Mahmood Mamdani

Citizen and Subject (1996)


Conclusion: Linking the Urban and the Rural

As the dawn of independence broke on a horizon of internal conflict, reconsideration of the African colonial experience began. Could it be that the African problem was not colonialism but an incomplete pene­tration of traditional society by a weak colonial state or deference to it bv prudent but shortsighted colonizers? Could it be that Europe's mission in Africa was left half finished? If the rule of law took centuries to root in the land of its original habitation, is it surprising that the two sides of the European mission - market and civil society, the law of value and the rule of law - were neither fully nor successfully transplanted in less than a century of colonialism? And that this fragile transplant succumbed to caprice and terror on the morrow of independence?

With the end of the cold war, this point of view has crystallized into a tendency with a name, Afro-pessimism, and a claim highly skeptical of the continent's ability to rejuvenate itself from within. Whether seen as a problem of incomplete conquest or as one of unwise deference to traditional authorities, both sides of the Afro-pessimist point of view lead to the same conclusion: a case for the recolonization of Africa, for finish­ing a task left unfinished. Part of the argument of this book is that Afro-pessimism is unable to come to grips with the nature of the colonial experience in Africa precisely because it ignores the mode of colonial penetration into Africa.

Yet another set of questions coheres around a perspective that is not evolutionist but particularistic, whose impetus is not toward highlight­ing African “backwardness" but underlining its difference. That differ­ence is said to be the tendency to fragmentation and particularism, hith­erto held in check and obscured by a shared dilemma, colonial racism. Was not racism the general aspect of the African experience - its colonial and external aspect - and tribalism its particular, indigenous and inter­nal, aspect? Generally emancipated from racism with the end of colonial­ism, did not Africa once again come to be in the grip of a specifically African particularism: tribalism, ethnic conflict, and primordial combat? Another part of the argument in this book is that it is too naive to think of racism and tribalism as simple opposites, for alien (racial) domination was actually grounded in and mitigated through ethnically organized local power. In the colonial period, ethnic identity and separation were politically enforced. Although forged through colonial experience, this form of the state survived alien domination. Reformed after indepen­dence, purged of its racial underpinnings, it emerged as a specifically African form of the state.


Colonial Genesis

I have argued that to grasp the specificity of colonial domination in Af­rica, one needs to place it within the context of Europe's larger coloniz­ing experience. The trajectory of the wider experience, particularly as it tried to come to grips with the fact of resistance, explains its midstream shift in perspective: from the zeal of a civilizing mission to a calculated preoccupation with holding power, from rejuvenating to conserving so­ciety, from being the torch bearers of individual freedom to being custo­dians protecting the customary integrity of dominated tribes. This shift took place in older colonies, mainly India and Indochina, but its lessons were fully implemented in Africa, Europe's last colonial possession. Central to that lesson was an expanded notion of the customary.

Britain was the first to marshal authoritarian possibilities in indige­nous culture. It was the first to realize that key to an alien power's achieving a hegemonic domination was a cultural project: one of har­nessing the moral, historical, and community impetus behind local cus­tom to a larger colonial project. There were three distinctive features about the customary as colonial power came to define it. First, the cus­tomarv was considered synonymous with the tribal; each tribe was de­fined as a cultural group with its own customary law. Second, the world of the customary came to be all-encompassing; more so than in any other colonial experience, it came to include a customary access to land. Third, custom was defined and enforced by customary Native Authori­ties in the local state-backed up by the armed might of the central state.

To appreciate the significance of this, we need to recall only one fact. Although the use of force was outlawed in every British colony in the aftermath of the First World War (and in French colonies after the Sec­ond), this applied to the central state and usually to European officials supervising Native Authorities in the local state, but not to the Native Authorities. For this, there was one reason. So long as the use of force could be passed off as customary it was considered legitimate, and – to complete the tautology - force decreed by a customary authority was naturally regarded as customary. No wonder that when force was needed to implement development measures on reluctant peasants, its use was restricted to Native Authorities as much as possible. In the lan­guage of power, custom came to be the name of force. It was the halo around the regime of decentralized despotism.

The customary was never singular, but plural. As far as possible, every tribe was governed by its customary law. Europe did not bring to Africa a tropical version of the late-nineteenth-century European nation-state. Instead it created a multicultural and multiethnic state.1 The colonial state was a two-tiered structure: peasants were governed by a constella­tion of ethnically defined Native Authorities in the local state, and these authorities were in turn supervised by white officials deployed from a racial pinnacle at the center.

Another peculiarity of this form of the state was that the relation be­tween force and market was not antithetical. It was not simply that force framed market institutions. It was more that force and market came to be two alternative ways of regulating the process of production and ex­change. To the extent that the scope of the customary included land and labor, that of the market was limited. To flush either labor or its prod­ucts out of the realm of the customary required the use of force. Clearly, there was and is no particular and fixed balance between force and mar­ket. Its degree remains variable: the customary was never a Chinese wall keeping the tide of market relations at bay; nor was it of nominal signif­icance. The customary was porous. Within its parameters, market rela­tions were enmeshed with extra-economic coercion. Free peasants were differentiated, and those better off were shielded from the regime of force.

Postcolonial Reform and Variations

Characteristic of Afro-pessimism, whether in its left-wing or right-wing version, is a "roots of the crisis" literature that reduces the past to a one-dimensional reality. The result is a reconstruction of the past as if the only thing that happened was laying the foundations of a present crisis. The result is not an analysis that appropriates the past as a contra­dictory mix, but one that tends to debunk it.

The core agenda that African states faced at independence was three­fold: deracializing civil society, detribalizing the Native Authority, and developing the economy in the context of unequal international rela­tions. In a state form marked by bifurcated power, deracialization and detribalizanon were two aspects that would form the starting point of an overall process of democratization. By themselves, even if joined to­gether, they could not be tantamount to democratization. Together, this amalgam of internal and external imperatives signified the limits and possibilities of the moment of state independence.

Of this threefold agenda, the task undertaken with the greatest suc­cess was deracialization. Whether formulated as a program of "indigeni­zanon by mainstream nationalist regimes - conservative or moderate from Nigeria to Zaire to Idi Amin's Uganda, or as one of nationalization by radical ones, from Ghana to Guinea to Tanzania, the tendency everywhere was to erode racially accumulated privilege in erstwhile colonies. Whether they sought to Africanize or to nationalize, the historical legit­imacy of postindependence nationalist governments lay mainly in the program of deracialization they followed. The difference between them, however, was an effect of the strategy of distribution each one em­ployed. Whether the tendency was privatization or etatism, both strate­gies opened opportunities for nepotism and corruption, for clientelism.

In contrast to deracialization, the task undertaken with the least suc­cess was democratization. Key to democratization was the Native Au­thority in the local state: its detribalization would have to be the starting point in reorganizing the bifurcated power forged under colonialism. The failure to democratize explains why deracialization was not sustain­able and why development ultimately failed. Without a reform in the local state, the peasantrv locked up under the hold of a multiplicity of ethnically defined Native Authorities could not be brought into the mainstream of the historical process. In the absence of democratization, development became a top-down agenda enforced on the peasantry. Without thoroughgoing democratization, there could be no develop­ment of a home market. This latter failure opened wide what was a crev­ice at independence. With every downturn in the international econ­omy, the crevice turned into an opportunity for an externally defined structural adjustment that combined a narrowly defined program of privatization with a broadly defined program of globalization. The result was both an internal privatization that recalled the racial imbalance that was civil society in the colonial period and an externally managed capital inflow that towed alongside a phalanx of expatriates-according to UN estimates, more now than in the colonial period!

But if the limits of the postindependence period were reflected in a deracialization without democratization, I will argue that the Achilles' heel of the contemporary "second independence movement" lies in its political failure to grasp the specificity of the mode of rule that needs to be democratized. Theoretically, this is reflected in an infatuation with the notion of civil society, a preoccupation that conceals the actual form of power through which rural populations are ruled. Without a reform of the local state, as I will soon show, democratization will remain not only superficial but also explosive.


The mainstream nationalists who inherited the central state at indepen­dence understood colonial oppression as first and foremost an exclusion from civil society, and more generally as alien rule. They aimed to redress these wrongs through deracialization internally and anti-imperial­ism externally. The new state power sought to indigenize civil society institutions and to restructure relations between the independent state and the international economy and polity.

In the absence of the detribalization of rural power, however, deracialization could not be joined to democratization. In an urban-centered reform, the rural contaminated the urban. The tribal logic of Native Authorities easily overwhelmed the democratic logic of civil so­ciety. An electoral reform that does not affect the appointment of the Native Authority and its chiefs - which leaves rural areas out of consid­eration as so many protectorates - is precisely about the reemergence of a decentralized despotism! In such a context, electoral politics turned out to be about more than just who represents citizens in civil society, because victors in that contest would also have a right to rule over sub­jects through Native Authorities, for the winner would appoint chiefs, the Native Authority, everywhere. More than the rule of law, the issue in a civil society-centered contest comes to be who will be master of all tribes. As a Kenyan political scientist once remarked to me, the ethnicity of the president is the surest clue to the ethnic tinge of the government of the day. This is why civil society politics where the rural is governed through customary authority is necessarily patrimonial: urban politicians harness rural constituencies through patron-client relations. Where des­potism is presumed, clientelism is the only noncoercive way of linking the rural and the urban.

Confined to civil society, democratization is both superficial and ex­plosive: superficial because it is interpreted in a narrowly formal way that does not address the specificity of customary power-democratization equals free and fair multiparty elections-and explosive because, with the local state intact as the locus of a decentralized despotism, the stakes in any multiparty election are high. The winner would not only repre­sent citizens in civil society, but also dominate over subjects through the appointment of chiefs in the Native Authority. The winner in such an election is simultaneously the representative power in civil society and the despotic power over Native Authorities.

Tribalism is more one-sidedly corrosive in an urban context than in the rural one. Stripped of the rural context, where it is also a civil war, tribalism in urban areas has no democratic impetus. It becomes inter-ethnic only. This practice is not confined to propertied strata. We have seen that migrants who became involved in the inter-ethnic politics of civil society did so partly to protect customary rural rights. In the ab­sence of the democratization of Native Authorities and the custom they enforced, the more civil society was deracialized, the more it came to be tribalized. Urban tribalism appeared as a postindependence problem in states that reproduced customary forms of power precisely because de­racialization was a postindependence achievement of these states.


The accent of mainstream nationalism was on deracializing civil society, but it is the radical regimes that sought to detribalize Native Authority. The institutional basis of that effort was the single party, the inheritor of militant anticolonial nationalism, which symbolized a successful linkup between urban militants and rural insurrectionary movements against Native Authorities. Militant urban nationalism was the social and ideological glue that cemented otherwise heterogeneous peasant-based struggles. From that experience arose the single party as yet another noncoercive link between the rural and the urban.

The single party was simultaneously a way to contain social and polit­ical fragmentation reinforced by ethnically organized Native Authorities and a solution imposed from above in lieu of democratization from below, for the militants of the single party came to distrust democracy, by which they understood a civil society-centered electoral reform. A democratic link between the urban and the rural was in their eyes synon­ymous with a civil society-based clientelism. Seen as the outcome of an urban multiparty project, clientelism appeared as the other side of a deepening fragmentation along ethnic lines.

Whereas multiparty regimes tended toward a superficial and explosive democratization of civil society, their single-party counterparts tended to depoliticize civil society. The more they succeeded, the more the sin­gle party came to be bureaucratized. As the center of gravity in the party-state shifted from the party to the state, the method of work came to rely more on coercion than on persuasion. Whether heralding devel­opment or waging revolution, the single party came to enforce it from above on a reluctant peasantry. Although depoliticization contained interethnic tensions within civil society - and as a consequence within the whole polity - the result of a forced developmental march was to exacerbate tensions between the rural and the urban. The single party turned from a mobilizing organ into a coercive apparatus; in the words of Fanon, militants of yesterday turned into informers of today. True, there was a significant break with the formal institutions of indirect rule, but there was no such break with the form of its power. An institution such as chiefship may be abolished, only to be replaced by another with similar powers. The ideological text may change from the customary to the revolutionary - and so may political practice - but, in spite of real differences, there remains a continuity in administrative power and tech­nique: radical experiences have not only reproduced, but also reinforced fused power, administrative justice and extra-economic coercion, all in the name of development.

The reform of decentralized despotism turned out to be a centralized despotism. So we come to the seesaw of African politics that character­izes its present impasse. On one hand, decentralized despotism exacer­bates ethnic divisions, and so the solution appears as a centralization. On the other hand, centralized despotism exacerbates the urban-rural division, and the solution appears as a decentralization. But as variants both continue to revolve around a shared axis-despotism.


The two tensions the specific form of the African state generated, the interethnic and the urban-rural, have also been faced by oppositional movements. In chapter 6 we surveyed some rural movements, and in chapter 7 the urban movement in post-1973 South Africa.

These movements bring out the two dimensions of the ethnic ques­tion, the internal and the external. Because the context in which rural movements organize is more often than not multiethnic, they face a common question: is a resistance against a tribal Native Authority possible which does not at the same time exacerbate interethnic tensions within the resulting movement?

I have argued that a peasant movement in the parameters of a Native Authority is at the same time an ethnic civil war. Yet this should not be understood as a claim that every such movement is committed to root­ing out the institution of chiefship. Of the South African movements we surveyed, only the Mpondo sought to implement a democratic program whose target was the institution of chiefship. The rest fell short of it, some more than others. But all of them, without exception, sought to redefine the notion of the customary, to limit the powers of chiefship within traditional constraints. If "customary" is the name that power gives to the untraditional force with which it arms Native Authorities, it is also the language of peasant movements that seek to reform the same Native Authority in the name of a custom anchored in notions more historical and popular. When I speak of tribalism as civil war, my notion of civil war is a continuum along which muted tensions coexist long be­fore they break out into open confrontation.

Notwithstanding the colonial claim that traditional Africa was a tribal checkerboard, with each tribe in its own place, we have seen that tribal culture was highly textured and elastic, with the stranger often present on rural ground. For no reason other than to expand their following, the tendency of chiefs was to encourage strangers to settle in their do­main. With a state-enforced and tribally circumscribed notion of cus­tom, two related changes occurred. First, the tendency was to homoge­nize and flatten cultural diversity within the tribe in favor of an official tribal version. Second, the imposition of a tribal law as customary, to be defined and dispensed by a tribal authority, necessarily turned the simple fact of ethnic heterogeneity into a source of tension.

If the rural movements I surveyed were all to some degree marked by ethnic civil war, only the experience of the NRA in Uganda brought out clearly the intertribal tensions that surface in peasant movements. As it sought to weld together a common oppositional movement in a context where settled and migrant populations rubbed shoulders, the NRA de­fined rights as an attribute not of citizenship but of labor. To say that rights belonged to all those resident in a locality, regardless of geographical or ethnic origin, was to say that all those who labor have a justifiable claim to rights. Once labor was understood as the life-sustaining activity of laboring humanity, and not as wage labor, the accent shifted from citizens' rights to human rights.

To view rights as an attribute of labor is also to transcend the opposi­tion between customary law and civil law, for the opposition notwith­standing, customary law and civil law share a common premise. Both see rights as an attribute of individuals belonging to a common land-based community. The difference lies in the definition of the community. From the point of view of customary law, that community is defined in ethnic terms, as the tribe; from that of civil law, the community is a na­tion, whether defined ethnically or territorially. Both subject and citizen derive their rights, customary or civil, through membership in a patri: a tribe for the subject, a nation for the citizen.

Although the NRA was able to bring a creative insight into resolving the interethnic tension in its rural base, the Luwero Triangle, it found great difficulty in addressing the urban-rural tension from a position of power in the city. It saw urban civil society demands for a representa­tional multiparty democracy - a demand reinforced by many Western donors - as a threat, both to its hold on power and to the unity of power holding the country together. It was faced with the old dilemma that had plagued single-party regimes. Would not a multiparty contest in the city be about not just who would represent citizens in the city, but also who would be the master of tribes in the countryside? Would not such a contest both exacerbate clientelism in civil society and extend it to the countryside, thereby also activating and reorganizing democratic politics around interethnic tensions?

If the failure of the NRA - at the time of this writing, in late 1994 - was in making a transition from the rural to the urban, from democratiz­ing Native Authority to democratizing civil society, that of the urban movements in post-1973 South Africa was the opposite. The indepen­dent unions successfully fought attempts by the apartheid state to drive a wedge between migrant and resident labor in the townships but re­mained prisoners of a civil society-centered perspective. Although they disagreed on many issues of principle, the.workerists and populists inside the independent unions agreed on one thing: the community meant the township. If the IFP - and in retrospect, also the ANC - succeeded in bridging the rural and the urban in the context of a multiparty contest, it did so at the cost of exacerbating interethnic tensions in civil society. Whereas the IFP was an urban extension of a Native Authority-based organization that it sought to conserve, the ANC was unable to arrive at a program to democratize Native Authority; instead, it turned to em­bracing those in the Native Authority who were willing to join it in an electoral alliance.

The Ugandan case shows that the democratization of the rural and the local cannot be stabilized unless extended to embrace the urban and the central, and the South African case illustrates the other side of the same proposition: without a democratization of rural customary power, urban civil power must inevitably degenerate. So long as rural power is organized as a fused authority that denies rights in the name of enforc­ing custom, civil society will remain an urban phenomenon. Surrounded by tribally organized customary powers, urban civil society is subject to a dual pressure: deracialization from within and retribalization from without. We can see this reflected in the dilemma the ANC faced in the 1994 elections.

Without a presence in either the reserve or the hostel - and without a program for democratizing customary rule in either - the ANC could reach the rural only from above, through Native Authorities. Confined to waging a democratic struggle in the urban and reduced to reaching rural communities through its customary authorities, the ANC found itself trapped in a Catch-22 situation. Its only option of linking the urban to the rural was through a tribal logic: either an intertribal alliance or an intertribal conflict, or more likely both, an alliance with those who are friendly (such as the chiefs of the Congress of Traditional Leaders of South Africa, CONTRALESA) and a conflict with those who are not (such as Inkatha). In either case, the structure of customary power would remain intact. This is the context in which the 1994 election as­sumed a significance both civic and ethnic.

Critical to the shape these elections came to assume were migrant workers, the social force that more than any other straddled the rural and the urban. Without a democratic reach to either hostels or reserves, the ANC remained alienated from hostel workers, the very reason In­katha raced to embrace those Zulu as the custodian of their customary rights in KwaZulu, a custom it promised to defend with arms if neces­sary. The more successfully Inkatha executed this project, the more the tribal logic of customary authorities came to contaminate urban civil society. As the hostel-township-shanty triangle was engulfed by violence in the early 1990s, this fact registered on popular consciousness with the impact of an explosion. Unable to isolate Inkatha from its social base, the ANC explored a tactical alternative to a fast-expanding conflict. To defuse an intertribal collision, it settled for an intertribal alliance from above. The promise of that alliance was a federated civic power. Its price was unreformed customary power in the reserves. The compromised federal solution of South Africa closely resembled the one arrived at in Nigeria after the civil war.


There is also a methodological significance to the argument advanced in this book. It is that issues of democracy and governance cannot be di­rectly deduced from the analysis of a mode of production; nor can they be read off as prescriptions from a general theory of democracy. In grap­pling with the question of democracy and governance, I have both shifted perspective from the mode of livelihood to the mode of rule and argued that there is a historical specificity to the mode of rule on the African continent. This shift underlines a critique - more in the nature of a sublating than a simple negation - of two kinds of contemporary discourses, that of political economy and that of civil society.

The critique of the standpoint of political economy is clearest in my analysis of the South African experience. In South African studies, the finest fruit of the political economy perspective was the cheap labor power thesis: it argued that apartheid was functional to capitalism, criti­cal to ensuring a regular supply of cheap labor power. I have shifted attention from the cheapness of labor power to its semicoercive and con­trolled nature, in the context of a broader shift, from a focus on the labor question to one on the native question. My locus of analysis has been less the mode of accumulation than the mode of domination.

More than a response to the question of securing cheap labor power in a semi-industrial setting, I have argued that apartheid needs to be understood as the outcome of an unending quest for order in a setting both semi-industrial and colonial. Without denying the importance of the semi-industrial context, I have illuminated the significance of the colonial context in understanding apartheid as a form of the state. Rather than debunk or discard the perspective of political economy, my purpose has been to build on its insights while questioning its holistic claims.

Whereas in South Africa political economists generated a rich debate on the role of the state in reproducing a regime of semicoercive labor, the same cannot be said of the political analysis offered by their counter­parts to the north of the Limpopo. Their claim was that the problem of Africa is one of "backwardness," of precapitalist relations of production, of insufficient proletarianization; in sum a lack of "development."2 The key to democracy, then, is development. Ironically, this crude reduc­tionism still finds defenders on both the Left and the Right, from mili­tants advocating a single-party solution to champions of an IMF-style structural adjustment.

No less convincing, however, is the multiparty discourse of the so-called prodemocracy movements in equatorial Africa. Unlike their sin­gle-party counterparts, theirs is an explicitly political discourse. But it has turned a concrete historical experience - of civil society in the West - into the basis of a general and prescriptive theory. It has thereby turned democracy into a turnkey institutional import. Arguing that the problem of Africa is the absence or weakness of civil society institutions, it speaks the language of exclusion and marginalization, unable to un­ravel the form of power through which large numbers of Africans - in many cases the majority - are ruled.

Both perspectives presume rural areas to be residual, signifying a lack and an absence. That absence may be defined as economic or as political; the rural may be seen as lacking in urban modes of livelihood or in insti­tutions of civil society. Whether it is activists in the trade unions and civics of South Africa or their more liberal counterparts in prodemocracy movements to the north, both have failed to arrive at a political program that addresses the mode of power containing rural populations on the continent. In contrast, my emphasis has been more on the mode of in­corporation than that of marginalization. It is an emphasis less on the regime of rights from which the colonized were excluded on grounds of race than on the regime of custom into which they were incorporated and through which they were ruled.

In an analysis concerned not just with the colonial legacy, but also with postcolonial attempts to reform it, the shift has not simply been from the labor question to the native question; it has also involved placing­ the native question in the context of a broader problematic, the sub­ject question. In practice, this latter shift took place with independence, with the birth of a decolonized and deracialized state. With it, the dual­ity native-nonnative gave way to another, subject-citizen. Inasmuch as reform in postcolonial Africa crystallized along two distinct paths, known as the conservative and the radical, one can speak of two subject prototypes. In the conservative states, which reproduced Native Author­ities as the locus of a decentralized despotism, the prototype subject was stamped with an ethnic identity. In the radical states, which detribalized Native Authorities but where reform degenerated into a centralized des­potism-most dramatically illustrated when the central state branded poor and unemployed urban residents as vagrants and forcibly repatri­ated them to their "home areas" in the countryside - the prototype sub­ject was simply a poor inhabitant in the rural areas, a peasant.

My point, then, is not only that the mode of rule is not deducible from the mode of livelihood. It is also that the specificity of the political in the African experience lies not as much in the structural defects of a historically organized civil society as in the crystallization of a different form of power. This is why the point of democratization cannot be just a simple reform of civil society. It also has to be a dismantling of the mode of rule organized on the basis of fused power, administrative jus­tice, and extra-economic coercion, all legitimized as the customary.

The antidote to a mode of rule that accentuates difference, ethnic in this case, cannot be to denv difference but to historicize it. Faced with a power that fragments an oppressed majority into so many self-enclosed culturally defined minorities, the burden of resistance must be both to recognize and to transcend the points of difference. If there is a lesson in the experience of oppositional movements-whether rural, such as the Ruwenzururu, the NRA, and the Sungusungu or the urban, such as the independent trade unions in post-1973 South Africa - it is that to create a democratic solidarity requires joining the emphasis on auton­omv with the one on alliance, that on participatory self-rule with one on representational politics. In the specific circumstances of contemporary Africa, to create a democratic majority is to transcend two divisions that power spontaneously imposes on resistance: the rural-urban and the interethnic.


The point of this book is that any effective opposition in practice, and any theoretical analysis that would lead to one, must link the rural and the urban in ways that have not yet been done. This is why Uganda and South Africa are the paradigm cases today. Uganda, though the home of the most serious attempt yet to democratize Native Authority, has been unable to address the democratic demands of civil society movements. In South Africa, though the home of the strongest and the most imaginative civil society-based resistance on the continent, reform has floundered on the walls of customary power. As paradigm cases, both allow one to see in one place phenomena that appear as fragmen­tary elsewhere.

What social forces can link the urban and the rural? The only success­ful attempt yet to bridge the two has been the militant nationalist move­ment that followed the Second World War. The political impetus of this movement came from the disenfranchised native strata of the towns. Whether the "verandah bovs" of Nkrumah's CPP, migrant workers in many powerful trade union movements, or the "boatmen" of Cabral's African Party for the Independence of Guinea and the Cape Verde Islands (PAIGC) a decade later, they shared a common social position: they lay beyond the reach of customary law and yet had few entitlements to civil rights. Though in civil society, they were not of civil society.

Faced with a growing and militant nationalism, colonialism embarked on its most ambitious reform program yet. Part of that postwar reform was a stabilization of migrant labor. Colonial governments raised "bach­elor" wages to "family" ones, technically upgraded and differentiated the work process, and extended official recognition to trade unions. A similar process unfolded in South Africa in the wake of the post-Soweto Wiehahn and Riekert Commissions, but the reforms were of limited significance. In South Africa the end of influx control brought a wave of migrants from rural areas to urban ones - mainly into hostels and shantytowns. To the north of the Limpopo, the "informal sector" bur­geoned as a combination of economic crises and structural adjustment led to a shrinking domestic industry alongside deregulated markets. Today it is migrant labor - and those in the informal sector - that forms a class that is in civil society but is not of it.

The social role of migrants varies, depending on the political choices available to them. As the South African case demonstrates, that role can be progressive or nonprogressive. The point about the prodemocracy movement of today is precisely that it lacks a program for linking the urban and the rural on the basis of democratizing rural power, as the ANC in South Africa. In the absence of such a democratization, the cus­tomary will remain a rallying cry lining up urban-based migrants behind customary authorities in their ethnic homes and behind city-based champions of the customary - so as to defend customary rights, however residual these may be. In the linkage between the urban and the rural, the rural is the key. So long as the rural is not reformed, the perversion of civil society is inevitable. This is why the limits of the current South African reform are so serious.

The most serious attempt yet to reform the rural was, as I have already noted, that of Museveni in Uganda, following earlier and more partial attempts, Qaddafi in Libya, Sankara in Burkina Faso, and the early Rawlings in Ghana. They all highlight one lesson: decentralized democ­racy confined to the local state is both partial and unstable. It harbors contradictory possibilities: the point of reform of rural power can just as easily be to link up with representative demands from urban civil society as it can be to check these. If the objective is an overall democratization, it requires a balance between decentralization and centralization, partic­ipation and representation, autonomy and alliance. But if it is to check­mate civil society, a one-sided glorification of decentralization, auton­omy and participation will suffice because, in the final analysis, it is bound to exacerbate the breach between the urban and the rural. Yet it is precisely such a tendency that is a growing orientation in left-oriented intellectual thought, on the one hand opposing and upholding partici­pation against representation, on the other championing autonomy against alliance. If the experience of oppositional movements and the record of regime strategies reviewed in this book are anything to go by, this tendency needs to be seen as a negative development.

Colonial legal theory justified the subordination of subjects to a fused power as the continuation of a customary law and gave it the name of indirect rule; in contrast, it termed as direct rule the racially defined ex­clusion of colonized persons from citizen rights guaranteed by civil law in a differentiated form of power that framed civil society. Postindepen­dence governments seeking to overcome this duality took one of two alternatives; either preserving the customary in the name of defending tradition against alien encroachment or abolishing it in the name of overcoming backwardness and embracing a triumphant modernism. But if indirect rule characteristic of Native Authorities was anchored in par­ticipatory forms, however distorted, and direct rule over civil society in representational forms, however exclusive, then was not the point to transcend both through a creative synthesis?

The reform of indirect rule systems in postindependence Africa built on the practice of participation without representation. In the second phase of radical African governments - from Qaddafi and Sankara to early Rawlings and Museveni - this reform became the basis of dismantling authority in the local state without democratizing power in the central state. Each of the peasant movements considered in chapter 6 tended toward participatory reforms, but none was able to stabilize these on the basis of participation alone. Participatory forms ("empow­erment") that stress the autonomy of a bounded group - only to undermine any possibility of an alliance-building majority-based representa­tion - can justify and uphold the most undemocratic forms of central power. One only needs to look at the experience of self-initiated squatter settlements in South Africa: many began with an emphasis on partic­ipation and ended up with a warlord.

At the other extreme, there is the phenomenon of representation with­out participation. This is characteristic of a multiparty electoral reform whose target is the central state while leaving intact the decentralized despotism crystallized in the local state. Without an accent on participa­tory forms (and, as we will soon see, autonomy), the tendency is for representation to turn into its opposite: instead of a representation of popular strata in the state, the representative turns into an agent of the state power to popular sectors.

If democratic politics calls for joining participation at the local level with direct representation at higher levels, a similar perspective also needs to be forged when it comes to the relationship between autonomy and alliance. If the rationale for autonomy is the legitimacy and particu­larity of the local, then the fragmentation produced by a one-sidedly localized perspective underlines the need for alliance as a way to tran­scend it. And if participation and representation, autonomy and alliance, cannot be viewed in a one-sided opposition, neither can the customary and the civil. In spite of the practice of Native Authorities to justify their writ as custom, the customary was never a single, noncontradictory whole. Not only the Native Authority but also many peasant movements spoke the language of the customary. For every notion of the customary defined and enforced by the state, one could find a counternotion with a subaltern currency. A democratic appreciation of the customary must reject embracing an uncompromising modernism or traditionalism. As a start, it needs to disentangle authoritarian from emancipatory possibil­ities in both.

The point is neither to set aside dualisms that mark social theory nor to exchange one set for another more adequate to describing the con­temporary situation. Rather it is to problematize both sides of every dualism by historicizing it, thereby underlining the institutional and po­litical condition for its reproduction and for its transformation. Al­though theory cannot by itself transform reality, without a theoretical illumination reality must appear a closed riddle.

The fall of Soviet-type regimes in the late eighties was followed by an uncompromising critique of single-party regimes. In the African con­text, these have been followed by equally single-minded and prescriptive reforms embracing multiparty elections on the one hand and decentral­ization on the other. With every fresh round of lessons, however, we seem to lose historical depth. Once again, the impact of multiparty elec­tions - in the absence of a reform of rural power - turns out to be not just shallow and short-lived, but also explosive. Too many presume that despotic power on this continent was always or even mainly a centralized affair, in the process forgetting the decentralized despotism that was the colonial state, and that is one variant of the African state today. In the absence of alliance-building mechanisms, all decentralized systems of rule fragment the ruled and stabilize their rulers. No doubt this is the great attraction of the current wave of decentralization - and the histor­ical amnesia accompanying it - to Africa's current rulers.

But for the opposition that must take stock of social fragmentation as its historical starting point, it makes more sense to appropriate critically the experience of militant nationalism of yesteryears than just to debunk it. The strength of that experience lay in its ability to link the urban and the rural - politically. Its Achilles' heel was the failure to ground the link in an ongoing process of democratic reform, one with a focus on re­forming the bifurcated state inherited from colonialism. Once in power, militant nationalists pursued reform in both civil society and Native Au­thority; deracializing the former and detribalizing the latter. But they reformed each sphere separately, and they did so from above. As reform from above substituted administration for politics, a bifurcated reform strategy re-created the bifurcated state. That failure corrupted a hitherto political link between the rural and the urban into a coercive one, cut­ting the ground from under their own feet. So the attempt to reform decentralized despotism degenerated into a centralized despotism, the other and more unstable variant of the African state.

The second round of reformers, those of the 1980s and 1990s, learned one lesson from their predecessors but also reproduced a limita­tion. They began the reform process from below, by dismantling Native Authorities and reorganizing village communities on the basis of self-administration. Detribalization was thereby joined to democratization, but in the local sphere only. Their dilemma is how to reform the center and thereby how to join the rural and the urban through a single-over­arching but differentiated-reform process.

The record of state reform has been mixed. The tendency of African governments has been to play reform in one sphere against repression in the other. The result, inevitably, is a truncated reform. Of the two ten­sions aggravated by the form of the African state, the interethnic and the rural-urban, the latter is key. Hitherto, there have been two ways of linking the rural and the urban: the administrative and the political. The administrative link has turned out to be coercive. The political link has taken the form of a noncoercive clientelism. To bridge the rural and the urban through a politics both noncoercive and democratic, it is neces­sary to transcend the dualism of power around which the bifurcated state is organized. To do so requires that the nature of power in both spheres, the rural and the urban, be transformed, simultaneously. Only then will the distinction rural-urban - and interethnic - be more fluid than rigid, more an outcome of social processes than a state-enforced artifact.


1. Notwithstanding Basil Davidson’s claim to the contrary in The Black Man’s Burden: Africa and the Curse of the Nation-State.

2. Expressed in nonracial terminology, the problem is “the politics of the belly” to cite the title of Jean-Francois Bayart’s latest book. It is a politics that he explains as “the rush for spoils in which all actors – rich and poor – participate in the world of networks.” My point about clientelism is that it is more an effect of the form of power than an explanation of it