Fostering critical dialogue and debate on socio-political and economic issues

  Site map  

National Lottery Logo
Dialogue events
Memorial lectures
Writings of Harold Wolpe
About the Trust
Contact us
Harold Wolpe Memorial Trust
Unit F
Colstine Terrace
88 Belvedere Rd
PO Box 44907
Tel: +27 21 674 0361
Fax: 086 670 6772



The theory of internal colonialism: the South African case

Harold Wolpe

The view that there are close parallels between the external relationships established by colonial powers over colonized peoples and the relationship of ethnic, cultural, national, or racial groups within some Latin American societies, the United States, and South Africa, has led to the use of the notion of `internal colonialism' in the analysis of these societies.1

The specific feature which is said to distinguish `internal' from `normal' colonialism is the fact that in the former the colonizing `nation' or 'race' or other group occupies the same territory as the colonized people. As Simons and Simons put it:2

The imperial colonial qualities of the society ... become visible by comparison with the typical colony. In its normal form, the colony' is a distinct territorial entity, spatially detached from its imperial metropolis.

In all other important respects, the implication is, the components of the 'normal' imperial-colonial relation are to be found within the borders of a single state to an extent which justifies the view that it constitutes an internal colonialism. In particular, it is argued in this approach, that the 'underdeveloped' (and 'underdeveloping') condition, of subordinate ethnic and racial groups and the geographical areas they occupy within the boundaries of the state, is produced and maintained by the same mechanisms of cultural domination, political oppression, and economic exploitation which, at the international level, produce the development of the advanced capitalist states through the imperialist underdevelopment of the colonial satellites.

Notwithstanding the apparently unproblematical use of the terms 'imperialism' and 'colonialism' in the passage cited above (and in the ,writing on 'internal colonialism' generally), it is obvious from the literature: that there arc differing conceptions of imperialism and colonialism, and that these are not all equally suitable for conversion into a notion of `internal colonialism'. Lenin's insistence in his Imperialism: the Highest Stage of Capitalism, that the export of capital is a crucial distinguishing feature of imperialism and, there­fore, of colonialism in the monopoly stage of capitalism, is only one relevant example. It is, therefore, of considerable importance to analyse the conceptions of colonialism and imperialism which serve as the model from which the notion of `internal colonialism' is derived by analogy.

In making such an analysis it will be argued first that, while the internal colonial thesis purports to rest on class relations of capitalist exploitation, in fact it treats such relations as residual. That is to say, the conceptualization of class relations, which is present in the theory, is accorded little or no role in the analysis of relations of. domination and exploitation which are, instead, conceived of as occurring between `racial', `ethnic', and `national' categories. To this extent, the `internal colonial' thesis converges with conventional race relations theory (iii particular, the theory of plural society), and, as I shall show, suffers from the same analytical limitations as the latter.

It will be argued, secondly, that in so far as the theory of internal colonialism does accord relevance to relations of capitalist exploitation it does so in a manner which denudes the analysis of all historical specificity and thereby deprives the concept of analytical utility.

The question remains as to whether or not it is possible to develop a satisfactory and rigorous concept of `internal colonialism' which will provide the foundation for an adequate analysis of the internal structure and development of certain social formations. This problem will constitute the subject-matter of the third section in which an attempt will be made to provide an historically specific account of South Africa which is both based on concepts which define capitalism generally and, at the sank tine, uncovers in a rigorous way its internal colonial character.

Internal colonialism

In the theory of internal colonialism, the colonial relation appears to be characterized by two main elements. First, the colonial relation-. ship is conceived of as occurring between different countries,. total populations, nations, geographical areas or between peoples of different races, colours, and cultures. As Blauner, for example, expresses it:3

The colonial order in the modern world has been based on the dominance of White Westerners over non-Westerners of colour: racial oppression and the racial conflict to which it gives rise are endemic to it, much as class exploitation and conflict are fundamental to capitalist societies.

Secondly, the colonial relationship is characterized, in a general way, as involving domination, oppression, and exploitation. Again Blauner provides a convenient statement:4

Colonialism traditionally refers to the establishment of domination over a geographically external political unit, most often inhabited by people of a different race and culture, where this domination is political and economic, and the colony exists subordinated to and dependent upon the mother country. Typically the colonisers exploit the land, the raw materials,. the labour, and other resources of the colonised nation; in addition a formal recognition is given to the differences in power, autonomy and political status, and various agencies are set up to maintain this subordination.

It is these two features which constitute the core of internal colonialism, that is, of colonialism internal to a particular society. Casanova, for example, states: `Internal colonialism corresponds to a structure of social relations based on domination and exploitation among culturally heterogeneous, distinct groups.5 And Tabb puts it thus: 'The economic relations of the ghetto to White America closely parallel those between third world nations and the industrially advanced countries.'6

With regard to South Africa, the argument lies been formulated as follows by the South African Communist Party:7

South Africa is not a colony but an independent state. Yet masses of our people enjoy neither independence nor freedom. The conceding of independence to South Africa by Britain in 1910 ... was designed in the interests of imperialism. Power was transferred not into the hands of the masses of the people of South Africa, but into the hands of the White minority alone. The evils of colonialism, insofar as the non-White majorit,' was concerned, was perpetuated and reinforced. A new type of colonialism was developed, in which the oppressing white nation occupied the same territory as the oppressed people themselves and lived side by side with them.

On one level, that of White South Africa, there are all the features of an advanced capitalist state in its final stage of imperialism. There are highly developed industrial monopolies, and the merging of industrial and finance capital. The land is farmed along capitalist lines, employing wage labour, and producing cash crops for the local and export market. The South African monopoly capitalists ... export capital abroad. But on another level, that of `non-White South Africa', there are all the features of a colony. The indigenous population is subjected to national oppression, poverty and exploitation, lack of all democratic rights and political domination by a group which does everything it can to emphasize and perpetuate its alien `European' character. The African Reserves show the complete lack of industry, communications, transport and power resources which arc characteristic of ... territories under colonial rule.... Typical, too, of imperialist rule, is the reliance by the state upon brute force and terror, and upon the most backward tribal elements and institutions which are deliberately and artificially preserved. Non-While South Africa is the colony of White South Africa itself. [My emphasis throughout.]

The characterization of colonial relations as occurring between nations, countries, races and so on finds its most rigorous formulation in the work of Blauner. He states:8

Unfortunately, social science lacks a model of American society and its social structure in which racial division and conflict are basic elements rather than phenomena to be explained (or explained away) in terms of other forces or determinants. To close this theoretical gap, in part, I rely on the framework of colonialism in the present study.

And again: `racism and racial oppression are ... independent dynamic forces (not ultimately reducible to other causal determinants)'.9

Since. in this argument. Blauner asserts the independence and irreducibility of `race' (although curiously he also argues that `racism developed out of the same historical situation [as colonialism] and reflected a world economic and power stratification')10 he cannot conceptualize the relationship of `race' to the social structure and we are, therefore, left only with racial and ethnic groups abstracted out of the social formation. Indeed, Blauner appears to be aware of this problem, but to have no answer to it:11

Yet the colonial perspective cannot by itself provide the theoretical framework necessary to grasp the complexities of: race relations and social change in America. When the colonial model is transferred from the overseas situation to the United States without substantial alteration, it tends to miss the total social structure, the context of advanced industrial ... This suggests a major defect of my study. It lacks a conception of American society as a total structure beyond the central significance that I attribute to racism.

In two other versions of the theory of internal colonialism no assumption of the independance of race, ethnicity, or culture is made, but the analysis nevertheless does not go beyond Blauner. In the first of these versions, the contrast is implicitly drawn between capitalist societies which are culturally, ethnically, and racially homogeneous, and in which relations of class exploitation are dominant, and those societies in which both capitalist exploitation and internal colonial relations exist side by side (with the latter frequently dominant). Two questions arise here. First, what is the relationship between the system of class exploitation and domination and the relations of racial, ethnic, cultural or national exploitation and domination characteristic of internal colonialism? Second, in what way 'does internal colonial exploitation differ from class exploitation? On these questions the theory is silent. Thus Casanova, for example, asserts:12

The colonial structure and internal colonialism are distinguished from the class structure since colonialism is not only a relation of exploitation of the workers by the owners of raw materials or of production and their collaborators, but also a relation of domination and exploitation of a total population (with its distinct classes, proprietors, workers) by another population which also has distinct classes (proprietors and workers).

While this passage is useful for the way in which it points to the, or rather to one of the, modes of class exploitation (that is, the appropriation of surplus value) entailed in imperialism, it nevertheless fails to link `the exploitation of the workers' to the exploitation of one `total population' by another and nor does it explain the meaning of exploitation in the latter case.

Similarly, Johnson states: `The population of internal colonies is subject to discriminatory practices over and above those characteristic of relations between dominant classes and underclasses.'13 But, despite a lengthy discussion of `Class relations and colonial relations', he is unable to clarify the relationship between `discriminatory practices' and class relations or the differences between the two. He can only assert the differences:'14

The major differences in the relations between the dominant classes and institutions of society and marginal under-classes on the one hand, and internal colonies [An internal colony constitutes a society within a society based upon racial, linguistic and/or marked cultural differences as well as differences of social class. H.W.J on the other hand, revolve around different means of social control. It is important to emphasize that till the classes of the dominant society rest upon the colonial population.

What appears from the above passages is that no attempt is made to identify the specific mode of exploitation and domination characteristic of internal colonialism which purports to differentiate it from class exploitation and domination. Instead, there is a general reference to exploitation, used in a descriptive sense, and to undefined states of racial or ethnic oppression and these arc in no way linked to the system of class exploitation. The consequence of this is that, as in Blauner's analysis, internal colonial relations are not only left obscure but are said to hold between racial, ethnic, and cultural groups which are analysed as if they are autonomous of the total social structure.

In the second version of the theory in which no claim is made that race and ethnicity are independent of the social structure, a similar result is arrived at by a different path. The analysis of the South African Communist Party is the case in point here. In this case, as is clear from the passage from the Party's Programme quoted above, class relations are simply assimilated to race relations. Thus `White South Africa' is identified with the `capitalist state' and the capitalist system, while 'non-White South Africa' is identified with `the colony'. From this point on the analysis of class relations gives way to the description of White domination and exploitation of Blacks in terms of the internal colonial analogy.

It is possible to argue that where there is a complete coincidence between race and class the concepts defining the relationship between classes may be utilized in defining the relationship between races or ethnic groups.15 Where this is not the case (and it is not the case in South Africa), the substitution of racial groups for classes in the analysis requires a specification of the nature of the relationship between the former groups. Such a specification is not forthcoming in the Programme and, consequently, we are once more left with racial groups which stand in a vaguely defrned relationship of domination and subordination and whose relations to the class structure are left completely unanalysed.

This conclusion is underlined by the fact that the. characterization of internal colonialism as a relation between racial or ethnic entities necessarily involves, despite the recognition that these entities themselves have complex class structures, an analysis which treats these categories as homogeneous. If this were not the case, if the analysis were to be made in terms of class relations, then the internal colonial relation could no longer be conceptualized as a relation between racial, ethnic, etc., groups. But the consequence of the failure to relate classes within racial or ethnic groups to the class structure of the society as a whole, is that racial or ethnic entities are treated abstractly and as if their internal class structures are irrelevant to their existence as groups and to 'their political and ideological practices. Indeed the simultaneous recognition of the diversity of classes within racial groups and their conceptualization as homogeneous categories of internal colonialism forces the Programme of the Communist Party into a contradictory position: `Power [in 1910] was transferred not into the hands of the masses of people of South Africa, but into the hands of the White minority alone.'16 This is followed with:17

All Whites enjoy privileges in South Africa. They alone can vote and be elected to parliament and local government bodies. They have used this privilege to monopolise nearly all economic, educational, cultural and social opportunities. This gives the impression that the ruling class is composed of the entire White population. In fact, however, real power is in the hands of the monopolists who own and control the mines, the banks and finance houses, and most of the farms and major industries.

The theory of plural society

The unexplained autonomy of racial, ethnic, and cultural groups and the obscurity of the relationships between them which we have shown to be the 'outcome of the theory of internal colonialism, brings this theory within a .conceptual. framework which is similar to that of the theory of plural society. Despite the very different origins of these theories they suffer from identical analytical limitations. The nature of these limitations, which have been briefly adverted to above, may be elaborated and made clearer by a discussion of the main propositions of plural theory.

It is well known by now that Furnivall 18 was the first to characterize a colonial society as `plural' and since then, in one sense or another, the notion has been widely used in the analysis of certain societies outside West Europe and the United States, particularly by M. G. Smith with reference to the West Indies.19 What was the intellectual context in which this concept emerged and became increasingly utilized by Western social scientists in their accounts of non-Western societies? The answer to this question sheds some light on one of the central theoretical difficulties inherent in the attempt to develop a concept of plural society.

Throughout the 1950s and 1960s structural-functionalism, as developed by Talcott Parsons, was the most widely accepted (among social scientists that is) sociological theory, at least in the United States and possibly Britain. The basic propositions of this theory,'' in so far as they arc relevant for present purposes, may be summarized as follows: all social systems are made up of the interaction of individuals. Such interaction is not random, on the contrary, it is highly structured. That structure is produced consensually through the values and norms of the common cultural system, which deter mine the actions of individuals. The differentiation of individuals into groups, and, most irn portantly, their allocation to institutionalized roles (including authority or power positions) is carried out in accordance with the common, accepted norms. Because cohesion is produced through voluntary adherence to the rules, conflict is at a minimum;{ in any event conflict is not systematically generated, it is managed in accordance with the norms and may even function to enhance cohesion.

Despite the fact that it was claimed that these propositions constituted the core of a general sociological theory, it soon became[ apparent that the theory was actually regarded as appropriate only to certain societies which, in fact, were believed to be highly stable,,[ relatively free of conflict, and consensual. That is to say, the general theory turned out to be a specific model for the analysis of social: systems which were thought to be integrated around a common` value system. It need hardly be added that the social systems to which' the model was held to be appropriate were the 'advanced industrial societies' of Western Europe and the United States.

By contrast, colonial and former colonial societies were seen by Furnivall and later by Smith to be characterized by conflict, cultural heterogeneity and' an absence of common values. Not consensus, but domination, is asserted to be the basis of social order and cohesion ,ii) such societies. If consent is the basis of social solidarity in Western societies then clearly a different 'model' had to be devised for societies held together largely by coercion. At this point the 'conflict theorists' enter the stage with the various. 'theories' of plural society.

While the construction of two quite different models in this way seems to imply the assumption that some societies are totally free of conflict and bound together solely by consensus, while others are racked with conflict and bound together only by coercion, in fact it is clear that no such assumption is made. Thus as Lockwood has pointed out in relation to Parsons:20

The presence of a normative order, or common value system does not mean that conflict has disappeared, or been resolved in some way. Instead, the very existence of a normative order mirrors the potentiality of conflict.

Both Rex and Van Den Berghe seem to adopt a similar position, as the following passages show:21

Could not a similar plan of study [i.e. the plural model. H.W.] be employed in the analysis of British society? For myself I am prepared to accept that the scheme is less applicable in a society like our own than in the colonial situation. Class divisions in Britain are certainly not as far-reaching as are the differences between race groups in the colonies. Nonetheless I think that our analysis of our own society would be enriched by studies which started by assuming conflicting valuations rather than some sort of social consensus.

I prefer to regard pluralism as a variable, and to include cases of stratification based on 'race', caste, estate, or class ... as instances of pluralism, even though the constituent groups share the same general culture.22

Since both consensus and conflict and 'plural' groups are common to all societies, what determines the decision to apply a plural conflict model rather than a consensus model? It would seem that the decision is not based on any conceptual distinction but on ad hoc empiricist judgments. Thus in the passages cited we have nothing more than the preference of the authors (Van Den Berghe: 'I prefer to regard pluralism as a variable'; Rex: 'For myself I am prepared to accept that the scheme is less applicable in a society like our own') based simply on an, assessment of the degree of conflict which manifests itself in any period. This is hardly a satisfactory basis for deciding what 'model' ought to be applied to the analysis of a particular society.

It is possible to argue, however, that the pluralism of a society does not. rest merely on the degree of conflict but also on the nature of the groups and the content of the conflict between them. While it is not always clear what the different plural theorists are arguing, it would seem that plural societies are sometimes differentiated from others on the grounds that the salient groups in the former societies are racial, cultural, religious, national but not classes or strata. This seems to have been Rex's argument at one stage and Van Den Berghe's in his South Africa, A Study in Conflict.23

Two problems arise here. First, since such ethnic, cultural, etc., divisions are obviously found in societies not normally regarded as plural or internal colonial (e.g. Flemish people and Walloons in Belgium; French and English in Canada; Catholic and non-Catholic in France and Holland; Irish and Welsh in Britain) it is difficult to see how plural societies or internal colonial societies can be distinguished on this basis. But second, and more importantly, to treat such groups as autonomous and as the salient groups in the society has the consequence of excluding from the analysis precisely those other structures and relations (in particular the mode of production, the class structure, and class relations) which axe necessary to an explanation of the nature and relationships of those groups. The point is that to base an analysis on the criteria (race, religion, etc.) by which groups define themselves and the conflict between them is to take as given precisely what requires explanation. For what needs to be accounted for is why these particular groups come into existence and into conflict with one another. This requires an analysis of the conditions which generate particular conflicts and which affect their nature and intensity.

Therefore, what is needed is, on the one hand, a description of the ideology and political practices of the ethnic, racial, and national groups and, on the other, an analysis of how they relate to the mode of production and social formation in which they are located. It is E thus insufficient to stop at the first stage because this is too abstract from the social totality in which the groups are embedded and which explains them.

Some writers, particularly M. G. Smith and Van Den Berghe, have attempted to define the nature of plural groups more rigorously. Although their conceptions differ in a number of respects (which need not be discussed here) they are in fundamental agreement that there are two basic features which differentiate plural groups and, therefore, plural societies from others. In Van Den Berghe's formulation these features are:24

(1) segmentation into corporate groups that frequently, though not necessarily, have different cultures or subcultures; and
(2) a social structure compartmentalized into analogous, parallel, non-complementary but distinguishable sets of institutions.

In this view, the corporate groups are incorporated around ; `complementary but distinguishable institutions'. These institutions, and therefore the groups, operate independently of the other groups and institutions in the society. If this were not the case, that is, if they-the institutions and groups-were integrated with one another, the society could not be regarded as. plural.

Given the plurality and autonomy of these institutions and groups, on what basis can they be regarded as constituting a society? Van Den Berghe and Smith give rather different answers to this question but both answers are open to the same objections. According to Van Den Berghe:25

At one extreme, societies characterized by a high degree of pluralism are integrated only through a set of central political institutions controlled by the dominant group, and of economic institutions in which members of different groups interact asymmetrically.

And Smith argues that26

Pluralism is a condition in which members of a common society are internally distinguished by fundamental differences in their institutional practice. Where present, such differences are not distributed at random; they normally cluster, and by their clusters they simultaneously establish deep social divisions between them. The prevalence of such systematic dissociation between the members of institutionally distinct collectivities within a single society constitutes pluralism.... In a plural society where the rulers form a culturally distinct numerical minority, the aggregate depends for its formation, unity, order, and form primarily on the concentration ... of regulative powers by the ruling section through the political framework.

It is to be noted that for Smith the pluralism relates only to the basic institutional system which `embraces kinship, religion, property and economy, recreation, and certain sodalities.... It does not normally include government.'27

The criticisms already made of plural `theory' apply equally to these conceptions and need not be repeated here. There is, however, an additional point to be made. It is one thing to argue that groups within a society may, in certain respects, follow different institutional practices; it is quite another thing to suggest that these institutional practices are independent of one another and of the structure of the social formation. On what basis, for example, can it be maintained that only the political system links the plural institutional orders so as to enable a plural society to be referred to as a society whereas the economy does not? Again, once it is argued that the political and the economic institution do hold the society together despite the plurality of the institutions, how can it be maintained that the other institutional orders remain autonomous? The assertion of institutional `segregation' and autonomy presses plural theory to its logical conclusion and emphasizes the abstractness of its formulation. This is so since we are asked to understand institutions independently of any relationship outside of their own `boundaries'.

One result of this is that there is no way in which it can be meaningfully asked (let alone be answered) within this `theory': how can the development and maintenance of distinguishable institutional practices be explained?

The above discussion of internal colonialism suggests that this ; theory presents socicty as a composite of class relations and ethnic, race, cultural, or national relations. To this extent the theory may be distinguished loin convcntional analyses of race, ethnic and similar relations. since in the latter approach these relations are accorded sole salience. On the other hand, the theory of internal colonialism is unable to explain the relationship between class relations and race or ethnic,etc., relations. As a consequence, the latter' relations come once more to he treated as autonomous and in isolation from the class relations. To this extent there is a close convergence between internal colonialism and conventional race relations' theory, more particularly, as our analysis has shown, when the latter is based on a plural modcl. of society.

For this reason, interned colonialism remains vulnerable to the analytical limitations which we have shown to apply to plural theory.

Imperialism and modes of production

The obfuscating consequences of an analysis in terms of racial, ethnic,, cultural, or national entities is nowhere clearer than in the use of the notion of exploitation to describe relations between such entities., The reason for this is that while the concept of exploitation can have! a rigorous and explicit meaning in defining class relations, it becomes a vague, descriptive term in the characterization of relations between; such entities as racial, national, or cultural groups. Bettelheim; in, commenting on the notion of the exploitation of the `poor countries' by the 'rich ones', has made the same point, in relation to `normal ' colonialism. He states :28

Because the concept of exploitation expresses a production relation-production of surplus labour and expropriation of this by a social class-it necessarily relates to class relations (and a relation between `countries' is not and cannot be a relation between classes).

He argues that 'it is not possible to give a strict meaning to the, notion of exploitation of one country by another country', and he concludes:29

With only slight amendments this passage applies equally to the case of internal colonialism. Thus, we may say, that in order to avoid the abstraction involved in treating racial or ethnic groups as undifferentiated and homogeneous, we must think of each such group as having a `specific structure, in particular because of the existence of classes with contradictory interests'. It follows that the concrete social totality is constituted by the complex articulation of class relations within racial or ethnic groups, as well a$ the relation of classes across these groups together, we may add, with the ideological and political practices which `fit' these relationships.

This consideration leads directly to the crucial further question of historical specificity. It should be clear from what has so far been argued that the concept of colonialism upon which the internal colonial thesis is based is extremely vague and unspecific. In part, this is due to the failure to distinguish between forms of colonial, political, ideological, and cultural domination and modes of imperialist economic exploitation. In turn this" conflation stems from the failure to distinguish differing modes of imperialist economic exploitation with the result that different forms of colonial domination .cannot be explieitly related to different modes of exploitation.

More specifically, much of the analysis of imperialism and underdevelopment (and of internal colonialism) has. been based on the assumption. that in the era of capitalist imperialism, exploitation everywhere takes place according to a single, invariant mode. There are two variants of this argument, but both contend that capitalist relations have, as Laclau puts it, `effectively and completely penetrated even the most apparently isolated sectors of the under! developed world'.30

In one variant capitalism is equated with commodity exchangewith, the market economy-and consequently the participation of the underdeveloped world in the market is construed as evidence of the total transformation of the indigenous economies into capitalist economies, albeit subordinate ones. This is the position advanced by Frank in his analysis of Latin America. But, Laclau argues,31

Frank's theoretical schema involves three types of assertion: 1. Latin America has had a market economy from the beginning; 2. Latin America has been capitalist from the beginning; 3. the dependent nature of its insertion into the capitalist world market is the cause of its underdevelopment. The three assertions claim to refer to a single process identical in its essential aspects from the 16th to the 20th century.

The consequence of this, as Laclau has shown, is that it,becomes impossible to define 'the specificity of the exploitative relationship' in operation at a specific moment and this flows directly from Frank's failure to base his analysis on the concept of relations of production. Thus, an analysis based on the concept of relations of - production would have shown, in the particular case of Latin America, not the complete penetration of capitalism but rather that - the32

pre-capitalist character of the dominant relations of production in Latin America was not only not incompatible with production for the world market, but was actually intensified by the expansion of the latter. The feudal regime of the haciendas tended to increase its servile exactions on the peasantry as the growing demands of the world market stimulated maximization of their surplus. Thus, far from expansion of the external market acting as a disintegrating force on feudalism, its effect was rather to accentuate and consolidate it.

It is thus clear from Laclau's,argument that it cannot be assumed from the emergence of a dominant capitalist market, that non- F capitalist economies which participate in that market are, thereby, automatically transformed into capitalist modes of production. ' In the second variant of this argument, the analysis is, indeed, based on the concept of the mode of production. In this case it is assumed that the effect of the emergence of capitalism as a dominant mode of production is the necessary and rapid disintegration of non-capitalist productive relations. This view seems to be based on Lenin's discussion of imperialism and Marx's analysis of primitive accumulation.

In Imperialism: lime highest Stage of Capitalism Lenin stated:33

The export of capital influences and greatly accelerates the development of capitalism in those countries to which it is exported. While, therefore, the export of capital may tend to a certain extent to arrest development in the capital exporting countries, it can only do so by expanding and deepening the further development of capitalism throughout the world.

In Capital Marx formulated the notion of primitive accumulation in the following terms.34

The capitalist system presupposes the complete separation of the labourers from all property in the means by which they can realize their labour.... The process, therefore, that clears the way for the capitalist system, can be none other than the process which takes away from the labourer the possession of his means of production; a process that transforms on the one hand, the social means of subsistence and of production into capital, on the other, the immediate producers into wage­labourers. The so-called primitive accumulation, therefore, is nothing else than the historical process of divorcing the producer from the means of production.

However, 'either Lenin's general characterization of the development of -capitalism through imperialism in the era of monopoly capitalism, nor Marx's theoretical analysis of the constitution of capitalism through primitive accumulation, can be construed as concrete historical accounts of the actual progression of imperialism and capitalism either within particular social formations or on a world scale. To interpret Marx and Lenin in this way, as, for example, the Programme of the South African Communist Party clearly does, is precisely to obliterate the analysis of the relationship of capitalism with non-capitalist modes of production and thereby to exclude the possibility of analysing the specificity of the exploitative relations which concretely characterize social formations.

In fact, the relationship of capitalist to non- or pre-capitalist modes of production may vary in a number of- ways and for different reasons. Thus, in one place the relationship of capital to a non­capitalist mode of production may revolve around the extraction in different ways-by plunder, or' the exchange of non-equivalents or by means of the process of price formation-of the commodities produced by the latter. Geertz's study of Inner Java is an example of this. At another place, the main focus of the relationship may be on the extraction, not of the product, but of labour-power. South Africa, as I will show below, is an example of this type of relationship. While in both of these cases the associated political policy turns on the domination and preservation of the non-capitalist societies, in other instances the particular mode of economic exploitation may be accompanied by a policy aimed at or having the effect of destroying the non-capitalist -societies.

The relevance of this for the present discussion may be clarified by the following elaboration. In the course of its development, the capitalist mode of production enters into relationships with other, non-capitalist, systems of production-the very origins of capitalism in the interstices of feudalism testifies to this. Relations with other modes of production first occur within the boundaries of the nation state. First with trade and later with the development of monopoly capitalism and the export of capital, capital increasingly enters into new relationships with other, non-capitalist, modes of production,. beyond the borders of the nation-state. These relations, which are exploitative in the strict sense of the term-they involve directly or indirectly the extraction of the surplus from the direct producers characterize, in general, the period of capitalist imperialism. These relations of imperialism are constituted within a particular context of political domination and arc sustained and supported by a mode of ideological and political practice which varies. with the mode of exploitation. But, as Lenin pointed out, both imperialism and colonialism undergo historical changes:35

Colonial policy and imperialism existed before the latest stage of capitalism, even before capitalism. Rome, founded on slavery, pursued a colonial policy and practised imperialism. But `general' disquisitions on imperialism which ignore, or put into the background, the fundamental differences between socio-economic formations, inevitably turn into the most vapid banality.... Even the capitalist colonial policy of previous stages of capitalism is essentially different from the colonial policy of finance capital.

In certain conditions of imperialist development, ideological and political domination tend to be expressed not in terms of the relations of class exploitation which they must sustain but in racial, ethnic, national, etc., tcrms and, in all cases, this is related to the fact that the specific anode of exploitation involves the conservation, in some form, of the non-capitalist modes of production and social organization, the existence of which provides the foundation of that exploitation. Indeed, it is in part the very attempt to conserved an control the non-capitalist societies in the face of the tendency of capitalist development to disintegrate them and thereby to undermine the basis of exploitation, that accounts for political policies and ideologies which centre on cultural, ethnic, national, and racial characteristics.

In certain circumstances capitalism may, within the boundaries of a single state, develop pr, -dominantly by means of its relationship to non-capitalist modes of production. When that occurs, the mode of political domination and the content of legitimating ideologies assume racial, ethnic, and cultural forms and for the same reason as in the case of imperialism. In this case, political domination takes on a colonial form, the precise or specific nature of which has to be related to the specific mode of exploitation of the non-capitalist society.

These points can be illustrated and perhaps made clearer by an analysis of internal colonialism in South Africa.

Internal colonialism in South Africa

It was suggested in the previous section that one important economic basis of colonial domination is the economic relationship which im-perialism establishes between capitalista nd non-capitalist modes of production. I also argued that that relationship may take different forms.

In volume II of Capital, in dealing with the circuit of capital and in particular the commodities which comprise the means of production Marx stated:36

Within its process of circulation, in which industrial capital functions either as money or as commodities, the circuit of industrial capital whether as money-capital or as commodity capital, crosses the commodity circulation of the most diverse modes of social production, so far as they produce commodities. No matter whether commodities are the output of production based on slavery, of peasants ... of state enterprise ... or of half-savage hunting tribes, etc.-as commodities and money they come face to face with the money and commodities in which the industrial capital presents itself and enter ... into its circuit The character of the process of production from which they originate is immaterial. They function as commodities in the market, and as commodities they enter into the circuit of industrial capital as well as into the circulation of the surplus value incorporated into it.... To replace them [i.e. the commodities entering the capitalist circuit in the above manner] they must be reproduced and to this extent the capitalist mode of production is conditional on modes of production lying outside of its own stage of development.

While in the above passage Marx's remarks are restricted to commodities which are also means of production, it seems clear that they apply equally to labour-power which is physically produced in a non-capitalist mode of production but which is converted into a commodity by its appearance on the capitalist labour market.

It is this feature, the introduction into the capitalist circuit of production of labour-power physically produced in a non-capitalist economy, that denotes one important feature of imperialism. This `crossing' of different modes of production modifies the relationship between wages and the cost of reproducing labour-power in favour of capital. It is precisely: this relationship which is the foundation of `internal colonialism' in South Africa.

In fact, the South African social formation is made up of several modes of production but it is not possible in this paper to discuss all of these or to explore the complex relations between them. For present purposes the analysis may be restricted to the relationship between the dominant capitalist economy and the mode of production in the African areas (Reserves).

The capitalist mode of production in South Africa (as elsewhere) is one in which

(1) the direct labourers, who do not own the means of capitalist production, sell their labour-power to the owners of the means of production who are non-labourers; and
(2) the wage the labourer receives for the sale of his labour-power for a certain pcriod is only a portion of the value created by him during that period, the balance being appropriated as unpaid labour (surplus value) by the owners of the productive means.'

This second condition is, of course, related to Marx's conception of labour-power as a commodity and expresses the specific form in which the surplus is extracted from the direct producers in the capitalist mode of production.

The ratio between the surplus product and the necessary product which accrues to the labourer in the form of wages is, in Marx's terms, the rate of surplus value. This rate will obviously vary !in accordance with changes in the distribution of the product between necessary and surplus labour. The greater the proportion of the working day devoted to necessary labour, the lower the rate of surplus value and consequently the rate of profit, all other things remaining equal. It follows that the conditions which determine the amount of time spent. on the necessary product are of crucial importance in capitalist production.

In general commodities exchange at their value. The value .of labour-power is determined in the same way as that of other commodities-by the amount of socially necessary labour time which has been expended in its production. As Marx put it:37

The value of labour power is determined, as in the case of every other commodity, by the labour time necessary for the production and consequently also the reproduction of this special article.... Given the individual, the production of labour power consists in his reproduction of himself or his maintenance. Therefore the labour-time requisite for the production of labour-power reduces itself to that necessary for the production of... the means of subsistence; in other words, the value of labour-power is the value of the means of subsistence necessary for the maintenance of the labourer.

The subsistence necessary for the reproduction of labour-power is extended in at least two ways by Marx: `the sum of the means of subsistence necessary for the production of labour-power must include the means necessary for the labourers' substitutes, that is, his children';38 in addition: 'The expenses of ... education ... enter pro tanto into the total value spent in its production.'39

There are a number of ways in which the proportion of the working day which is allocated to necessary labour may be decreased. Thus, for example, the value of labour-power may be decreased or, again, the length of the working day may be increased and most importantly for the present argument, labour-power may be ac quired at a cost below its value.

As Meillassoux has pointed out, the means of subsistence acquired by the labourer can be divided into two, parts-the direct wages paid to the worker in and during employment, and indirect wages which he receives in the form of social security benefits, for example, unemployment payments, family allowances, health services, education and so on.40 In its most advanced form indirect wages are institutionalized in the social welfare arrangements of the welfare state, but obviously, these arrangements are the outcome of a lengthy historical process.

Under certain conditions the capitalist mode of production is able to avoid, to a greater or lesser extent, the payment of indirect wages; that is, it is obliged to pay only the immediate sustenance of the labourer but it can avoid paying for his subsistence during unemployment, or for the subsistence of children or costs of education, etc.

The most important condition enabling capitalism to pay for labour-power below its cost of reproduction in this way is the availability of a supply of labour-power which is produced and reproduced outside the capitalist mode of production.

In South Africa this condition was (and still is, although to a decreasing extent) met by the presence of a non-capitalist mode of agricultural production in the areas of African concentration (particularly, but not exclusively, in the Reserves). In this mode of production land is held communally by the community and worked by social units based on kinship, and the product of labour is `pooled' and then redistributed directly by means of an allocation through the kinship units in accordance with certain rules of distribution. Alternatively, where the land is held in individual tenure, it is worked by a kin group between the members of which, certain reciprocal obligations of support are in. force. Whatever the mode of production, however, the crucial element is the existence of reciprocal, obligation of support.

Given the "nature of the relations of production and distribution in such an economy, the potentiality exists of utilizing labour-power drawn from it into the capitalist sector without fundamentally alter ing those relations. Thus, as Meillassoux argues, if the necessary sub sistence for the entire year can be produced by labour which is limited to a part of the year, then labour-power will be potentially available to the capitalist sector for the remainder of the year. This potential labour-power can be brought into the circuit of capitalist production provided That the capitalist sector `finds the means to extract it practically, without the direct intrusion of capital into the self-sustaining sector', an intrusion which would destroy the relations of production and, therefore, the basis of the production of labou-rpower in the sector external to capitalism.41 It is presumably in this kind of situation that various 'political' measures may be taken to force labour-power onto the market. On the other hand, if the subsistence produced during the productive season is insufficient to meet all necessary needs then, provided there are no actual productive possibilities beyond the period of agricultural production, the propulsion of labour-power onto the market may occur through the operation of economic forces.

In either case, the significant aspect is that the capitalist sector benefits from the means of subsistence produced in the non-capitalist mode of production to the extent that it is relieved of paying a portion of the necessary means of subsistence by way of indirect wages. This, as I have shown in a previous paper, has the important effect of raising the rate of surplus value.42 The uniqueness or specificity of South Africa, in the period of capitalism, lies precisely it I this: that it embodies within a single nation-state a relationship characteristic of the external relationship between imperialist states and their colonies (or neo-colonies).

Bettelheim has pointed out that43

Inside social formations in which the capitalist mode of production is dominant, this domination mainly tends to expanded reproduction of the capitalist mode of production, that is, to the dissolution of the other modes of production and subsumption of their agents to capitalist production relations. The qualification 'mainly' indicates that this is the predominant tendency of the capitalist mode of production within the social formations under consideration. However, this predominant tendency is combined with another secondary tendency, that of 'conservation-dissolution'. This means that within a capitalist social formation the non-capitalist forms of production, before they disappear are 'restructured' (partly dissolved) and thus subordinated to the predominant capitalist relations (and so conserved).

Within the advanced capitalist states themselves the dominant tendency more or less rapidly brought about the complete or almost complete dissolution of the non-capitalist relations of production. The explanation for this, in each society, and the specification of the processes involved require, of course, their own historical analysis.

In South Africa, on the contrary, the dominant tendency has been inhibited by the secondary tendency of 'conservation-dissolution'.

That is to say, the tendency of capital accumulation to dissolve the very relationship (with the non-capitalist economies) which makes that accumulation possible (at a particular rate) is blocked by. the contradictory tendency of capital to conserve the relationship and with it the 'non-capitalist. economies, albeit in a restricted form for the reasons already outlined.

The political expression of this imperialist-type relationship takes on a colonial form. This is so because, at one level, the conservation of the non-capitalist modes of production necessarily requires the development of ideologies and political policies which revolve around the segregation, and preservation and control of African 'tribal' societies. The ideological focus, it must be stressed, is always necessarily on the 'racial' or 'tribal' or 'national' elements, precisely because of the 'tribal' nature of what is being preserved and controlled.44

:So, too, the policies pursued and the laws passed must have the :same focus. Therefore, the attempt to conserve these societies in the face of disruptive tendencies centres on guaranteeing the availability of some land (1913 Land Act) to the 'tribe', the preservation of the ;social and political organization of the 'tribe', and thus the retention of much 'Native' law and so on. At the same time the disruptive tendencies create problems of control for the capitalist state and these are met by a vast superstructure of administrative control both through the state and through 'tribal' authorities. The counterpart of, all this is the. structure of domination exercised over the African labour force through the pass laws, urban area§ acts, police, Bantu administration department, and so on.

In the paper referred to above I showed concretely and in some detail how the specific changes in ideology and political policy the transition from 'Segregation' to 'Apartheid'-reflected changing relationships between the African redistributive economics and the capitalist sector with particular reference to the supply of cheap labour-power. In brief the preservation of the conditions (migrantlabour, fixed land area, low capital investment in African agriculture)which enable labour-power to be extracted from the African societies serve to destroy the productive capacity of these societies (given the increase in population, and consequent over-population on the fixed land means, backward farming methods, etc.). The diminution of the product from. these Reserve economies generates rural impoverishment and, also, 'in the absence of the assumption by the capitalist sector of responsibility for indirect wages, extreme urban impoverishment. The consequence is increasing African pressure on wages and on: rural conditions, pressure which becomes elaborated into an assault on the whole political and economic structure in the 1940s and 1950s. Apartheid may be seen as the attempt of the capitalist state to maintain the system of cheap migrant-labour, in the face of this opposition, by means of the erection-of a 'perfected' and 'modernized' apparatus of political domination.

Although, in this section, the focus has been on the extraction of labour-power by a capitalist mode of production from non-capitalist productive systems, it must be stressed that it is not intended to suggest that this is the only form such a relation may take. I indicated above that imperialism may also operate by appropriating the product of non-capitalist societies or, indeed, by destroying those societies such that the producers are 'freed' of the means of production. These types of relations give rise, to varying forms of political domination 45


As I indicated earlier, the analysis presented here is partial in that it does not deal with the full complexity of the relationships between all the different modes of production in South Africa. Nevertheless, sufficient has been said to show the fundamental difference between this approach and those criticized in the chapter. Although only briefly, specific racial ideologies and political policies, despite their pervading racial content, have been shown to reflect and to relate to a specific reality `external' to themselves-the modes of production and their interrelationships. In particular, the point has been stressed that specific modes of political domination which assume a racial or ethnic and, therefore, a colonial rather than a class form have to be analysed in terms of the specific relations of economic exploitation.


1 See, for example: R. Blauner, 'Internal colonialism and ghetto revolt', Social Problems, vol. 16, no. 4, 1969, pp. 393-408, and Racial Oppression in America, Harper & Row, New York, 1972; G. M. Carter, T. Karis and N. M. Stultz, South Africa's Transkei, the Politics of Domestic Colonialism, Heinemann, London, 1967; P. Gonzalez Casanova, 'Internal colonialism and national development', Studies in Comparative International Development, vol. 1, no. 4, 1965; A. G. Frank, Capitalism an r Underdevelopment in Latin America, Monthly Review Press, New York, 1967; A. Lerumo, Fifty Fighting Years, Inkululeko Publications, Johannesburg, 1971; L. Marquand, South Africa's Colonial Policy, South African Institute of Race Relations, Johannesburg, 1957: II. J. Simons and R. E. Simons, Class and Colour in South Africa, 1850-1950, Penguin, Harmondsworth, 1969; South African Communist Party, The Road to South African Freedom, Ellis Bowles, London, n.d.; R. Stavenhagen, 'Classes, colonialism, and acculturation: a system of inter-ethnic relations in Mesoamerica', Studies in Comparative International Development, vol. 1, no. 6, 1965.
2 Simons and Simons, op. cit., p. 610.
3 Blaunder, 1972, op. cit., pp. 12-13.
4 Blauner, 1969, op. cit., p. 395.
5 Gonzalez Casanova, op. cit., p. 33.
6 W. K. Tabb, The Political Economy of the Black Ghetto, W. W. Norton, New York, 1970, p. 15.
7 South African Communist Party, op. cit., pp. 25-6.
8 Blauner, 1972, op. cit., pp. 11-12.
9 Ibid., p. 2.
10 Blauner, 1969, op. cit., p. 395. 11 Blauner, 1972, op. cit., p. 13.
12 Gonzalez Casanova, op. cit., p. 33.
13 D. L. Johnson, 'On oppressed classes', in J. D. Cockroft, A. G. Frank and D. L. Johnson, Dependence and Underdevelopment, Doubleday,
New York, 1972, p. 282.
14 Ibid., p. 281.
15 Cf. Stavenhagen, op. cit.
16 South African Communist Party, op. cit., p. 25. 17 Ibid., p. 27.
18 J. S. Furnivall, Colonial Policy and Practice, Cambridge University Press, 1948.
19 M. G. Smith, The Plural Society in the British West Indies, University of California Press, Berkeley, 1965.
20 D. Lockwood, 'Some remarks on "The Social System"', British Journal of. Sociology, vol. .7, 1956, p. 137.
21 J. Rex, Race, Colonialism and the City, Routiedge & Kegan Paul, London, 1973, p. 256.
22 P. L.-Van Den Berghe, 'Pluralism and the polity: a theoretical exploration', in L. Kuper and M. G. Smith, eds, Pluralism in Africa, University
of California Press, Berkeley, 1969, p. 68.
23 J. Rex, 'The plural society in sociological theory', B JS, vol. 10, 1959, and 'The plural society: the South African case', Race, vol. 12, no. 4, 1971; University of California Press, Berkeley, 1967. :
24 Van Den Berghe, 1969, op. cit., p. 67.
25 Ibid., p. 71.
26 M. G. Smith, 'Institutional and political conditions of pluralism', in Kuper and Smith, op. cit., pp. 27, 32.
27 Smith, 1965, op. cit., p. 82.
28 C. Bettelheim, 'Theoretical continents', in A. Emmanuel, Unequal Exchange: A Study of the Imperialism of Trade, New Left Books,
London, 1972, p: 301. Ernesto Laclau in a personal communication has pointed out that, the unequal exchange of non-equivalents also
constitutes an exploitative relation. This, however, in no way affects the general point being made by Bettelheim since an analysis of the class structure of 'countries' in a relationship of unequal exchange is no less important than in the case of production relations.
29 Ibid., p. 300.
30 E. Laclau, 'Feudalism and capitalism in Latin America', New Left Review, no. 67, May-June 1971, p. 21.
31 Frank, op. cit.; Laclau, op. cit., p. 22.
32 Ibid., p. 30.
33 V. 1. Lenin, Selected Works, Lawrence & Wishart, London, 1969, p. 214.
34 K. Marx, Capital, Foreign Languages Publishing House, Moscow, 1962, vol. 1, p. 714.
35 Lenin, op. cit., p. 228.
36 Marx, op. cit., vol. 11, pp. 109, 110.
37 Ibid.. vol. 1, p. 171.
38 Ibid., p. 172.
39 Ibid.
40 C. Meillassoux, 'imperialism as a mode of reproduction of labour power', unpublished seminar paper, 1974. 41 Ibid.
42 1-1. Wolpe, 'Capitalism and cheap labour-power in South Africa: from segregation to apartheid', Economy and Society, vol. 1, no. 4, 1972.
43 Bettelheim, op. cit., p. 297. 1
44 Stavenhagen, op. cit., makes a similar analysis in relation to the r 'corporate' nature of the Indian Community but he does not articulate the relationship between this and the precise mode of economic exploitation. ,
45 I leave open whether the notion of 'internal colonialism' has any proper application in conditions of racial discrimination where, how- { ever, the internal relations within the society are overwhelmingly capitalist in nature, that is, where non-capitalist modes of production, if they exist at all, are marginal.