systematically studied and can so far only be gleaned from ad hoc information. Clearly, the extent of colonial oppression would have limited such a process, but this should not be confused with its absence. Just across the border, for example, in Southern Mozambique where colonial oppression was, if anything, more severe than under British tutelage, Hermele finds evidence of a process of differentiation among the peasantry during the colonial period, although he notes that "it was rather weak and its dominant feature was the growth of non-agricultural activities such as trade and transport" (Hermele, 1988: 15). Where such differentiation did occur, it was mainly as a result of the access to cash from mine wages (ibid.).
This is a crucially important remark which receives further corroboration in studies of the post-colonial period to which I now turn. Ruth First's Black Gold (1983) for example, was an exception to the overall migration literature in that it understood the necessity of investigating the rural side of semi-proletarian life. First looks at the question of differentiation explicitly in Southern Mozambique, although disappointingly this is only treated in a very short chapter of six pages and seems to be added as a kind of afterthought. First did not find any rich peasants in the area she investigated (Inhambane Province), but only a differentiation between middle and poor peasantry, and in this process of differentiation the differing impact of mine wages was not a sole but often the most determining factor (First, 1983: 128, emphasis added).
What this means of course, is that migrants to the South African mines were reinvesting their wages in agriculture, and that this was the most important determinant of the higher position of the middle-peasants:
The middle peasantry... relied on mine wages, first to invest in agricultural production and then to supply means of production and a market for artisan production. (They sold a significant volume of crops). The poor peasantry likewise depended on mine wages, but not to reinforce agricultural production but for lack of it (ibid.: 130).
She also notes that, "on the whole middle peasants have fewer family members away at the mines than poorer peasants in the same area" (ibid.: 131). The importance of these findings is that they radically undermine the linear proletarianisation thesis. In the case of the middle peasants at least, wage-labour is subsiding petty-commodity production, although the reverse process stressed by Wolpe and others is also taking place. In other words, super-exploited labour on the mines and petty-commodity production in rural areas mutually condition each other's existence. It also follows that high levels of migration cannot be uniformly visualised as an "index of poverty among rural Africans", as I have noted Mafeje as stressing.
These findings are not restricted to Southern Mozambique. My own research on the differentiation of the peasantry in Swaziland (Neocosmos, 1987a), reveals similar processes at work. Out of a sample of around 200 Swazi homesteads, 20 percent were found to be poor peasants (i.e. semi-proletarians proper), 36 percent were lower-middle peasants, 36 percent upper-middle peasants, and 7 percent rich peasants (ibid.: 48). It was also found that the lower the peasant stratum, the greater the reliance on wage labour (ibid.: 51), and that all classes had some family members away at work, with the result that the rich and upper-middle strata utilised the proceeds of that labour for accumulation (not necessarily in agricultural production), given the lack of alternatives avenues for accumulation from p.c.p. in Swaziland (ibid.:72). Thus these results largely corroborate First's conclusions.
Black Gold and my own work on Swaziland, are the only two works which I know of to investigate explicitly the differentiation process among the peasantry in the region, by studying production relations, although I am aware of research in progress which follows similar lines of thought in Zimbabwe and Mozambique.13 In Lesotho too there is evidence of differentiation although Murray (1981) ascribes it to the Chayanovian biological forces of the `life cycle' and eschews social explanations of the phenomenon. In Botswana, the trend seems to be to employ a notion of `peasantariat' (e.g. Parson, 1984). Although this term does overcome some of the problems of linear proletarianisation as it is an attempt to avoid its assumption of necessary `downward mobility', it conflates those producers who are forced to sell their to ensure their reproduction with those who sell it as a source of accumulation.
In Zimbabwe, as I have noted, there is increasing evidence of differentiation in regional and class terms, although the latter is largely measured in terms of distribution relations (e.g. Coudere and Marjisee, 1988; see also Weiner, 1989). Interestingly enough, recent research has been suggesting implicitly or explicitly that during the independence war in Zimbabwe, ZANU relied most heavily on the `traditional leaders' who were among the better-off peasants,14 while Hermele (op. cit.: 24) makes the same point with regard to FRELIMO in the liberated zones during the liberation war.
In South Africa we are confronted with a literature which almost uniformly, stresses the dispossession of the African peasantry in the interests of white mines and farmers, from the early twentieth century to the present. The effects of state policy are seen as governed by a simple logic of the creation of wage-labour for expanding capitalism and its effects as monolithically destructive on the African population. While this literature served and continues to serve the valid purpose of demythologising segregationist and Apartheid ideology, it is still largely framed within the nationalist problematic inaugurated by Plaatje's (1982) seminal work, and is thus largely limited in its ability to elucidate the contradictory effects (let alone content) of state repression (whether legislative or otherwise). This dominant tendency largely reflects the objective fact that in South Africa we are confronted with historically the most repressive state in the region. It is also limited by the equally objective fact that the national question has still to be resolved there. Thus the South African literature on the agrarian question seriously lags behind that which is beginning to emerge from the independent countries in the region (let alone that which has been produced in Eastern Africa for example).15
As is well known, it was the 1913 land act, later consolidated in 1936, which laid the foundation for the African reserves which were in the 1960's to be transformed into `homelands' or `bantustans'. It was under these acts and a panoply of legislation which was to follow, that peasant owners, sharecroppers, money tenants, and later (in the 1960's) labour tenants, were to be removed from `white' agricultural land and dumped into the reserves. While the historical literature is beginning to chronicle the resistance of the peasantry to these processes (e.g. Beinart and Bundy, 1987; Bradford, 1987), such struggles are largely being portrayed as interesting episodes, but ultimately losing battles. And yet the Surplus People's Project, the main recent chronicler of the destruction of forced removals estimates that between 1960 and 1983 about 1.1 million `farm workers' and 475 thousand peasant owners have been removed from white agriculture (Platzky and Walker, 1985). Although the category of `farm worker' is confusing because it tells us nothing regarding the nature of that labour, it is reasonably clear that this refers primarily to peasant labour tenants (TRAC, 1988: 26).
We should thus be wary of the categorisation employed in South African statistics. Thus when we are told (Wilson and Ramphele, 1989: 24) that in 1980, just under 30 percent of Blacks lived on white owned farmland16, we should be careful not to assume that all these people just constitute different kinds of wage-labourers (as Marcus, 1986, for example seems to want to do), especially as so many white farms are organised in a highly paternalistic and repressive manner reminiscent of Latin American latifundia (IDAF, 1988). Thus Claasens (1989:15) notes that "there are vast numbers of Black people living on white farms who are not wage-labourers their dependents. And this number of people is continuing to increase". Thus it seems that the ratio of Africans to whites on farms has soared since the 1950's and that the total number of rural blacks outside the `bantustans' increased in absolute terms by two million between 1951 and 1980 despite the high levels of removals noted above (ibid). Although I have not been able to study these figures in any detail, if reliable they certainly testify as Claasens notes, "to a process of tenacious resistance and return in the face of overwhelming odds" (ibid.). They also evidently testify to the fallacy of the linear proletarianisation thesis and the determinism inherent in radical political economy in general.
Unfortunately, we have little detailed information regarding production relations on white farms and we have to rely primarily on impressionistic accounts by activists. Thus TRAC's (1988) description of labour-tenants in one of the obvious pockets of `survival' in South Eastern transvaal and Northern Natal notes that tenant families sometimes employ labour to pay their labour dues to the farmer and that "this is very common" (ibid.: 10) a process which implies differentiation amongst labour-tenants and which is similar to practices recorded elsewhere in the world (e.g. in Chile in the 1960's). We do not know how common this practice is, but we do know that such labourers receive wages of between R60 and R130 a month (R30 from the farmer and between R30 and R100 from the tenant family) (ibid.). This practice is of course, not understandable from within the parameters of the linear proletarianisation thesis; how can `proletrianising peasants' employ wage-labour?17 The same account of these tenants also notes:
It is not uncommon for families to have bred cattle of excellent stock which are coveted by other farmers, Black and White alike. I have seen big oxen change hands for over R100 a piece and goats and sheep for R80. Often people own ploughs and sometimes even tractors which they have acquired and maintained over generations(ibid.: 12).
This account also notes that farm workers sometimes also have access to their own arable land as well as being allowed to keep stock (in one instance 25 cattle and 9 goats) on the farm (ibid.). I do not wish to imply that these instances are somehow generalisable, but only to suggest the deficiencies of conventional wisdom and the desperate need for serious investigations of production relations on white owned land.
As far as Black peasant owners outside the bantustans are concerned (the so-called `black spots' as they are known in racist state terminology), claasens notes that most have been removed to the bantustans and that there are fewer than ten `black spots' left in the Transvaal all of which are resisting removal. However, the fact that 75 years after the land act was passed, the state is still removing peasant owners (mainly communities) bears testimony to their resistance which is certainly not abating.18 It also shows that the application of oppressive state legislation is never unproblematic.
The bantustans are usually described by `radical political economy' as `labour reserves' or bases for the `reserve army of labour' and attention is drawn to their declining environment due to overgrazing, erosion and water shortage so that a strong case is made for the `non-viability' of agricultural production in these areas. The classical statement of this argument is Yawitch (1982). While these arguments do provide a strong corrective to the class notions of apartheid ideology, they are limited by their inability to comprehend the contradictory nature of social relations in the bantustans which mean that state policy, even when uniformly applied, has contradictory effects. Thus Yawitch (op.cit.) makes much of the disaster of the so-called `betterment schemes', a kind of forced villagisation programme (which resulted in overcrowding on too little land) which the state enforced in various phases from the 1930's to the 1950's. While there is no denying the destructive effects on peasants of such programmes, precisely because they are enforced by state agencies, they do not have uniform effects. Cooper (1988: 95) for example, notes that land was not divided equally and that larger plots were allocated to chiefs, headmen and their supporters. In addition he notes importantly that there is a large amount of variation between bantustans. In Qwa Qwa and Kwandebele, 80 percent of families are landless, while in less crowded areas such as Northern Transkei 40 percent have no plots.
While there is no doubting the overcrowding in relation to land availability in the bantustans, it is also the case that the bantustan authorities themselves have been restricting access to land by land-hungry peasants, preferring to lease this out to ambitious large-scale capitalist (so-called commercial) farming schemes under the pretence of `development' exigencies. It must also not be forgotten that the overwhelming majority of the land in the bantustans, is under so-called `traditional' tenure and thus allocated and controlled by chiefs (ibid.: 92), a point the importance of which I shall discuss in the next section.
There is little doubt regarding the existence of substantial agricultural (and non-agricultural) p.c.p. in the bantustans (see Beinart, 1988). The difficulty resides in ascertaining the different classes of differentiated peasants there. Mbeki, in his celebrated critique of the bantustan system and his study of the Transkeian peasants revolt of the 1960's, already distinguished between the chiefs, the middle class, the traders and landholding peasants, and the mass of the landless (Mbeki, 1984: 109-110). A notion of differentiation for the same period is also used by Moll (1988), while for the more recent period, Innes and O'Meara (1977) have discussed figures which show that 91.6 percent of Transkeian households were engaged in some kind of agricultural production, although only 8.4 percent of that total farmed for both home use and sale, while only 0.1 percent produced exclusively for the market (Innes and O'Meara, 1977: 72).19 More recently, Southall (1983) and Josana (1989) have discussed the emergence of a bourgeoisie in the Transkei, part of which has a basis in landownership. Keegan (1988: 151) remarks that in the bantustans there are also "wealthy businessmen" such as tractor owners (Black or White) who work the land of migrant families in exchange for 90 percent of the crop.
As far as Kwazulu is concerned, the socio-economic differentiation of the peasantry is shown by May (1987), although his discussion is couched in distributionist terms,while Vaughan (1989: 34) briefly documents the expansion of sugar p.c.p. in outgrower schemes in Natal over the past fifteen years. Finally, it is worth ending on this telling remark by Keegan:
The strategies of survival that Blacks have employed (vis-a-vis the South African state - M.N.) have been enormously varied. But, in particular, urban Blacks in the twentieth century have remained remarkably committed to the land, mainly because so many early revenues of Black entrepreneurship have been closed off, and more advanced avenues were monopolised by whites from the start. The insecurities of urban life under apartheid have meant that even those who are most secure against urban dispossession and ejection have often continued to channel some of their income into rural resources. The cattle herds of the Transkei, it has been found, are more commonly the property of urban dwellers with so-called `section 10' rights, rather than of migrants whose home base is in the reserves. The relationship between higher income-earning capacity in the urban economy and control over rural resources persists to this day (Keegan, 1988: 150).
There is perhaps no better indication of the fallacy of the linear proletarianisation thesis than the well-known widespread popular demand for land in South Africa, whether from the urbanised petty-bourgeoisie as Keegan notes, or from the peasantry and agricultural labourers themselves. It would be a strange proletariat indeed which clamours this vociferously for land redistribution, even though this demand is not organisationally articulated by an independent peasant movement, for example (although see note 18 above). On the other hand, all liberation movements contain sections in their programmes regarding land redistribution, despite the fact that this struggle has (with rare exceptions) been largely urban based in the 1970's and 1980's.20
I return to a discussion of the link between the national and agrarian questions in South Africa in the final section of this paper. For the present, having presented some of the evidence for the existence of agricultural p.c.p. and its differentiation, I want to turn to a brief discussion of the political economy of imperialism in the region, because it is precisely a misunderstanding of this political economy which lies at the root of the misinterpretation of history characteristic of the position of `radical political economy'. This discussion is also necessary as a step towards my assessment of the connection between the agrarian and national questions in part four of this paper.
The fact that `radical political economy' has been unable to recognise the production and differentiation of p.c.p. in the region is not simply due to theoretical errors; nor is it due to the conscious espousal of a specific political position. Most writers in this tradition would be unaware of, and would probably vehemently deny, their objective adherence to a petty-bourgeois nationalist position. Ideologies after all are fundamentally the social products of real material relations. This misconception was (and still is) partly due to the real fact that such differentiation is restricted to a great extent during periods of intense political repression, most evidently prevalent under colonial and apartheid state rule.
The relative process of democratisation which independence unleashes tends to lead to a clearer demarcation of classes from among the peasantry and the population as a whole (in economic, political and ideological terms). Although this may not always seem obvious, there is a clear relation between extended (bourgeois) democracy, especially in the form of popular democracy, and class demarcation/constitution. It was Lenin who consistently argued that the deeper and more widespread this democracy, the clearer the demarcation of classes (especially that of a proletariat).21 For Lenin, the `leadership of the proletariat' in the democratic revolution consisted precisely in effecting the clearest demarcation revolution consisted precisely in effecting the clearest demarcation of classes via the greatest possible extension of democracy, while under `bourgeois leadership' there was always a tendency for the bourgeoisie to compromise with reaction by maintaining many of the non-democratic features of the pre-revolutionary period. A good example is the maintenance of pre-independence repressive legislation and its use against the people and their representatives in Zimbabwe - a common feature of neo-colonial regimes. (Levin and Neocosmos, 1989).
In any case, the main point to be recalled here is that the repressive forms of state associated particularly with colonialism and apartheid, systematically restrict the demarcation of classes from among the peasantry and the oppressed people as a whole. In addition, it should be recalled that explicit state policy was not always directly motivated by a `desire' to proletarianise. Rather this policy often alternated between `proletarianisation' and `peasantisation'. The shift in policy towards the South African bantustans, first as labour reserves,then in more recent years as `rural development areas' to restrict the flow of migrants to the cities is a case in point (Unterhalter, 1987).Similar processes also operated throughout Rhodesian history.
Moreover, state policy did not simply vacillate between these two alternatives, but had at times, to pursue both simultaneously, despite the obvious contradictions. This seems to me to constitute the essence of the South African state's `betterment' policy for example, in spite of the fact that the Tomlinson Commission's recommendations for the creation of a rich peasantry in the bantustans were not followed (Yawitch, 1982; Levin and Weiner, 1989).
It should be recalled that the South African racists' ideal, represented in Verwoerd's attempt to socially engineer W.A. Lewis' `dual sector model' in the real world, required the separation of the commodity labour-power from the laborer himself, a point explicitly made by state representatives themselves.22 Now, the separation of labour from the laborer in social not just in geographical terms, requires the ability of the said laborer to reproduce himself partly independently of the wage form, otherwise the pressure to increase wags and thus overcome this separation would be too great. To that extent the `ideal' racist policy required at one and the same time the partial freeing of labour from the land and the partial tying of labour to the land. It is this process which Wolpe and others have seen as one of `preservation - dissolution' of `pre-capitalist modes' but, as recent research has shown, unfree labour is perfectly compatible with and often produced by capitalist relations so that it is not necessarily `pre-capitalist' in content (Corrigan, 1977; Brass, 1986). On the contrary I would suggest that such forms of unfree labour are a specific attribute of capitalism in its imperialist form in the region.
First (op.cit.: 130), makes the important point that in some areas of Mozambique, peasant production had been virtually destroyed as a result of extensive land alienation, but that in other areas where mining-capital had extensive influence, the peasantry was systematically reproduced alongside wage labour. This is a very important remark for it seems that in many instances, mining capital reproduced p.c.p. to an extent where other forms of intensive capitalist development could not. Thus it is reasonably apparent that, far from just having a proletariansing effect, the development of mining capital in particular had much more contradictory effects, one of which was to produce and reproduce p.c.p..
The causes of these contradictory policies and effects briefly sketched above, are not simply to be laid at the door of the `desire' of capitalists to secure for themselves the supply of extra-cheap labour-power, although there is no doubting that this conforms to what capitalists always `desire' and that labour is super-exploited in the region. Nevertheless, capital has historically known other ways of reducing wages and increasing its profits (e.g. by reducing the value of labour-power through increasing absolute or relative surplus value extraction or deskilling). Accounts which rely on the `will' of capital or capitalists as an explanation can never explain the particular strategy employed other than with reference to an accidental choice. Such accounts are obviously steeped in voluntaristic assumptions.
The causes of the existence of super-exploited labour and the attendant production and re-production of p.c.p. are to be found in the social structure of imperialist relations in the region, rather than in the `will' of capitalists or the `interests of capital'. It is this fundamental nature of imperialism which the linear proletarianisation thesis fails to comprehend, which accounts for its misunderstanding of the simultaneous production or a proletariat and p.c.p., either separately or especially in its relatively stable combination of `worker-peasants' or `peasantariat'. This combination must not be seen as just a temporary stage on the ineluctable road to full proletarian status, but as a necessary product of current social relations which could thereby `dissolve' in either direction once these relations are systematically democratised.
As Shivji (1988b) has recently reminded us, the operation of imperialism - as a specific form of the capitalist mode of production - systematically modifies the laws of motion of capitalism in some fundamental ways which are central to the reproduction of the social formations of the region.23 Imperialism, as Lenin showed, is monopoly capitalism. As such its form of operation is geared not just towards the formation of profits, but of super-profits. What this means is that commodities are not always bought and sold at their value. In Africa this process of `cheating' takes place systematically (not as a matter of accident but as a matter of course) in some dominant markets and is usually refereed to as `unequal exchange'. Its operations have been discussed at length in the literature, in at least three different areas:
- internationally (between countries) whereby capital accumulation is restricted in the neo-colonial countries and dominated by imperialist/monopolist interests - a system discussed at length by dependency theory:
- between town and country (or the state and the countryside; or industry and agriculture) whereby a so-called `surplus' is systematically extracted from the peasantry. In this relationship generally, the majority of peasants are exploited (i.e. they only reproduce themselves below value levels so that they appear as wage-labour equivalents - see Bernstein, 1977), while a minority have their profits reduced or a forced to subsist as middle peasants. This process receives extensive treatment in the literature from East Africa in particular;
- on the labour market, where wages are systematically depressed below the value of labour-power, the typical Southern African condition.
I want here to discuss some of the effects of these relations, which are, it must be reiterated, the normal conditions of the operation of capitalism in the region, through a brief discussion of the latter two exchange mechanisms. Four important points need to be made. First, as I have noted, the systematic extraction of a `surplus' from petty-commodity producers is felt differentially. Some will be exploited, others will reduce their accumulation and `squeeze' their wage-labourers, others still will be confined to middle peasant status. Clearly, this unequal exchange mechanism depresses the differentiation process so that p.c.p. may take the form of homogeneity, especially under severely repressive conditions, either as wage-labour equivalents (Bernstein, 1977, on Tanzania) or as middle peasants (Gibbon and Neocosmos, 1985, on Tanzania; Neocosmos, 1987a, on Swaziland). But this does not mean either a lack of differentiation, or an ineluctable process of proletarianisation. If total proletarianisation took place, the extraction of `surplus' could no longer operate and this feature of imperialist relations would cease to exist. Further it would threaten the existence of imperialism itself, hence as has been observed, this process stops short of full expropriation (Bernstein, 1977).
Second, if the super-exploitation of labour is to be systematically reproduced by paying for labour-power below its value, then some link to p.c.p. must be reproduced, either by tying labour to the land, or through the reproduction of other forms of p.c.p. (such as the so-called `informal sector' in towns). These links to p.c.p. imply a relatively effective limit to the proletarianisation process. In the present context of social relations, this is the most effective way for capital to reduce wages on a large scale, compared to say the reliance on relative surplus value extraction which would necessitate inter alia the capitalisation (as in Western countries), although these methods are also employed in a subordinate way in the region. At the same time, this dual reproduction of wage labour and p.c.p. `in combination' will tend to allow for the reproduction of `pure' wage labour and pure p.c.p. `at the margin' so to speak. This is simply because the operation of the capitalist mode of production is only modified and not fundamentally transformed by monopoly conditions. Ultimately, it creates the prerequisites for the reproduction of the essential relation between capital and wage labour, and hence for the appearance of the latter in its `purer' forms.
Third, this systematic process of unequal exchange has the effect of giving a greater centrality to the market in society, so that market relations assume an overriding importance in social life. Simply put, it appears (this is a real appearance) that the market itself is the source of exploitation as `cheating' becomes the norm. Incidentally, this is obviously the source of the error of dependency theory. More importantly, as exchange relations operate between entities defined by a capitalist division of labour (e.g. between specialized enterprises, between sectors of the economy such as agriculture and industry, between employers and owners of specialised useful labour in Marx's sense, and so on -Gibbon and Neocosmos, op.cit.), this division of labour assumes an even more prominent position within political economy. Thus what is often referred to as `uneven development' becomes more noticeable.
Its expressions, as Shivji (1988b: 10) notes, such as geographical inequalities (we could add inequalities between nationalities), distinctions between town and country, industry and agriculture, mental and manual labour and so on, become more acute. It follows that under these conditions, class divisions do not obviously dominate the social formations in question. These are displaced in importance by nationality and regional differences, sectoral differences and countless distinctions between buyers and sellers, so that the primary objective of the struggle of the oppressed becomes one of redressing capitalist market inequalities (a struggle for bourgeois freedoms and rights) rather than capitalist exploitation as such.
Finally we are driven in this way to a discussion of the question of democracy. For Lenin, imperialism was "indisputably the `negation' of democracy in general, or all democracy; he added that "democracy corresponds to free competition, political reaction corresponds to monopoly".24 This is simply because systematic `cheating' on the market requires not only "the dull compulsion of market forces" as Marx put it, but also what is sometimes termed `extra-economic coercion'.25 As a result, these conditions tend to produce highly repressive states. While colonial and apartheid states are typical of monopoly conditions, Shivji also argues that "the anti-democratic nature of imperialism finds its most concentrated expression in the neo-colonial states of Africa" (ibid.: 9) which are overwhelmingly repressive in character.
As a result we have a multitude of democratic struggles in the region most clearly represented by the South African liberation struggle. Although these struggles have a class content, the dominance of the division of labour (e.g. between black and white in South Africa) ensures that they are fought out between organisations which represent a number of classes (and/or sections of classes). It also follows from my earlier assessment that it is in such periods of democratic transformation that classes are likely to become more clearly democratic resolution possible to such struggles. The agrarian question is one of those issues the democratic resolution of which is necessary for future advance.
Thus in brief, it can be argued that in the Southern African region in particular, the dominant tendency has been for monopoly capitalist conditions to produce not a clearly demarcated proletariat, neither a proletariat in the making - nor for that matter a large `pure' peasantry - but a strange amalgam as Wallerstein (1976) for one recognised some time ago. This semi-proletariat which is a specific and necessary product of these conditions cannot simply be visualised as a `creation' of the state or of capitalists, neither is it an aberration or a `stage' to full proletarian status. The state and capitalists were only able to `create' it so to speak because the material social relations under which they operated, provided the conditions for such `creation'. The (partial?) dissolution of the semi-proletariat which is likely to result from the liberation of South Africa is likely in reality to operate in the direction of a proletariat and of p.c.p., as it has done in the other now independent countries. In particular no lasting solution to the migrant labour problem can be provided unless more land is available for rural production, unless one wishes to see the `informal sector' (urban p.c.p.) burgeoning beyond control.
Any thorough process of democratisation would thus require the freeing of the working class from its ties to the land (which allow for its super-exploitation). Moreover, the same process also requires a solution to the agrarian question through a thoroughly democratic land reform (including land redistribution) in order to free the differentiation of p.c.p. from its oppressive constraints.26 This is necessary, not because of the supposed perseverance of pre-capitalist relations, neither is it a retrograde step towards the `recreation' of a `backward peasantry' (as is sometimes argued in South Africa - e.g. Krikler, 1987). It is necessary because the existence of p.c.p. is a reality and because socialism can only be achieved on the basis of thoroughly democratic resolution of the crises of imperialism. Such a conclusion should perhaps be more apparent in the era of `peristroika' and `glasnost'.
In the independent countries of Southern Africa there is, by and large, very little question of the link between the national and agrarian questions as the largest proportion of the population is agriculturally based, or at least partly secures its existence from agricultural production. However in the case of South Africa, this link is not immediately evident given the relatively high degree of proletarianisation and industrial development in that country. A recent collection of political articles on the national question in South Africa makes no mention whatsoever of the agrarian question.27 This is very disappointing but explicable in terms of the relative neglect of the agrarian question by the national liberation movement in that country (Bundy, 1984). I wish in this, the final section of this paper, to provide the beginnings of an assessment of the important link between the agrarian and national questions in South Africa. My treatment will be primarily theoretical and will concentrate on providing and account of the question of `tribalism' or `ethnicity', or what I prefer to call, following Mamdani (1984), the "nationality question" and its link to p.c.p. in South Africa.
Apartheid must be understood as a form of capitalist state in the era of imperialism. One of the fundamental ways in which it differs from the segregationist state in South Africa (1910-1948) is in its pushing of the notion of `reserves' common to settler colonial regimes, to its `logical' conclusion of `independent' or quasi-independent `homelands' or bantustans. This formed the basis for the ability of the state to forcibly `de-nationalise' a large percentage (under Verwoerd it was hoped a majority) of the Black population, 53 percent of which live in these areas.
The colonial and segregationist states had already laid the foundations of racial oppression through a systematic creation of nationalities in common with colonial states elsewhere in Africa. Thus Mamdani states of Uganda during the colonial period that:
Every ruling class tries to use the state machinery to organise itself and to disorganise the oppressed. The main thrust of colonial policy was not only to deepen existing nationality contradictions but also to introduce new ones where none existed before. In fact the colonial power turned nationality into an organising principle for both the state structure and for economic life, thereby institutionalising it (Mamdani, 1984: 1048).
A major pillar of this process of institutionalisation was the creation of `traditions' - the `tribalisation' of African life.
As Ranger notes about pre-colonial Southern Africa, "people defined themselves politically - as subjects of a particular chief - rather than linguistically, or culturally, or ethnically" (Ranger, 1985b: 4). The linguistic, `tribal' or `ethnic' entity or division was thus primarily a colonial product. In this way, whole `traditional' cultures and polities were systematically manufactured, written-up by colonial officials with eager help of anthropologists and dropped on an unsuspecting population, often with the connivance of local chiefs who benefitted from increased powers in the process (e.g. ibi 6-9; Macmillan, op.cit.). In fact a whole body of sophisticated literature has emerged of late which shows the intricacies of these developments in some detail (e.g. Ranger, op.cit.; Hobsbawm and Ranger, 1983; Vail, 1989).
What emerges from this literature amongst other things, is that this creation was not just a result of state policy, but that it provided the parameters within which socio-economic relations would operate so that the oppressed themselves often played along with `tribal' distinctions in order to secure employment for instance (Ranger, op.cit.: 13-14). Quinlan gives the following example of the economic reasons for the formation of `tribes' on the South African highvelt in the 1860's:
indigenous pastoralists and farmers needed land which was rapidly coming under the control of the colonial settlers, and which was only being allocated to recognisable groups. Hence, it was in the interests of many African refugees to form groups according to colonial conception of African society (Quinlan, 1986: 33).
Hence this `tribalisation' or formation of nationalities was a complex process in which both oppressors and oppressed participated, and which went beyond the formation of a consciousness or `tribal' identity. This development was about the production of specific social relations (from the confines of earlier ones) within a division of labour which achieved its most developed expression in the epoch of imperialism.
The creation of reserves in Southern Africa after the 1890's by the colonial state was to entrench this division into nationalities through its fundamental production of a chieftaincy whose powers were based on the control over land. In pre-colonial Southern Africa, the powers of the chieftaincy were not founded on their control over land, but on their control of bridewealth (cattle) within lineage modes of production. As I have argued in the case of Swaziland, the creation of reserves in a period of rapid and increased commoditisation, helped to transform the powers of the chieftaincy from ones based on the control over cattle to ones based on the control over land (Neocosmos, 1987b). In addition, of course this new form of control could only operate satisfactorily because it occurred simultaneously with a process of `peasantisation'. In the absence of a base of agricultural p.c.p., the chiefs would not have been able to exert control over, and extract fees from migrants quite so easily as they obviously did.
While I have used this argument to show that the Swazi state is fundamentally capitalist and that `tradition' is not a simple left-over from the past, I have also used it to suggest that the formally traditional aspects of the Swazi state are explained by the fact that state power is fundamentally founded on its control over p.c.p. which is then generalised to the population as a whole (ibid.). Thus agricultural p.c.p. provides the foundation for the reproduction of the rule of the chieftaincy. This point is generalisable with appropriate alterations, to Southern Africa in generalisable with appropriate alterations, to Southern Africa in general and to South Africa in particular. Thus Keegan for example note that:
Independent access to land was dependent on access to chiefly patronage. Only through the institution of chieftainship could access to resources be legitimated. Although clothed in the garb of `tradition', there was in fact very little historical continuity in the new foci of legitimacy and patronage that were emerging under the auspices of the industrial state in the new segregationist era (Keegan, 1988: 149).
Because in South Africa, the operation of the (post 1948) apartheid state is based on the creation against the will of the majority of the people, of `independent' or quasi-independent bantustans, the national liberation struggle in that country has not taken the classical form of demanding independence or secession but its opposite viz. the demand for the creation of a unitary state by overcoming divisions based on race and nationality. The history of the ANC in particular, is one of a struggle to forge unity, first among Blacks, then between Blacks, Coloureds, Indians and democratic Whites which culminated in the adoption of the Freedom Charter in 1955. Struggles against enforced separation (secession) have continued to this day, the most obvious example being the recent (1987-8) struggle of the people of Kwandebele against `independence'.
The fundamental aspect of the national struggle in South Africa, therefore, has been the question of uniting the masses of the oppressed in order to form a new nation. In Leninist terms, the national struggle in South Africa has been for the right to unite and not for the right to secede (although for Lenin the one implied the other). The demand for unity is a democratic demand as it is expressed by the overwhelming majority of the oppressed, while secession has been systematically enforced on the oppressed people (although it does have support among a minority of Blacks in the bantustans as we shall see) who have in recent years, consistently resisted the pressures for their division.
Since the 1960's the most extreme form of division - the enforced secession of nationalities - has been the hallmark of the apartheid state which, combined with its other ultra-repressive measures (see Wolpe, 1972), distinguish it from the earlier segregationist forms of the South African state. This division thus comes to form the foundation of the rule of the racist apartheid state - i.e. the division between Whites and Blacks under apartheid (internal colonialism) is premised on the theoretically prior division and ultimate balkanisation of the Black population. Without such specific divisions, the other aspects of racist oppression (such as legislation for example) would be inoperable and the state would be unable to secure its rule.
In any discussion of the contemporary bantustan system, it must be recalled first of all, that on assuming power in 1948, the Nationalist government went about systematically increasing the powers of the chieftaincy under the 1951 Bantu Authorities Act and the Bantu Laws Amendment Act of 1952. In terms of these acts, chiefs became paid government officials responsible for the first level of local government. Their powers of land allocation and settlement of disputes were thus codified in law and hence `traditionalism' was entrenched (Haines ans Tapscott, 1988: 166; Copper, 1988: 92; Segar, 1989: 82-3). Previously the chiefs had been responsible to local government officials. As Mbeki put it with reference to the run up to Transkeian `independence': 62).
The chiefs have been used with skill. Every little chief, even those not getting large stipends from the government, has been placed in a position where he can make money for himself at the expense of the people. Fresh powers of civil jurisdiction have been placed in the hands of the chiefs, for example, and many have amassed small fortunes in fines and bribes (op.cit.: 62).
The importance of the chieftaincy's control over land is manifold. In the first place, it not only enables them to extract bribes from the peasantry in return for allocating them a plot on which to produce, but the threat of banishment from the land constantly hangs over the peasantry especially in conditions of land scarcity thus enabling the development of patronage relations which systematically fleece the people of their resources (Haines and Tapscott, op.cit.: 170). In addition this power provides the basis for the mobilisation of unpaid labour for the construction of public works. "Road building, contour building, the removal of noxious weeds, the repairing of dipping tanks and the construction of houses for the chiefs, come under the category of services to which compulsory unpaid labour is drafted" (Mbeki, op.cit.:99). Chiefs can even impose levies for the purposes of buying the chief a car (ibid.: 102). Mbeki (op.cit.: 98) also notes that chiefs courts were used in order to force the resistance to `betterment' schemes into submission. What these powers also mean is that they provide the framework within which development projects operate in the bantustans (Haines and Tapscott, op.cit.: 172-3). The result is that as in other countries of the region where similar practices operate (Neocosmos, 1987b; Mamdani, 1987), state run `development' as a whole contributes to the reproduction of the plunder and oppression of the peasantry and to the reproduction of state power.
This chieftaincy is firmly imbedded in all the state apparatuses such as the executive, the legislative and the top civil service with access to large salaries and perks as a result (Mare and Hamilton, 1987: 36; Quinlan, op.cit.: 42). For the Transkei alone it has been noted that in recent years official expenditure on salaries for chiefs and headmen has risen from R323 500 in 1974/5 to R536 500 in 1975/6 to R1 440 000 for 1987/8 (Haines and Tapscott, op.cit.: 168).
The basis of this wealth and power is the so-called `traditional' and `communal' land tenure system which covers the greatest part of the land area of the bantustans (Cooper, op.cit.: 92). The difference of this tenure system from private property in land lies in its form whereby petty commodity producers and thence the people in general can be systematically controlled and plundered. The `traditional' customs are conveniently set aside when it comes to forcibly resettling people or to clearing land for large capitalist farming. It thus bears only a superficial (legalistic) difference with private property in the usual sense. This difference is crucially important however, for it is on such a difference that the power of the chieftaincy is based. Thus when the Tomlinson commission reported in 1955 that freehold tenure be granted to the African population in the `homelands' on the condition that it was `adequately used' in the interests of `development', the government itself rejected the proposal with Verwoerd himself stating that it "would undermine the whole tribal structure" (SAIRR, 1973:79-80).
At the same time capitalist farmers in particular, such as those in Natal, have understood the benefits to be gained by returning the chiefs to areas from which they had previously been evicted. Thus one of them stressed in 1987:
Our aim is now to return the chiefs to their original place of residence with their families where they will once more form the legal nucleus of the area (cited in Mare and Hmilton, op.cit.: 37).
The same spokesman for agrarian capital continued by noting that one year after signing an accord with the `Zulu authorities' to return the chiefs to the land, the incidence of stock theft had not diminished, but that the `community' no longer protected such thieves on their release from goal but banned them from the district (ibid.). Thus white capital in Natal at least, has understood that so-called `traditionalism' and `communalism' can be effectively enlisted in the service of private property, which was precisely what the famous `Kwa-Natal indaba' was about and seems to constitute the direction towards which the present Deklerk government is moving.
In sum the `tribalisation' of the oppressed population in South Africa and its division into quasi-independent nationalities which is the foundation stone of the apartheid state, is itself based on the reproduction of the powers of the chieftaincy over a population of petty-commodity producers which can then be extended to other areas where agricultural production is less in evidence. While this has succeeded to a degree in Kwazulu/Natal because of the power of Inkatha in some urban areas to control access to housing and jobs and while it is also evidenced by the so-called `induna' system on the mines for example, it is more difficult to enforce in areas where wage labour predominates, precisely because of the diminished importance of land as a means of production. Hence, Quinlan notes that in Qwa Qwa where land is scarce, the chieftaincy has been weakened as a result (Quinlan, op.cit.: 43).
In any case, I hope that I have shown that there is a clear relation between the structures of imperialism in South Africa,the apartheid state, the existence of diverse nationalities among the oppressed, the reproduction of the chieftaincy and agricultural p.c.p. This overall system allows for a far more repressive political structure in the bantustans themselves (than in the urban areas of South Africa) where popular organisations are systematically banned and vigilant squads have a free reign. It is thus theoretically absurd to see the land tenure system in the bantustans as communal like Marcus (1989) does, and -politically reactionary to suppose that such a system of tenure could provide a guide for a future land reform in South Africa like others do.
It is sometimes suggested that the communal nature of traditional African land tenure systems can provide the basis for a future transition to socialism as they are relatively egalitarian. This argument, which is most clearly associated with Nyerere's `African Socialism', is firmly rooted in Russian Narodnik (populist) theory of the 1890's, which argued that the capitalist stage of development could be avoided in a direct move to socialism based on the peasant commune. This view is also reflected in the current Soviet theory of the so-called `non-capitalist path' of development in Africa (Nkom, 1983).
In Southern Africa this conception is not always seriously discussed, yet it appears in some `Africanist' literature influenced by the position of `radical political economy'. Thus Mafeje for example, sees the maintenance of "African traditional land tenure systems, which recognise no individual ownership rights in land but rather usufruct rights to which all members of the community are entitled" as the prime condition for the resolution of the agrarian crisis in Southern Africa (Mafeje, 1988: 120). While there may arguably be a basis for holding such a view in other parts of Africa, in our region the so-called `traditional' land tenure system is based on the most oppressive system of social relations. These are not confined to South Africa itself but are only exhibited in their most extreme form in that country.
Perhaps even more surprising is the fact that someone should actually advocate such a policy explicitly for South Africa itself (rather than for Southern Africa in general as Mafeje does). Yet this is precisely what Letsoalo suggests when she states that the repeal of the 1913 and 1936 land acts would "depoliticize the land issue" (sic!) and that this "depoliticisation ...should be accompanied... not by privatisation... [but] ...by the African system of communalism. Land should be vested in the community" (Letsoalo, 1987:82-3). This writer even goes so far as to cite a text ostensibly about Lesotho and which was published in Pretoria, as a source for her view that the future of the chieftaincy is "not irretrievably lost", and which states among other gems that the people (one is surprised not to read "the Bantu") are "strongly attached" to the institution of the chieftaincy which "has served them well in the past" and for which they have "deep respect" (ibid.: 68).
Unfortunately for those `Africanists' who wish to return to the past, the communalism of African life has long been destroyed in Southern Africa and there is nothing left of the genuinely democratic relations on which pre-capitalist African societies were founded. It is one of the dangers of articulationism that it gives the impression (through its argument of `preservation') that this may be so. Rather such relations have been fundamentally subverted and transformed by imperialist political economy into some of the most oppressive relations known in the African countryside serving in South Africa, to reproduce the racially oppressive apartheid regime. Any genuine liberation of that country requires a radical transformation of these oppressive relations in order to create a process of national unity. The arguments of `radical political economy' can therefore unfortunately not help to provide a solution to the national question in South Africa.
In this context, the recent formation of the Congress of Traditional Leaders of South Africa (CONTRALESA) which brings together chiefs who are opposed to apartheid structures is a step in the right direction. It has raised the ire of bantustan leaders such as chief Buthelezi of Kwazulu. The Ciskei and Kwazulu authorities together are recently reported as having issued a statement which notes, among other things, that CONTRALESA is "a divisive force, attempting to drag divisive party politics into the traditional authority structure where consensus is vital stability and progress" (Weekly Mail, Nov. 24-30, 1989, p.5).
It is as yet too early to evaluate the direction which this organisation is taking, but a representative was recently quoted as saying that the organisation wanted "to restore the chieftainship so that our leaders work like their forefathers, who used to take decisions with community approval instead of commissioners (sic)" (New Nation, Nov. 17-23, 1989, p.2). The contradictory nature of this statement is interesting as under present conditions, a democratically accountable chief would in no way differ from any other accountable leader as there can be no return to genuinely democratic pre-capitalist forms of organisation. The consequence of this statement seems to be therefore the disappearance of the chieftaincy (as we know it), which precisely what the bantustan authorities fear. Referring to the foundation of `independent' Transkei, Mbeki wrote that:
The last thing the chiefs wanted, of course, was any weakening of their authority ...And so the Transkei Legislative Assembly is to be overweighed with chiefs, who constitute the basis of the entire Bantustan scheme. Without them it would never have been accepted in the Transkei, and without them it has no prospect of lasting (Mbeki, op.cit.: 61).
This remark made over thirty years ago is still valid today with the proviso that, unlike the Pondoland revolt of the late 1950's and early 1960's which was to a great extent predicated on a defence of the `traditional' chieftaincy against its corruption by the state (Moll, 1988) and which was one of the reasons for its failure, the contemporary national struggle must go beyond the narrow ideological confines of a peasant movement to the questioning of the chieftaincy as such. In this way, if it is to achieve a lasting victory over the apartheid state, the national liberation movement must recognise that a solution to the national question in South Africa of necessity requires a solution to the agrarian question both in the bantustans and on the so-called `White platteland' where together, an overwhelming majority of 80 percent of the Black population lives.30
I have in this paper surveyed a large volume of literature which has taken me beyond the confines of historiography in the usual sense. Nevertheless, this was necessary both in order to explore what I see as being the errors of the dominant paradigm of `radical political economy', and also to provide the beginnings of an alternative to such a conception. In the process a discussion of the kind of history we should be undertaking in Southern Africa has been treated in a rather cursory way. It may be pertinent therefore to conclude by making a few explicit remarks on this question.
The first point which comes to mind is the fact that agrarian history in Southern Africa seriously lags behind that produced in Eastern Africa for example. This has been due to the fact, as I have argued, that for conventional wisdom at least among `radicals', peasant agriculture has been seen as overwhelmingly destroyed, or at least as confined to residual pockets of `tradition' in a sea of `capitalism'. It is also due to the real fact that the national question has only recently started to be resolved in the region, while in East Africa political independence occurred long ago, so that there is a substantial volume of literature there which reaches beyond the confines of narrow nationalist history.
It is important to note in this context that if intellectuals in the region are not just to tail behind the struggles of the people, they need to be more theoretically advanced. In
particular this means that they should not restrict themselves to the use of mechanical slogans as a substitute for theory but make an attempt to face complexities of the latter.
The recent literature on the creation of tradition is a notable advance in this respect but there are as yet few histories of the post-colonial states in the region, or of the contradictory complexity of capitalist relations in the imperialist era. Most historians (indeed most social scientists) are still operating under the assumption that capitalism can only exist in one form usually found in the West, which clearly shows a developed polarisation between capitalists and free wage labourers (whether this is an accurate representation of Western societies is another issue). As such they ignore the multitude of specific forms which capitalist relations have taken in their development through contact with indigenous societies. Any slight deviation from this supposed `ideal capitalism' is referred to ad nauseam as pre-capitalist or traditional, as if the real world constituted a concretisation of the intellectual's ideal. These problems reflect distinct lacunae in theory as well as in the practice of detailed historical investigation.
The overwhelming deterministic view of `imperialism', `capitalism', `colonialism' and the `state' which are all ascribed with anthropomorphic qualities and endowed with a will, seems also to be a major problem with much Southern African historiography and social science. This has the effect of underemphasising the contradictory nature of capitalist relations and of stressing the monolithic power of oppression thereby relegating collective agency to an epiphenomenal position. Surely the basis of any genuine social science (including history) is to be able to show that the oppressed can and do change their world which is not out of their control, in spite of the fact that their actions are constrained by social conditions which exist beyond their will. Surely the object of such a science is to keep on arguing this point in the face of the unrelenting and changing offensive of bourgeois ideologies which systematically portray the existing world as eternal and unalterable. There is certainly no room in such a science for a pandering to the vulgar notions and intellectually safe mythology of state ideology.
The mirror image of crude determinism is provided by the relativism and eclecticism which exists mainly in South African historiography, dominated as it is by Whites trained in the British tradition. This relativism and eclecticism dominates in much of the new South African `revisionist' school of agrarian history where theories are opportunistically utilised or discarded as `heuristic devices', or where we are presented with such intricate historical minutiae under the guise of `scholarship' that the important historical processes at work are completely lost sight of. Both determinism and its opposite-relativism - tend to hide rather than reveal the ability of the oppressed to forge their own destinies.
Finally of course, this means that we should not just be writing the history of the oppressed but (as much as we can possible achieve) that we should be writing history from the point of view of the oppressed sections of our societies. This is not easy to achieve from our class position as intellectuals, but it is necessary, in present conditions, as prelude to the oppressed writing their own histories themselves. For we have the resources both intellectual and material, which the oppressed do not yet possess and as such we have, it seems to me, a responsibility to them as well as to future generations to ensure that future state of affairs is brought closer sooner.
1. This paper constitutes part of work in progress on the Agrarian Question in Southern Africa. Its conclusions are preliminary. I would like to thank particularly Ebrahim Jassat and Issa Shivji for debating some of these issues with me. The usual caveats apply. I wish to dedicate this work to the memory of David Webster, an organic intellectual in the true sense of the term, murdered by the apartheid state on May 1st 1989.
2. I am not concerned here with the literature of the Right, i.e. primarily colonial and neo-colonial literature - although it should be noted, for reasons I shall explain below, that this literature has come to dominate overwhelmingly the contemporary assessment of rural relations in Southern Africa.
3. See inter alia S. Amin (1972) and Kowett (1978) on the region as a whole, Legassick and Wolpe (1976) on the bantustans, Arrighi (1973) on Rhodesia.
4. See inter alia Phimister (1986, 1987), Coudere and Marjisse (1988).
5. For a theorisation of the petty bourgeoisie in an African context see Gibbon and Neocosmos (1985).
6. Incidentally, Mafeje (Op.cit) is guilty of this conflation too as when he states (p.114), referring to the present period, that "Colonial Capitalism is not able to resolve the problem of production and social reproduction in Africa". If this statement constitutes a "slip of the pen" it certainly seems to be a "freudian slip".
7. For a discussion of the notion of National Democratic Revolution (as indeed for that of compradorisation) in Africa, see Shivji (1985, 1988b).
8. By "national bourgoisie" I am not referring to an industrial bourgoisie, but to that class which is caught between imperialism and the people and hence vacillates between the two. It seems to me that Kaunda is a clear representative of such a class. (See Gibbon and Neocosmos, op.cot.: 191)
9. As far as the change at independence to questions of development rather than politics is concerned, see Mandaza (1986) which is entitled Zimbabwe: the Political Economy of Transition. The transition to what is never specified. Nowhere in the book is any discussion of oppression or mass politics to be found neither is there any attempt at class analysis. This is especially true of the section entitled "the agrarian question" which deals primarily with developmental issues and policies. But perhaps the clearest and most disturbing example of the change in attitude towards the peasantry is to be found in the writings of Macheal himself. In a text dating from 1971, Macheal notes that, in Mozambique, "we also find illiterate peasants with a knowledge of the world limited to the village horizon, where colonial domination has instilled the notion that they form an ignorant and brutish mass incapable of rational thought or initiative" (Munslow, 1985: 23). After liberation, in a statement dating from 1982, his position changes completely thus: "Peasant societies essentially are underdeveloped because they are fatalist. Mozambican society, even in the urban sector, carries the dead weight of the fatalist, resigned, passive legacy of the peasant society... The peasants use a short-handled hoe, although it is inconvenient and inefficient because they are accustomed to it" (ibid.: 130). He even goes so far as to ascribe vandalism and dirtiness in cities - indeed seemingly most of the ills of society - to this "peasant mentality" (ibid.).
10. For a discussion of the position taken in Gibbon and Neocosmos (1985) see Bernstein (1987); for a direct application of this position to Southern African conditions, see Neocosmos (1987a, 1987b) and Levin and Neocosmos (1989).
11. One of the few historians of Swaziland to take seriously the question of the construction of "tradition" is MacMillan (1985).
12. Recent research provides some evidence to suggest again that this form of land tenure is only "traditional" in form as this "tradition" can be circumvented by the state, large capitalist farms (who can expel "squatters" and charge outgrowers for irrigation) and richer peasants who put up fences and so on. See e.g. Levin (1987); Matthews (1987).
13. See CEA, (n.d.) for example. On the problems of ignoring the differentiation process in Mozambique see in particular Hermele (op.cit.), Mackintosh and Wuyts (1987) and Wuyts (1989).
14. See especially Lan (1985), Phimister (1987), Bratton (1987) and to a lesser extent Ranger (1985a).
15. Cobbett (1988) from a social democratic perspective, and Krikler (1987) from a trotskyite perspective, are both more recent examples of advocates of the linear proletarianisation thesis from which they draw different conclusions. In both cases, aggregate official statistics are used uncritically to "prove" the overwhelming disappearance of peasants in South Africa. As I have argued already, especially in Neocosmos (1987a), aggregate statistics can do nothing of the sort.
16. As opposed to 53 percent in the "bantustans" and 18.5 percent in metropolitan areas (ibid.).
17. I have argued in my Ph.D., thesis (Neocosmos, 1982, part two) that the parties of the Left in Chile also misrecognised the class nature of labour tenants confusing them with labourers paid in kind. This had disastrous consequences for Allende's Popular Unity government during the early seventies, as these parties relied on this class for their agrarian programme and ignored the real semi-proletarians and agricultural wage labourers. This was one of the reasons for the failure of the class struggle in the countryside and the ultimate coup.
18. See the recent formation in the Northern Cape of The Namaqualand Residents Association, an organisation of peasant owners which involves eight communities (South Aug. 24-Aug. 30). One of its main aims is to investigate the land claims of its affiliates and to reclaim the land. The chairperson of the organisation was quoted as saying that the people of Namaqualand have a deep land hunger (p.2). Despite the fact that such people live in communities they are not undifferentiated. It is clear that only some, usually the leaders of the resistance, are land-owners and that others work for them. This differentiation does not preclude these peasants providing a "united front" across class boundaries against state repression. See for example, Morriss(1989), Claasens (1989).
19. Innes and O'Meara (1977) do not draw a conclusion of differentiation from their paper. For a brief discussion of their figures, see Levin and Neocosmos (1989: 235-6).
20. See Bundy (1984) for a history of the treatment of the agrarian question by the South African National Liberation Movement.
21. On this link between class demarcation and democracy, see in particular LSW1 (pp.485-565), LSW3 (pp.114-128), LCW13.
22. A prominent Nationalist MP declared at the time that "the Bantu... are only supplying a commodity, the commodity of labour... It is labour we are importing not labourers as individuals... Numbers make no difference" (cited in SAIRR, 1966:1-2).
23. In the countries of the West the most important effects of imperialism sometimes differ from those in the colonial and neo-colonial countries. there they include (or have included) the formation of an aristocracy of labour, the dominance of finance capital, the formation of fascism, and also, in particular, a transformation of the labour process which has the effect of systematically deskilling labour, on which see Braverman (1974).
24. LCW 23: 43.
25. See the article by Mamdani in P. Anyang Nyong'o (1987) for example.
26. It is note worthy that no such redistribution has taken place in Southern Africa even under the most explicitly "socialist" regimes. It looks unlikely that it will take place in Namibia either as SWAPO has promised to appoint an (ex-) Nationalist Party leader to the post of minister of agriculture. The main reason for this failure in the region it can be surmised, has been the lack of any lasting independent peasant organisation, unlike in Latin America or in Asia.
27. See: The National Question in South Africa, London: Zed Press, 1988.
28. In this context see Oscar Dhlomo's (the general secretary of Inkatha) response to the ANC's constitutional guidelines in the Weekly Mail.
29. See Neocosmos (1987b) for a discussion of the land tenure system in Swaziland.
30. This figure is from Wilson and Rhamphele (op.cit.:24). Given the obsession of the apartheid state with "nationality" the figures for residence in the bantustans are probably over-inflated; moreover "residence" differs substantially from "place of work" of course. Nevertheless, these figures do imply a majority living in rural areas ("platteland" and bantustans).
S. Amin (1972) "Underdevelopment and Dependence in Black Africa", Journal of Modern African Studies, Vol. 10, No. 4.
P. Anyang' Nyong'o (ed.) (1987) Popular Struggles for Democracy in Africa, London: Zed.
G. Arrighi (1973) "Labour Supplies in Historical Perspective: a Study of the Proletarianisation of the African Peasantry in Rhodesia" in G. Arrighi and J. Saul (eds) Essays in the Political Economy of Africa, New York: Monthly Review Press.
G. Arrighi and J. Saul (1973) "Nationalism and Revolution in Sub-Saharan Africa" in G. Arrighi and J. Saul (op.cit.).
W. Beinart (1982) The Political Economy of Pondoland 1860-1930, Cambridge: CUP.
W. Beinart (1988) "Agrarian Historiography and Agrarian Reconstruction" in J. Lonsdale (ed) South Africa in Question, Potsmouth N.H.: Heinemann.
W. Beinart (1988) "Agrarian Historiography and Agrarian Reconstruction" in J. Lonsdale (ed) South Africa in Question, Potsmouth N.H.: Heinemann.
W. Beinart and C. Bundy (1987) Hidden Struggles in Rural South Africa, London: James Currey.
W. Beinart Et. Al. (eds) (1986) Putting a Plough to the Ground, Johannesburg: Ravan.
H. Bernstein (1977) "Capitalism and Petty Commodity Production" in A. MacEwen Scott (ed.) Rethinking Petty Commodity Production, Special Issue Series of Social Analysis.
H. Bernstein and H. Nicholas (1983) "Pessimism of the Intellect, Pessimism of the Will: a response to Gunder Frank", Development and Change, Vol. 14, No. 4.
P. Bonner (1983) Kings Commoners and Concessionaires, Johannesburg: Ravan.
H. Bradford (1987) A Taste of Freedom, the ICU in Rural South Africa, 1924-1930, New Haven: Yale UP.
M. Bratton (1987) "The Comrades and the Countryside: the Politics of Agricultural Policy in Zimbabwe", World Politics, Vol 39, No. 2, June.
H. Braverman (1974) Labor and Monopoly Capitalism, New York: Monthly Review.
A. Booth (1983) "The Homestead, the State and Migrant Labour in 20th Century Swaziland", Paper presented at the African Studies Association Convention, Boston Mass.
T. Brass (1986) "Unfree Labour and Capitalist Restructuring in the Agrarian Sector: Peru and India", Journal of Peasant Studies, Vol. 14, No. 1.
C. Bundy (1972) "The Emergence and Decline of the South African Peasantry", African Affairs, Vol. 71, No. 285 Oct.
C. Bundy (1979) The Rise and Fall of the South African Peasantry, London:
C. Bundy (1984) &qu