Figuring Democracy

An Anthropological Take on African Political Modernities

Publication Date: 06/07/1997

Journal: Journal of Anthropological Research

Reissue Date:

Page start: 123

Page End: 146

Volume: 2

Edition: 53

12.01 a.m., 25 April 1994. Wale Street, Cape Town, South Africa: The last strains of the anthem of the ancien regime – part requiem, part death- rattle – drift off into the night. A local choir, carefully rehearsed for the occasion, begins to belt out the new national song, with its familiar, once- banned libretto of liberation, its hymnnotic harmony of hope. The old flag, long an emblem of colonialism and apartheid, is folded away for the last time. Its replacement, a brash, multicolored icon of consensus, is raised. The symbolism, by intent, is too obvious to miss. Calico curtains ring down and up as the world’s latest Midnight’s Child, the “new” South Africa, is born.


Perhaps it was sacrilegious, at that precise moment, that moment of unreserved optimism, to recall Malcolm Bradbury’s Doctor Criminale. In this novel about a postmodern philosopher, a fictional fusion of Foucault and Derrida, the Hungarian ex- wife of the hero, one Gertla Riviero, reflects upon the recent transition to democracy and free market economics throughout much of the contemporary world (1992:276):

“Democracy, the free market,” she muses, “do you really think they can save us?…Marxism [was] a great idea, democracy [is] just a small idea. It promises hope, and it gives you [Kentucky] Fried Chicken.”

Ms. Riviero’s commentary is sad, cynical, salutary. Especially so when read in the cooling afterglow of post-election South Africa. Especially so as we call to mind the queues that waited for hours outside polling stations in those last heady days of April, 1994, some in almost sacral silence, some in carnivalesque revelry. Those snaking, eternal queues reminded us of the interminable lines that graced McDonald’s in Moscow a few years back as people voted with their feet not merely for hamburgers or cheeseburgers, but for a market economy and capitalist consumerism. The association may seem irreverent. Yet Gertla Riviero’s question carries an obvious, ominous punch, precisely because it calls into doubt our taken-for-granted narrative of democratization, a heroic liberal myth which links the conventional practices of modernist politics to the prospect of material and social salvation. So, too, if in a different way, does the image of patient, passive people standing in millennial lines to choose either cheap food or political candidates; all the more so as we recall Bayart’s (1993) discomforting aphorism for African public life, to wit, “the Politics of the Belly.”

Let us pursue this question, and follow these lines, for a while. They lead us into an unexpected encounter with very different philosophies of governmentality, democracy, and modernity.


It became commonplace during the 1990’s, especially in Europe and North America, to ascribe the fin de siècle push for democracy in many parts of the world to the end of the cold war and the triumph of the free market over communism. In fact, as many have pointed out, this view was flawed from the first. Apart from all else, the push began well before 1989. But no matter: the association is itself a symptom, often misrecognized, of something much longer in the making, namely, a fundamental reconstruction of the modernist world-order. We have ourselves suggested before that the events of 1989 were evidence of an unfolding Age of Revolution, an epochal process akin to the one that began in 1789 – the European Age of Revolution, that is, which gave us modernity, the seeds of the nation-state form, industrial capitalism, the second colonialism, and much besides (J.L. Comaroff 1995). The present revolution has been marked, in particular, by the rise of a planetary political economy in which sites of production and consumption are widely dispersed; in which social class is rendered barely visible by being scattered promiscuously across the earth; in which finance takes precedence over fabrication, flexibility over fixity, the short-run over the long; in which the state outsources many of its received operations, not least those involving the exercise of violence; in which the nation is confronted by the irreducible fact of increasing demographic heterogeneity; in which governance is represented primarily in the argot of technical oversight; in which politics, more a matter of ID-ology than ideology (see Chapter 3), is increasingly focused on the simultaneous calculi of right, interest, and entitlement, often pursued by judicial means.

For many, these things are cause for despondency. Let us return to Dr. Criminale, Bradbury’s figurative philosopher. Ours, he says (1992:330), is the media age, the age of simulation…The age of no ideology, only hyperreality…Too little reality, also too much. Everywhere, wild fantasies, everyone wants a violent illusion. Life is a movie, death a plot ending, no stories are real. And even the philosophers think in unrealities, [as] they describe a world of no ethics, no humanism, no self.

In this new Age of Revolution, fear of the atomic bomb subsides. But anomic bombs explode all over the place. People across the globe – alienated, disempowered, dispossessed – commit extraordinary acts of violence in the name of ethnic and national aspiration. The “me” generation folds into the “we” generation. And the end of politics, at least politics as anything more than the pursuit of brute interest, appears visible on the horizon.

The scenario, like Doctor Criminale himself, might be fantastic. It is, however, becoming ever less fictional, ever more recognizable.

But how is this darkly pessimistic view of the contemporary world to be reconciled with the rise of late 20th century democratic movements in so many far-flung places? Were those movements not a positive, liberatory sign of the times in that premillennial moment, that Great Time of Signs? And how, in particular, ought we to understand them in Africa, long seen in the West as the continent-least-likely-to- democratize-itself?

It is difficult to gainsay those who draw connections between the recent rise of democracy and the triumph of consumer capitalism – even if the line of causality that joins them is at once complicated and the subject of ongoing debate. Capitalism, to be sure, does not require democracy; it has done perfectly well under authoritarian regimes in the past, and continues to do so in many parts of the late modern world. But those nation-states that seek to democratize themselves appear, these days, to require at least the figment of a free market. An elective (or is it electoral?) affinity connects the ballot box to business. Nor is it a passive affinity (cf. Young 1993:299f). U.S. overseas aid has become largely conditional on the establishment of “democratic institutions.” For which read “regular elections.” To wit, in 1996, Robert Mugabe – then still a leader of some standing, now a discredited dictator who takes every opportunity to censure the West – drew a direct connection between ballot box, business and foreign involvement in African politics: “Western countries,” he said, push “multi-party [systems] for Africa because it enables them to “buy influence” and “manipulate parties” into creating congenial economic environments.2

The contemporary Western concern with the democratization of the global south, however, is not reducible to utility alone, important though that may be. It has roots in the hegemonic, indeed ontological, association throughout the global north of freedom and self-expression with choice. Democracy has become to homo politicus what shopping has long been to homo economicus: a sacred, cosmic fusion of free will and righteous human satisfaction. They are, so to speak, two sides of the same coin, two regimes of consumption underpinned by the same mode of ideological and material and production.

On 1 May, at 11.48 p.m., during the counting of votes after the first free election in South Africa, SATV Channel 2 broke into its local news coverage to broadcast a meta-advertisement, an advertisement for advertising. “ADVERTISING,” blared the message on the primal screen, “THE RIGHT TO CHOOSE.”

It is no coincidence, then, as several theorists have observed, that democracy has increasingly been reduced, in practice, from the substantive to the procedural (e.g. Farer 1989; Barsh 1992); that, purged of any ideological density, it has come to connote little more than the periodic exercise of preference, the satisfying of desire, the physics of pure interest. To wit, it does not take a political theorist, or the fictional Ms Riviero, to make the point that, understood thus, democracy is a small idea, one that is more likely to bring with it Kentucky Fried Chicken and McDonald’s than an amelioration of the human condition. We might go yet farther: to argue that processes of democratization bespeak a historical paradox, namely, that “the people” are being empowered in the politics of state at the very moment when, as we have noted, the politics that count are moving elsewhere — to global processes and institutions, into the corporate world and non-governmental organizations, the media and the law, new social movements, “grass roots” coalitions, and other domains of civil society.

To phrase all this in the interrogative voice, is it possible that Dr. Criminale is correct: that democratization is a product of the death of politics, of its dispersal to everywhere and anywhere and nowhere in particular? Is democracy rising because it has become politically beside the point?

An echo here from home. Speaking of democracy in a workshop at the University of Chicago many years ago, Wayne Booth – author, tellingly, of The Rhetoric of Fiction (1961) and The Rhetoric of Irony (1974) – observed that freedom of speech is guaranteed in America only to the extent that no one is listening; that, while everybody has a right to talk, nobody has an obligation to pay attention; that democracy disempowers by encouraging a cacophony in which voices cancel each other out.

Put these two things together – the reduction of the Idea of Democracy to the exercise of choice and the decentering, de-institutionalization of politics – and what do we get? For one thing, some of the concerns that many Africans, academics and intellectuals and every-persons alike, raise about the export of modernist European models to the global south: that they are founded on an “extremely narrow” conception of public life, one that places too much emphasis on “votes and free-market economics”3 and too little on the realization of universal human rights, civil liberties, the commonweal, and transparent, accountable government – all of which, according to recent survey research, tend to be embraced in popular African definitions of democracy (cf. Bratton 2002:5), definitions that also take on heavy local inflictions across the continent. Given that the meaning of the term is hardly unambiguous or uncontested in the global north, as Mahmood Mamdani (1986, 1990, 1992) has noted – we paraphrase him heavily here – how much more murky does it become in Africa, whose vast array of dynamic, evanescent cultures have their own theories and practices of politics, of personhood, of power, of representation. As this suggests, the cultural transitivity of the concept cannot simply be presumed, as it so often is by comparative political scientists. The more general implication? That the common presumption in the West according to which Africa ought to adopt the liberal modernist Euro-American model (see e.g. Bratton and Mattes 2001), an ideology floating free of its social and historical moorings, leaves Africans with a unenviable dilemma: to opt for either (i) a highly un-African political order, wherein the body politic is composed of autonomous, individualized, right-bearing citizens whose primary political being is congealed in the exercise of the ballot;4 or (ii) an “indigenous” alternative, usually characterized as anti-modern, ethnically-based, patriarchal, traditionalist, customary, communalist, clientalist, and authoritarian – and/or, more insidiously yet, populist. What kind of choice is this? Even more fundamentally, what, in its own terms, might democracy actually mean in Africa?

Mikael Karlstrom (1996:485) observed, in the mid-1990’s that, notwithstanding the burgeoning literature on democracy in contemporary Africa, surprisingly little heed had been paid to this last question. As long as it is not adequately addressed, he added, we have little hope of grasping postcolonial politics at all, little hope of making sense of such things as, say, the Ugandan insistence that political parties are inimical to representative government. But there is yet another corollary here. Some African counter-discourses on democratization, as we have already intimated, are grounded in a vernacular political anthropology that offers a substantive critique of conventional Western political theory and practice. By confronting this narrative we stand not merely to understand African politics better than we do now – to understand what lies beyond the “politics of the belly,” beneath the “banality of power…in the postcolony.5 We might also arrive at a more reflexive, critical appreciation of our own received political forms.

On the Levi-Straussian principle that one good case may illuminate an entire world, let us offer an exemplary instance to make our argument. Our choice will be surprising perhaps. We do not take a country in which representative government or electoral politics have been repudiated, subverted, or misappropriated. Such examples are either too easy or too stereotypic to be useful. Rather, we take Botswana, the African nation-state most widely regarded as a “model” democracy6 – and the closest, by common agreement, to the Western ideal. This very similarity, at least in appearances, will serve to underscore a brace of revealing differences.


Consider the following facts. In October 1974, Botswana held its third national elections, in which the ruling Botswana Democratic Party (BDP) won an emphatic 85 per cent majority of the seats in the national assembly.7 Both before and after the ballot there was a great deal of public discussion, seemingly spontaneous and unprovoked, of the advantages to be gained from the introduction of a one-party state. Many people, clearly, favored a move away from the existing British-style multiparty system; so much so, in fact, that the president at the time, Sir Seretse Khama, felt compelled to comment repeatedly on the subject, to refuse even to ponder the possibility – and to encourage people both to vote and to consider the merits of all parties. His public statements were featured prominently in the Botswana Daily News at the time.8

In hindsight, and from the vantage of the global north, this expression of popular support for a one-party system would seem odd. First, it did not come primarily from BDP voters. Adversaries of the government were among its more enthusiastic protagonists; to wit, Puo Phaa, official organ of the opposition Botswana National Front – which was led by Bathoen Gaesitsiwe, the ex-ruler of a large chiefdom – urged the formation of a “national government, fusing all political groupings into a single administration.9 Second, it was not engineered by a cadre of bosses or captains in the name of a mass ruling party. At the time, as Colclough and McCarthy (1980:41) note, the BDP was anything but that: “indeed it scarcely exist[ed] between elections.” Third, in refusing insistently to change existing electoral arrangements, much to the delight of South Africa and other Western powers,10 the Khama regime was aware that the BDP was passing up its best chance to gain a legitimate monopoly of the organs of state. Fourth, and most puzzling of all, demands for a one-party system were typically justified on the ground that it would foster both better government and more participatory democracy.

These demands resonated with informal views we encountered in rural Botswana at the time, especially in the south.11 As we shall see, they were consistent with the way in which village populations tended to participate in electoral processes. What is more, they echoed opinions we had heard before. During the previous general election, in 1969, we had been delivered a memorable lesson in comparative political anthropology by a local teacher, an organic intellectual from the edge of the Kalahari. This man, who described himself as “neither a radical nor a traditionalist,” had argued that one-party systems were the “only true social democracies.” With due respect for old European verities, he added politely, the very idea of a multiparty democracy is a contradiction in terms. It abases politics, shrinking them to nothing more than an occasional act of choice. And, by erasing all real government accountability between elections, it licenses the indifference of regimes in power both to popular participation and to public criticism – thereby alienating the citizenry at large from the everyday functioning of the state. President Khama seems to have been aware that views of this ilk had been gaining currency among people in the countryside. Speaking before the 1974 election at Oodi, a small town near the capital, he went to great lengths, in defending multiparty democracy, to stress that “the Government’s intention was not to fetter or discourage… criticism.” That, he said, would be “against our Setswana tradition.”12 Of which more later.

How, then, do we explain such manifestations of antagonism against multiparty democracy, especially where it seems to have taken root so successfully? Why did it appear to these people as an oxymoron, as antithetical to participatory politics, even as an elaborate Western mystification? What accounts for the positive light in which a one- party system came to be regarded here? And to historicize these questions, one or two more: Was this outburst of vox populi merely a passing moment in the history of the public sphere in Botswana, its civil society crying out, ever so briefly, against the postcolonial state? Or did it speak to something more enduring. If so, what? And how?

After all, foreign observers have been quick to comment on the non-involvement – “apathy” or, worse yet, “ignorance” are the words commonly used – of the electorate in matters affecting national politics in the 1970’s and 1980’s.

In order to address these issues, we begin by turning to so-called “traditional” Setswana political theory and practice, a vernacular theory and practice that, albeit contested and constantly transforming itself, persisted through the colonial epoch; then to its conjuncture with the postcolonial politics of the nation-state. For it is here, we believe, that the answers lie. Observe, in all this, that we have three subtexts, three not- so-hidden agendas. One is to show that African political anthropology, despite repeated criticisms of theoretical aridity, has something yet to add to the analysis of world- historical phenomena. The second grows out of an old axiom. Long ago, Fortes and Evans-Pritchard (1940:4) said that Western political philosophy, because of its lack of comparative perspective, has had nothing useful to say to political anthropology. We seek to turn this on its side: to argue that a political philosophy found in another social world may be the basis of a critical anthropology of our own. The third is to provide a corrective to the persisting tendency in the global north to reduce Africa to adjectives – communalist, clientalist, patriarchal, and the like – thus to reproduce tired racist archetypes. And, worse yet, to mistake those archetypes for empirical descriptions in which to ground political theory.


Botswana, until 1966 the Bechuanaland Protectorate, is usually said to be made up of eight distinct chiefdoms (“tribes”). These, prior to incorporation within the British Empire in 1885 and with modifications during the colonial period, were the major, but not the only,13 political communities into which indigenous populations were grouped.14 Although the system of local government has changed over the years, and chiefs have been denuded of much of their authority,15 the eight chiefdoms still exist. Some of the larger ones are today (more-or-less) coterminous with the jurisdictions of districts and their councils. Moreover, while its urban centers have grown enormously, Botswana is stereotypically seen as a predominantly rural nation: much of its citizenry was raised in, and sustains active links with, “villages.”

While chiefdoms varied in size and in the minutiae of their institutional arrangements, the dominant features of their political organization, cultures, and ideology were broadly shared. These have been thoroughly documented;16 although there does remain some controversy about the politics of succession to high office17 – and an unfortunate tendency among Western social scientists to typify the public sphere in Tswana communities, often glossed as “the kgotla system” (see below), in rather too simple terms.18 For present purposes, the briefest of summaries will do.

From the earliest documentary accounts we have of centralized “Bechuana” polities, dating from the first half of the 19th century, three things are clear.19 The first is that the chiefship was seen to be the axis mundi of the social world. It was, as one Tshidi-Rolong elder said to us in 1969, like the pinnegare, the central pole, of houses of old. Everything – the fertility of the earth and the abundance of the rains, security from attack and success in war, the passing of the seasons and “giving of the seed-time,” material wealth and spiritual well-being, the crafting of legislation and courts that judged fairly – all these things, and much besides, turned around the apical office. Its holder, in principle at least, personified his people, signified their sovereignty and subjectivity, embodied their essence. He was known by an honorific whose form was the metonymic singular of the name of his “nation” (morafe): Mokwena, the ruler of the Bakwena (mo-, sing; ba-, plural), Mongwaketse, the ruler of the Bangwaketse, and so on.

But, second, a clear line was drawn between chiefship (bogosi) and chief (kgosi), office and incumbent. The former stood for the very existence of the polity. It was the public sphere incarnate, the morafe made manifest and represented back to itself as a political principle. The authority vested in it – albeit historically shifting over the long-run – was taken, at any moment in time, to be beyond question. The latter, by contrast, was merely human. He might be more or less effective a ruler, more or less influential, more or less adept at mobilizing the political capital available to him. Early European visitors to the Tswana were impressed by the charisma and command of some “kings” who, it is said, struck awe into their followers and whose slightest whim elicited the strictest compliance. But they were also fascinated by the frankly critical way in which most sovereigns were addressed at their own courts. And by the fact that their power was often constrained by the sheer unwillingness of their subjects to do their bidding.20 Elsewhere (e.g. J.L. Comaroff 1975, 1978; see n. 17) we have shown that a chief who lost all legitimacy, who was said publicly to be “not fit to rule” (Campbell 1822,2:157), was likely to find his genealogical status successfully contested by a rival. This in spite of the prevailing rules of ascription according to which sovereigns held office by virtue of birth, not election. It was always possible to unfix the fixities, to unscrew the inscrutabilities, of ascribed rank by reconstruing the relations that gave rise to it.

The third thing of note is that great store was placed here on what might be glossed as “good government.” Substantively speaking, chiefs were responsible for all aspects – political, judicial, administrative, material, spiritual – of collective well-being; that is, for everything in the public domain. This, furthermore, is to be understood in historical terms: sovereign responsibility embraced the fluid realities of time, space, and situation. Where transformed conditions demanded, say, that the colonial state be dealt with in a particular fashion, or that dams and storage depots be built for purposes of agrarian “development,” rulers were held to account for the discharge of these functions. But, and this is the crucial point, the ideology of good government paid less attention to the content of public affairs than to the means by which they were managed.

Tswana ideas about the proper means of governance were elaborate, nuanced, and enduring; we heard any number of discourses on the topic in the 1970’s. Above all, they stressed (i) the participatory, consultative aspect of the public sphere, in which there was, ostensibly, “perfect freedom of debate” (Philip 1828,1:133), and in which all male citizens (more recently, all adults)21 were entitled to a voice – just as they had the right to be represented by headmen on chiefly councils; (ii) the proportional relationship between the performance of any ruler (assessed against the cannons of good government) and his legitimacy (as indexed in his recognized capacity to wield control over people, policy, and public life); and (iii) the fusion of what, in Western social science, is nowadays distinguished as civil society and the state.

In sum, chiefs were expected to rule “with” the people. Kgosi ke kgosi ka morafe went the most quoted adage in the Tswana political lexicon; “a chief is chief by the nation.” What this meant, in practice, is that sovereigns were expected to surround themselves with advisors to guide the everyday life of the polity, men for whose advice and actions they were held responsible; to hold regular meetings of councils of headmen and other chiefly conclaves; to summon public assemblies of various kinds from which emerged policy that reflected popular views and attended to the common weal; to ensure that the hierarchy of courts over which they presided did not favor the rich over the poor, royals over commoners, or men over women (even though the latter, as “jural minors,” had to be represented by male kin); to be open always to approach by their subjects, whose physical welfare they were also obliged to heed, redistributing food and other requisites in times of need.

In Southern Tswana chiefdoms, in fact, past rulers were — in some places they still are – recalled by the legislation they introduced (cf. Schapera 1943) and by the wisdom of those whom they recruited as advisors. They are also remembered by their capacity to bring rain, itself a sure spiritual gauge of political mastery; but that is another story. Ultimately, in this respect, chiefly success was numbered in observable achievements: “improvements,” in the Protestant-saturated language of modernist governance. But delivering improvements, in turn, hinged on the public cooperation that a ruler could command. Which, tautologically, depended on the degree to which he was seen to measure up to the ideals of good government. Note, by way of example, the following text, which we published more than thirty years ago (J.L. Comaroff 1975:145). It comes from a speech made by a local elder statesmen in February, 1970 at the installation of Besele, the new ruler of the southernmost chiefdom of Barolong:

A chief can only be judged by what he does…If you treat [people] with respect, they will treat you with respect. If you shun them, they will shun you. And if you frighten them they will run away… We will be watching to see whether you are going to make improvements. Chiefship is not an easy job. A chief never sleeps. A chief does not discriminate. Batswana say that a chief is chief because of the nation. If we cannot see you in the court [kgotla] we shall draw away from you. And if we do will you still call yourself chief?

In analyzing this text when we first published it, we noted, in particular, how it underscored the significance attributed, in the local political imaginaire, (i) to the Hegelian interdependency between ruler and subject; (ii) to the measurability of chiefly success in terms of practical, palpable accomplishment (“what he does, his industry…[his] improvements”); and (iii) to the possibility that an authoritarian or an inattentive sovereign may be repudiated [“shunned”], even removed (“if we [draw ourselves away from you, will you still [be able to] call yourself chief”), notwithstanding the ideology of ascription in terms of which succession to high office is represented (see above; also n.17).

This, self-evidently, implied the existence of a model of incumbency, a paradigm of political legitimation in terms of which the actions of rulers were evaluated and their authority negotiated; by which, that is, the equation of performance to power was given practical, realized form. At the core of this equation was a simple socio-logarithm: the willingness of political subjects to comply with the commands of a chief was held to depend on the degree to which he could demonstrate, in public, that he had properly discharged the obligations of his office.

It follows – pace received wisdom that goes back to African Political Systems (1940) and persists in some quarters – that the “rights and duties” of Tswana (and, for that matter, other African) sovereigns was never immutable, never fixed by “tradition.” To the contrary. Their authority varied widely. As we have already said, some appeared, alike to their subjects and to outsiders, as mighty kings. Having established their legitimacy, they could exercise almost dictatorial power. Others found it hard to impose their wills, or their executive decisions, at all. Most, however, traversed the line between these extremes during their reigns.22 To be sure, many of the scholarly arguments that surround the analysis of Tswana politics, past and present, flow from an inattention to precisely this capacity for transformation over time and space.

How, then, did the model of incumbency, the equation of performance-to-power, work out in everyday practice? The answer to this question begins with the fact that, whatever their formal agendas, public meetings were also forums in which chiefly regimes were subjected to debate and evaluation. The process was founded on a crucial assumption: that there existed, tacit but nonetheless well understood, an incremental scale of sovereign authority; that, as the legitimacy of a ruler increased, the more inclusive (and exclusive) became his recognized right to regulate the various ways and means, the instruments and institutions, of governance23 – expanding, potentially at least, until it embraced virtually all aspects of social life. Thus, for example, before the passage of the Tribal Land Act (1968), a strong chief enjoyed, among other things, sole control over the distribution of fields, pasturage, and residential plots – either allocating them himself or appointing surrogates to do so – and a monopoly over the creation of new political constituencies (wards, sections, villages, provinces), along with the offices that ruled over them. He also could expect to be obeyed when he summoned labor for communal works and improvement projects, to receive sundry forms of tribute, to minister over the timing of the ritual and agricultural cycles, and to have his legislative initiatives, executive orders, and legal judgments implemented with dispatch.24

Conversely, a ruler who lost his legitimacy, a process that occurred slowly rather than suddenly, found it ever more difficult to exert control as, cumulatively, he forfeited the various rights of office. The exact composition of this scale of rights differed from chiefdom to chiefdom. But it appears to have existed in some form everywhere; again, with contrasting degrees of explicitness. In Barolong, for instance, the first thing a chief would lose was his sway over the activities of voluntary associations, which were likely to listen to him politely and then ignore him utterly; thence he would forego his monopoly over the allocation of land, this usually being effected by public demand that a committee be appointed to “help” him make decisions. Next went the taken-for- granted presumption that judgments and sentences handed down in his court would be executed without question. This was followed by the erosion of other capacities and entitlements: to call people to labor on public works, to enact legislation, to establish new constituencies or regulate space and time, to demand tribute, finally even to summon meetings.

But this leaves one part of the question unanswered. By what rhetorical means and concrete measures was the indigenous equation of performance to power actually resolved? How was the legitimacy of a reigning chief – the substance and scope of his command over the public sphere — actually negotiated? How, in short, did sovereign authority actually come to expand or contract?

Through mahoko, words. Words spoken in kgotla, in the public sphere, which were assumed to have great pragmatic power to affect the world; words spoken in the genre of political oratory, a genre not specifically named in Setswana but one for which Tswana are justly famed. Theirs is a rich aural culture, in which the aesthetics of utterance are potent indeed. And in which the negotiation of chiefly legitimacy takes on a very particular form.

Before saying more about that genre, however, a point of clarification. The kgotla might have been where chiefly authority was negotiated, but the production of that authority, and the power that lay behind it, was an altogether more complex matter. To hold that legitimacy was determined by the unconstrained consent of the governed, that it was decided purely by argument in town meetings, or that rulers bore passive witness to their own evaluation – all of which is implicit in the vernacular model of incumbency – is to simplify reality. Public debate, always the object of careful strategy and management, was a site of struggle, not a neutral enactment of vox populi. The distribution of support to which it gave voice depended, in major measure, on prior power relations, relations forged in offstage dealings of various kinds. The discourse of chiefly evaluation provided a medium by which the invisible calculi of patronage and influence congealed into social “facts,” collectively recognized lines of alliance and antagonism. There is a tautology here, of course: civic discussion was taken both to reflect and to determine sovereign legitimacy. But the tautology is more apparent than real. Verbal exchanges in kgotla made manifest, and so converted into the currency of politics, all the transactions that occurred, dispersed and individuated, across the axes of everyday life.

Tacit in all this is a political dynamic of some moment for the more general question at hand. Inasmuch as discourses of chiefly evaluation expressed alliances and antagonisms, support and opposition for the ruler – inasmuch, that is, as they were a partisan theater of the political, they tended to be articulated around identifiable factions. The existence of (usually a pair of) such factions was endemic in local public life. (The reasons for this are too complex to go into here. They flow from the fissiparous character of Tswana polities of the past, which were often wracked by rivalries over the chiefship. These invariably pitted the reigning sovereign against an agnatic adversary, thus dividing the morafe into two blocs, each around its royal leader.) One of the factions was always composed of “king’s men,” core supporters from among whom the personnel of his regime were drawn; 19th century missionaries, tellingly, sometimes referred to them as “the chief’s party.” The other, which might be more or less articulate(d), bounded, and assertive, depending on circumstance, typically clustered around senior royal patrikin who were, potentially and often in practice, the ruler’s primary adversaries for position and property. Again, all this been well documented. The point, as far as we are concerned, was the taken-for-granted, almost inevitable presence of factional alignments in local politics. For out of these blocs came the primary players, the dramatis personae, of the public sphere — as well, significantly, as the political and dialogical motivation that gave shape to discourses of chiefly evaluation.

The aesthetics of public discourses about governance and chiefly performance – the poetic play, that is, of form and substance – held the key to their politics. The latter derived from the juxtaposition, in “parliamentary” speeches, of two kinds of utterance; two styles, whose difference was closely connected to the vernacular distinction drawn between office and incumbent. One style (elsewhere we have referred to it as a “formal code”; J.L. Comaroff 1975) spoke of the ideals of good government, and of the regnant ideology of chiefship, largely in idiomatic form; phrases like kgosi ke kgosi ka morafe (see above), batho ga se ba melamu, ba bokwa ka lotlhare (“people are not ruled with clubs, they are waved with winnowing fans”), and others that specified expectations of office-holders. These utterances relied heavily on formulaic speech, were rarely phrased in the first person singular, their author usually being the collective “we” (“We Barolong say that…;” “It is our way/custom…” “Our fathers taught that…”). What is more, because they invoked shared values, they presupposed the consensual agreement of speaker and audience.

Strikingly different to these formulaic utterances, the second kind addressed the performance of the chief. Phrased always in the first person singular (“I must speak my mind, Chief!…”; “I have heard what others say. It is my view that …”), statements made in this register were not formulaic at all. Typically frank and forthright, sometimes even brutally censorious, they tended to be syntactically more elaborate, to deploy a wider vocabulary, to rely more on evidentiary argument than on shared assumptions, and to be voiced with a view to their persuasive force. These statements were made in a spirit of political argumentation. In observing such speech acts, we were also struck by the fact that, in contrast to more formal utterances – which, at best, were heard in polite silence – they were typically listened to in rapt attention.

These two styles were deployed in careful counterpoint to one another during the course of most political speeches. For their part, “king’s men” sought to convince the public at large of the convergence between the ideals of good government and the reigning incumbent’s record of actions and accomplishments; this by iterating the first, in formulaic speech, as a point of reference, a template almost, against which to mount first-person polemics, propositional claims and political arguments. Conversely, opposition factions would try to force the greatest plausible divergence between the mantras of good government and the material performance of the office-holder, at least as they construed it in their narratives of failure.

For chiefly protagonists, it follows that the greater the degree of convergence they could establish in the public eye between ideal and performance, the broader the claims they could make for expanding the authority of the ruler. Ultimate success, in theory, was when office and office-holder became as one, when statements in the formulaic mode about the first might be said to apply to the second; in practice, this condition of absolutism was never reached in Tswana polities, there being counter- forces which put constraints on the accumulation of sovereign power beyond a certain point. The inverse is also true. For opposition blocs, final victory occurred when the divergence between the ideal of good government and the performance of an office- holder became so great – and, concomitantly, sovereign authority so truncated – that the ruler was no longer a “real chief.” Whereupon, as we implied earlier, he could well be deposed.

Participation in discourses of chiefly evaluation was not confined to those who identified with one or other faction, although the close supporters and active antagonists of a ruler were likely to be most vocal; also the most caught up in the political tactics and intrigues that often lay behind, and broke through to the surface, in the dramaturgy of public dialogue. The unaligned, however, did not merely add their voices to the debate. They acted, at once, in the manner of a chorus and a jury, echoing or disagreeing with the arguments of those more partisan, commenting on their plausibility and persuasivess, and suggesting implications that might follow for the standing of the chief. From these interventions a measure of consensus was likely to emerge as speakers began slowly to draw closer in their views; this measure serving to confirm, expand, or redelimit the state of sovereign authority for the time being.

In sum, the kgotla was more than a forum for the discussion of social policy, although it certainly was that too. Nor was it just an African analogue of the classical polis (see n.18). It was also (i) a context for ongoing discourse about governance and sovereign authority – and, simultaneously, (ii) a space of contestation in which the powers of a living ruler were negotiated and given social currency. Its primary constituencies were factions rather than political parties, one a chiefly bloc and the other an opposition. These constituencies, patently, did not differentiate themselves according to ideology or matters of principle. Their arguments, recall, were about the means of government, not its content. In striking contrast to Western nation-states, where policy is seen from within to be the provenance of partisan politics, here it was taken to be a product of public discourse.

There is much more to the subtleties and the substance of Tswana political culture, past and present. Also to the workings of its public sphere. Enough has been said, however, to allow us to revisit, and to make sense of, contemporary discourses of democracy and the postcolonial politics of this nation-state.

Two brief, final observations before we do.

One is that there has been a revisionist tendency, in some circles, to portray “the kgotla system” as an altogether more repressive, more authoritarian institution than we and others allow. Good (1992:70; cf. Parson 1984:6f), for instance, says that “the kgotla essentially operated to facilitate social control by the leadership,” the implication being that it had less to do with the politics of public deliberation than with the sheer exercise of power by ruling cadres (cf. van Binsbergen 1995). This might have been true, some of the time, of some of the stronger, more centralized chiefly regimes – such as that of the Ngwato, the largest of all “tribal” polities in Botswana and the one usually treated as paradigmatic. But, as a general statement about the Tswana public sphere, the claim does not bear scrutiny. The documentary record shows that the kgotla was always a site of active political contestation in which, far from merely being exercised, sovereign authority had to be negotiated. And could be forfeited as well as fortified, withdrawn as well as won.

The other point is that, in the passage from the past to the postcolonial, the kgotla has remained a crucial element in the political imaginaire of Botswana. This in spite of its roots in the “village.” Or its “traditionalist” connotations. Since independence, in fact, public forums, called “freedom squares,” have been created all over the country, including in urban contexts. The resonance with an older vernacular public sphere could not be more obvious. Furthermore, as we shall see in a moment, national politicians have found themselves drawn back to the kgotla even in the course, and cause, of distinctly nonparochial political processes. In short, what we speak of here is far from a quaint anachronism, a romantic remnant of days gone by. It describes a cultural context, and a set of discursive practices, that are very much of the continuing present.


Let us return, then, to postcolonial politics and discourses of democracy.

In 1965, some months before Botswana became independent, national elections were held for the first time. Here, as in many other parts of Africa, decolonization – in the formal, political sense of that term – was fairly rapid. Three years earlier, the Botswana Democratic Party (BDP) had been established under the leadership of Seretse Khama and “other bourgeois nationalists” (Good 1992:72) drawn largely from a cattle-owning elite with strong connections to the countryside. From the start, the BDP promised to relegate chiefship and “tribalism” to the peripheries of postcolonial governance. It pledged itself to the evolution of a secular liberal nation-state, European in style; to the growth of a secure capitalist economy based on a mix of agrarian and industrial development, conventionally conceived; to a politics of moderation, the rule of law, and broad principles of social justice.

From the start, too, as Picard (1987) notes, the BDP groomed itself to be a “government party.” Enjoying strong support from the colonial administration, it acquired “a monopoly of the resources and apparatus of the state” (Good 1992:72). Other parties were formed as well, some of them earlier. But they never approached the levels of organization, the material and cultural capital, or the broad-based following of the BDP. The latter was helped by the fact that its members of parliament and district councillors “frequently [had] close kin ties with the traditional aristocracy” (Colclough and McCarthy 1980:41). Although the party set out to marginalize chiefs, and to distance post-independence Botswana from its indigenous political culture/s, there is no question that Seretse’s own popular status at the grassroots was due, in part, to his royal rank –which had been dramatically underscored by imperial intervention. Heir to the Ngwato chiefship, he had famously been forced, by Her Majesty’s Government, to renounce his rights to office as a condition of return from an involuntary exile occasioned by his marriage to a white woman.

That first election, as we said above, yielded an overwhelming victory for the BDP. What was most notable about it, though, was the very high turnout: 74% of all those registered. This was in spite the fact that, in some parts of the country, voter education had been severely limited. Moreover, because distances to polling places were often large and transport was not always available, it was physically difficult for many actually to cast a ballot. Nor was the organization of the election entirely problem- free. All of which made the high rate of participation altogether remarkable. And interesting, too, in light of accusations, voiced in the media and by foreign observers in the 1980’s, that a disturbing proportion of the populace of Botswana evinced indifference to, or ignorance of, the democratic process. It is even more striking in light of what was to happen later.

What, then, did happen later?

Several things, of which four stand out. The first was a radical drop in voter turnout in subsequent elections, down, for example, to 31% in 1974. There is one conspicuous exception, however: 1984, the national ballot after Seretse Khama’s death, when his successor, Quett Masire, had to go to the country as its new president-in- waiting. And go to the country he did. Literally. He went from kgotla to kgotla in an effort to persuade people to vote, to prove his willingness to listen to their demands, and to assure them that he would govern them well (Shepherd 1984:28.) An explanation for these patterns of voter turnout? According to Holm (1987:124), “a segment of the public” thought that, “as has always been the case with a chief, there is no need to reelect the President. Thus they do not go to the polls until a new President is chosen.” He is correct to draw the parallel, although we would take it further. As incumbents of apical offices, chiefs and presidents were subject to similar ideologies of governance (cf. Charlton 1993:331): both were expected to demonstrate their acumen and accomplishments in office; neither could assume their legitimacy; each was held to account for his actions, for the wisdom of his advisors, for the performance of his regime; and each had to subject himself to evaluation – all of which Masire appears to have appreciated. But, as long as they ruled “with the people,” and delivered the fruits of good government, there was no particular need to vote for or against them; under which conditions, ironically, as Colclough and McCarthy (1980:44) conclude, “declining turnout [may] be taken as a mark of approval.” Indeed, Holm (op. cit.) implies, it is only when a new incumbent has to be designated, for reasons of death or deposition, that there is a felt need for an expression of mass public opinion. Then, too, the process runs in close parallel. In each instance, a candidate is identif ied by a ruling cadre (the majority party in the case of the president, powerful royal factions in the chiefdoms), and is presented to the polity for its consideration. Hence the high turnouts in 1965 and 1984. And the indifference on most other occasions.25 In such circumstances, procedural democracy – defined (i) by elections whose primary justification is the abstract passage of time, (ii) by an ethos of choice and change, and (iii) by mass public participation – seems a somewhat curious creature. Of which more in a moment.

In this respect, second, another statistic is noteworthy. In the 1970’s and 1980’s, it was said that only a small proportion of the populace “knew” their parliamentary representatives, except where they were major public figures. Struck by this at the time, we did a preliminary survey, asking the question of 105 people in five southern villages. Around 55% said that they had no idea of the person concerned. Another 30% could give a name, but nothing else. Just under 15% answered in the affirmative. Yet more remarkable was the fact that over a quarter offered, unsolicited, that it made little real difference: that BDP members of parliament were the advisors and councillors of the president and that he was responsible both for appointing them and for their actions. Echoes, again, of a model of governance whose genealogy stretches deep into setswana, “Tswana ways.” In point of fact, politicians have become less anonymous in recent years. Still, one part of the idea, that a leader is responsible for the personnel of his/her regime, remains firmly intact.

Third, notwithstanding low voter turn out and the relative anonymity of political representatives, election meetings held by the BDP tended to draw crowds in the countryside. Among the opposition parties, in contrast, only those visited by important personages – an ex-chief back home, a charismatic with a big following, and the like – were well attended. We sat at many with only the candidate and a few of his friends. At BDP meetings, too, local people expected the president or a “close advisor” (i.e. a cabinet minister), to present themselves. Constituency politicians, those parliamentarians whom they “did not know,” were not good enough. After all, and this is the point, these meetings were knowingly modeled on the kgotla, that space of intersection between civil society and the state, between the public sphere and the politics of incumbency. Their object was not just to discuss matters of social concern, to play at popular, consultative democracy. It was also to evaluate the performance of the president and his party. And to hold him accountable for the extent to which the BDP had met the demands of good government. In this light, it seems injudicious to conclude, as van Binsbergen (1995:25-8) does, that the appeal to “the kgotla system,” dubbed a “neotraditional facade,” is merely a cynical effort by an authoritarian “state elite” to subjugate, appropriate, and manipulate local institutions. This was not the spirit in which Masire went to the country in 1984, nor the tenor of the BDP meetings which we attended over the years. Perhaps, though, it is the trend of the present and future. But that is another story.

In both their poetics and their politics, BDP election meetings evoked earlier discourses of chiefly authority. Speakers tended to line up into blocs of pro- and antagonists – the former being local party members, the latter, a coalition of dissent – surrounded by an unaligned public. Most of them spelled out the requirements of good government, typically in formulaic terms and in the authorial name of the transcendant “we” of nationhood and/or setswana. And then they offered their appraisals, often in starkly frank, pragmatic prose, always in the first person singular. In so doing, depending on their political positioning, they either proclaimed a convergence or a divergence between ideal and reality. Supporters, in particular, told a teleological tale of improvement. They spoke of the very successful “material performance of the post- independence state” (van Binsbergen 1995:27); also, usually, of the disbursement of resources “to all parts of the country equally” and the absence of clientelism (Charlton 1993:341). Others disputed just these things. The specifics of their counter arguments were contingent on place and circumstance, but they were frequently couched in accusations that government had “forgotten them.” (A popular pun in the south played on the name of the capital, Gaborone, named for the local chiefly dynasty; BDP critics called it ga re bone, “it does not see us.”) On both sides of the debate, however, there was the tacit assumption, utopian perhaps, that the BDP could only expect to enjoy legitimacy and the cooperation of the populace to the degree that it established the quality of its governance.

This is not to say that the electoral process mimicked the workings of the kgotla, past or present. The politics of the nation-state were not those of the chiefship writ large, nor are they today. Nonetheless, they did converge in two things. One was a deep aversion to autocracy at all levels of governance; hence Khama’s insistence that to “fetter criticism” is “against…Setswana tradition.” The other was the unspoken conviction, widely distributed across the various publics of Botswana, that substantive democracy depended on the simultaneity of (i) discourses of policy, seen here, as we said, to be the product of deliberative processes, not of partisan interest; and (ii) discourses of accountability, in which the proportionate relationship between performance and power was negotiated. The outcome of that negotiation, expressed in a quantum of sovereign authority, might have been heavily influenced by offstage dealings, by the capillary workings of the state, and by the social capital mobilized by ruling elites. And it might have been perverted by the covert forms of authoritarianism of which van Binsbergen speaks. But, for now, what is significant is this: underlying all public spheres was a civic culture that specified the means of producing a certain kind of participatory politics, a politics grounded in an articulate, popular ideology of good government.

In this civic culture, it will be clear, elections were important to the degree that they opened up a space, periodically at least, for substantive democracy. On the other hand, voting – procedural democracy – was much less salient, save at moments of crisis. Which is why people in the countryside would attend protracted political meetings and then often not cast a ballot, or do so more to express their dis/approval for the governing party than to exercise choice. Thus, for example, in 1974, the Botswana National Front (BNF) candidate from Barolong, O.B. Marumolwa, voted for the BDP – against himself. After hearing the president and a cabinet minister speak at several meetings, and give account of their performance, he declared that they should remain in office. “You do not just remove a ruler,” added Marumolwa, himself of royal descent.

This brings us, fourth, to the curious character of political parties here. Recall Colclough and McCarthy’s (1980:41) comment that the BDP was “not a mass party” at all; that it barely existed between elections; that it was, more than anything else, an immanent reservoir of support centered on the president and his cabinet. Nor, for their part, have any of the minority parties been an enduringly significant presence in the public domain. Even at their most active, these parties have served less as coherent ideological alternatives to the BDP than as a critical opposition, pure and simple. Some of them have been odd ideological hybrids. The BNF, for one, grafted a “traditionalist” wing, headed by a former chief, onto a “radical” one, led by a left wing Euro-intellectual. Custom and communism partying together is hardly what Weber had in mind in his classic typification of this species of voluntary association. In fact, both the BDP and BNF seem to have behaved more like the factional blocs we encountered in kgotla. This impression is reinforced by their conduct in the national assembly (Colcough and McCarthy 1980:46):

[T]he daily business of the National Assembly is conducted in a manner closer to the best of the African one-party states than to the Westminster model. The alignment is not so much the government benches against the opposition as Ministers against the backbenchers. Sometimes, indeed, opposition members are seen to support the Government when its own backbenchers are critical. Thus the role of the National Assembly, like that of the traditional Kgotla, is to audit proposals made by those in authority: to approve them and occasionally reject them. The Ministers respect this function of the assembly.

Talk here, once more, of a one-party state, and its juxtaposition to the workings of the kgotla, brings us full circle to the problem with which we began, and to the denouement of our argument.


Put together these various points and it will be clear what the call in the 1970’s for a one-party state was all about. It was an argument, in effect, against procedural democracy. Against democracy as the mere exercise of electoral options. Against the idea that freedom may be equated with choice. Against democracy, to return to Gertla Riviero, as a small idea, the kind of European export that promises the world and delivers Kentucky Fried Chicken. Given their own conception of participatory politics, their own ideologies of sovereign authority, legitimacy, and accountability, it is obvious why so many citizens of Botswana were alienated by the Western model, at least as presented to them. And why, by threatening to confine mass public involvement to a fleeting season every five years, it opened up a chasm between the state and civil society. For some, the very fact that the BDP leadership was so keen to sustain a Euro- styled multi-party system was itself an indictment.

More positively, the agitation for one-party government – towards which, interestingly, the national assembly was then moving in its own routine procedures – was a demand for a (re)turn to substantive democracy, to a civic culture in which participatory politics would be the stuff of everyday life. And in which the ruling regime was authorized to act for the nation in proportion to its warranted performance in office. Put another way, it called for a vernacular, indigenously rooted version of the kind of liberal democracy that Euromodernity has long idealized but scarcely realized – let alone implanted successfully elsewhere, especially when other interests have intervened. In hindsight, the gesture might appear to have been utopian, quixotic even. It also dated to a particular moment in the early history of this postcolony. But it gave voice to a deeply felt critique of taken-for-granted European political practices and institutions.

That critique spoke of a specifically African alternative, one that demanded not less popular sovereignty but more, not less accountability but more, not just choice but a public culture of criticism. All of which, of course, the global north has been moving steadily away from in recent times; prescient here is Julius Nyerere’s piquant comment, made already in the 1960’s, to the effect that the United States has “only one political party, but…[has] created two versions of [it].”26 Euro-American heads of state tend these days to act with ever greater impunity, to claim ever wider executive authority, and to promise as little government as possible. Concomitantly, large numbers of their citizens appear willing to forego freedoms, sovereignty, and the rule of law in the name of security and material well-being; vide the Patriot Act in the USA and the introduction of detention without trial in the UK, both post-9/11 measures that recalled the days of high apartheid in South Africa. Except in moments of rupture, moreover, levels of political involvement in the north seem steadily to wane, amidst accusations of epidemic apathy. In some European countries – Spain, Portugal, and Sweden being notable cases – an even smaller proportion of voters are currently able to name electoral candidates than was the case in Botswana in the first years of its independence (Norris 2004:230-48).27 As citizens of that nation-state sought ways to move from procedural toward substantive politics, so the West seems intent to move in the opposite direction.

In September 2009, a public intellectual and journalist in the USA, well known for his centrist political views, could quite plausibly title a widely syndicated essay on contemporary America, “One-Party Democracy.” Echoes of Nyerere, several decades on.28

The process that we have described here, we reiterate, was firmly located in the social realities of Botswana at the time: in its comparative ethnic homogeneity, its small size, its proximity to a particular historical past, all of which made the dream of a demos founded on popular sovereignty and direct state accountability appear eminently viable. These realities do not obtain everywhere. To the contrary: Botswana was, and is, relatively unique. And yet the vernacular political forms found there bear strong similarities to others in Africa (cf. Chabal 1986), some of them clearly visible, some submerged, some violently suppressed. Which raises a familiar conundrum, if in unfamiliar terms: Why it is that, for the most part, “democracy,” however it may be defined, is so fragile across the continent? What is it that intervenes between the conditions of its possibility, which are patently present, and its practical realization? How is it that the possible is rendered, if not quite impossible, then so difficult to accomplish? Why, where “democracy” may be said to prevail in the nation-states of the global south, does it seem more procedural than substantive, more “thin” than “thick”? Could it be that Euro-America’s contemporary move in the same direction, toward a “thinned out” version of representative government, provides a clue? That Africa has merely seen the emptying out of the large idea, its reduction to a small one, before the global north? That, in this regard too, the latter is evolving toward the former? And for the same reason, namely, that politics itself is escaping the formal public sphere and the institutions of state more and more as it migrates elsewhere. Could the de- democratization of north and south simply be a devolutionary counterpoint coming to us everywhere as part of the neoliberal age — an age in which, Archbishop Ndungane of Cape Town recently argued,29 citizens everywhere are valued purely as “voter fodder,” in which “good government, transparency, accountability, integrity and honesty” are known largely by their absence? If so, does it not demand that we address this counterpoint in taking on the Big Question of Theory, ca. 2010: Wherein lies the future of politics and the public sphere, sui generis, as the new century unfolds? Is it, as we have begun to suggest in previous chapters, in new social and religious movements and other forms of mass action, in politics of life, their strident mobilization of “the street,” their ever more assertive resort to lawfare, their deployment of the internet, and all the other means of experimental insurgency that have emerged so powerfully in the south and appear to be migrating northward?

Other Essays