September 23-25, 1999
University of Kansas
Mikael Karlstrom
University of Kansas
Department of Anthropology
Thank you very much John. I won’t actually focus entirely on civil society in my talk. I’m going to try to cast my gaze a little more widely than that, although I will come back to the civil society topic. The question that was posed to me for this talk had to do with the relationship between democratization and social trends in Africa over the past decade. Thinking about it I came up with a feeling that it’s not really possible to generalize about social trends in the way that it is about economic or political trends, and maybe this is just because I’m an anthropologist and anthropologists are somewhat reluctant to make large-scale generalizations, but I think there’s also a problem with the available data and what I want to focus on is the character of that problem and what might be done about it. At yesterday’s keynote lunch we heard an eloquent appeal to African academics to democratize the production of knowledge and art by returning to African languages and thereby to an active engagement with African cultures. Today I’ll appeal to academics interested in the democratization process to pursue an analytical engagement with African cultures on the grounds that democratization cannot be understood without understanding the broader social and cultural contexts in which it takes place. My argument will draw heavily on my research in Uganda where democracy has been a topic of much debate over the last 15 years or so. I’ll begin therefore with a brief overview of the Ugandan democratization process which has followed a somewhat different trajectory from that of most African countries in the past 10 years.
The latest wave of African democratization began in the early 1990s, as we’ve discussed in this conference, when longstanding autocrats were challenged by homegrown democracy movements and by donor pressure and often forced to hold either national conferences or multi-party elections or both. Uganda’s democratization process began earlier than this with a popular guerilla struggle in the early 1980s against a regime that was elected in an election that was generally viewed as fraudulent. Although a lot of popular energies went into the struggle, the democratization process since the current national resistance movement and president Yoweri Museveni came to power in 1986 has been considerably more state-led than most. Museveni and the NRM formulated a program of controlled and gradual transition to the full form of Western liberal democracy in the late 1980s. Since then they’ve considerably slowed the process to an even more gradual pace, some would say grinding it entirely to a halt, but still I think most would acknowledge that some gains have been made. Local government elections and indirect parliamentary elections were held in 1989, direct elections for a constituent assembly to write a new constitution in 1994, direct presidential and parliamentary elections were held under that new constitution in 1996, and each of these was deemed reasonably free and fair by external observers and judged to be legitimate by most Ugandans. What is controversial about the Ugandan model is the NRM’s reluctance to allow political parties to operate freely and to contest elections. Instead they’ve sought to institutionalize what they call a no-party or a movement system of democratic governance. They argue that political parties are a feature of Western capitalist democracies where class divisions give ideological content to political party platforms. In underdeveloped African countries with tiny bourgeousies and huge rural peasantries they argue that political parties can’t develop principled political ideologies, but only struggle for power by manipulating such local partisan loyalties such as tribe and religion. In support of this they site Uganda’s own political parties, which undeniably were formed on the basis of regional religious and tribal cleavages and which many Ugandans feel have manipulated and enflamed those divisions for purposes of electoral victory. According to the NRM therefore, Uganda needs to follow an alternative path towards democratization without this sort if divisive competition between political parties. On these grounds then the NRM has maintained restrictions on certain political party activities since 1986. Most significantly parties are barred from organizing public meetings and delegates conferences from issuing new party memberships and from coordinating election campaigns. This has not by any means evicted Uganda’s existing political parties and party leaders from the political process. A number of prominent party politicians were included in the NRM’s inaugural coalition or “broad-based” government which was central to their original movement philosophy. Party leaders and supporters are free to take advantage of the considerable freedom of the press that has been maintained since 1986. Party politicians as individuals are free to run for office and they constitute in fact a quite vociferous opposition minority in the current parliament. The old political parties also sponsored a joint presidential candidate against Yoweri Museveni in the 1996 elections, albeit under considerable logistical and structural handicaps. In the year 2000 a national referendum will be held on the question of whether to return to a system of multi-party elections or to retain the current no-party system for another five years. Judging from many secondary accounts and also from the loud protests of the suspended political party leaders themselves, this moratorium on political party activities would seem to be a self-evident abrogation of democratic rights and of the universal human right of freedom of association. Whatever we might think as outside observers, however, ordinary Ugandans, at least most of those among whom I did research in the early 90s, don’t see things quite that way. In fact I would argue that the entire domain of political reform, democratization, and state legitimacy looks very different from this perspective, from sort of a perspective from below than from the bird’s eye view of outside analysts and even from that of the Kampala political elite, both government and opposition, and I’m going to try to give a very synoptic version of that view from below highlighting certain discrepancies or differences between an outside perspective and this one. First of all, the understanding of democracy itself as a concept among the rural Baganda among whom I did my field work in the crucial region of Baganda in central Uganda, the understanding democracy itself has been assimilated to local conceptions of political legitimacy, and these are based on an enduring royalist or kingship oriented political culture in what is a kingdom that dates back to the 15th century. The standardized vernacular equivalent of the tranliterated word for democracy can be most literally translated as “freedom of the people” rather than any version of “rule of the people,” so there’s no notion of popular empowerment involved in this conception. The conception does involve a sense of freedom from oppression, but this isn’t a strictly liberal notion of freedom because it doesn’t conceive state oppression as an excess of state power so much as a kind of deficit or loss of legitimacy of state power, which is felt to be necessary in order to maintain a stable socio-political order. There’s also a notion of freedom of speech, although again it’s not an egalitarian notion of freedom of speech in an egalitarian public sphere, but rather a freedom of speech of the subjects towards the ruler, whose role is to listen to the subjects and address their concerns. It also contains a conception of fair judgement and justice by the transcendent ruler who precisely stands above the divisions and cleavages in the population of his subjects. So each element of this conception reinforces the sense that rulership is singular and transcendent rather than being anchored in popular choice and that mechanisms of accountability are founded on the solidity of that transcendence. The concept of democracy you might say has been converted, in this instance, into something like its opposite. Second, multi-party elections did not figure, either in local definitions of democracy nor in popular political aspirations more generally. Elections themselves were viewed as desirable but by no means crucial to political legitimacy and good governance. Political parties, on the other hand, were widely despised and viewed as anathema to the principles of democracy as locally conceived. This hostility towards political parties, of course, suited the NRM very well, but it wasn’t one that they created. Uganda hostility toward political parties has a fairly deep history and was created or overdetermined by a combination of historical circumstances, local experience, and socio-political ideology. First of all, it was the political party system after independence that most people blamed for the abolition of the kingdom itself in 1967, which was carried out by one political party, but the blame is generalized. At the local level Uganda have experienced national elections, those that took place in the early 60’s and another one in 1980, as highly disruptive and violent, with villages and even families bitterly divided along party lines, fields and homes burnt to punish people who voted the wrong way. Finally, at the ideological level political party competition does not fit well into a conception of the properly functioning political order as founded on regulated competition of subjects within an inclusive hierarchical system under the unifying aegis of the transcendent and immovable king. Against this conception, political parties seem to be engaged in a kind of divisive, unregulated form of competition in which the political apex, which is supposed to be untouchable, is the prize of the competition rather than the regulating principal, and this results in an exclusionary hierarchy in which the winner takes all the spoils. For these reasons the NRM’s decision to suspend political parties and conduct elections on the basis of individual candidacies played very well in rural Buganda, regardless of the fact that the NRM’s own reasoning, a more Marxist-inspired reasoning, in support of this, was not really understood by Baganda in rural areas. By the same token the NRM’s broad-based “movement” government at the national level was perceived as just the sort of unifying inclusive political order that was needed in Uganda after a period of civil war in order to promote the ideals that people encoded in the local concept of democracy. Third, rural Baganda viewed and treated elected officeholders in the higher reaches of government, such as members of parliament and ministers, not as their representatives or as public servants in any sense of that term, but rather as transcendent dignitaries, deserving or, as the case may be, undeserving of lavish exaltation and ritual generosity from their subjects whenever they came to visit from the capitol. Such visiting dignitaries were greeted in the villages in much the same ritual idiom of staged performances and competitive gift-giving that is mobilized for the king when he tours his kingdom, but it is worth noting that despite the structural assertion of hierarchy and political inequality in these ritual forms, they don’t only serve to glorify political subordination, or they don’t serve to glorify political subordination, in a sense. What they enact, rather, is a culturally specific mode of political accountability. Instead of the liberal democratic exchange of individual votes for a satisfying form of political representation, this is a collectivized ritual exaltation being exchanged for open and communicative governance, for fair judgement and for attention to local concerns. The ultimate sanction against an unsatisfactory officeholder is the refusal to stage this kind of reception, which deprives him or her of the stamp of local legitimacy. Fourth, local government reforms were far more salient for rural Baganda than such national level issues as the timetable for writing a new constitution or a return to a liberal democratic electoral procedure. The NRM had introduced a democratization of local government in 1986, replacing most of the functions of the previously appointed civil service chiefs with a nine member at the time called Resistance Councils who were elected by the local population. These were very popular in Buganda, at least in the early 1990s. They were generally considered very democratic by local standard, not because they were elected but because they constituted the kind of inclusive hierarchical non-oppressive structure of local self-governance that people held as an ideal. In fact, one of the reasons that villagers were so reluctant to have political parties reintroduced was because they wanted to preserve the viability of these new local institutions, since most people believed that political party divisions would factionalize them and destroy their inclusive character. Fifth and finally, the easing of state restrictions on ethnic political and cultural institutions was the highest priority of most rural Baganda with regard to the restructuring of the national public sphere. In particular they sought the restoration of the Buganda kingship and the revival of clan solidarities and associations that had declined since its abolition. This avenue of collective engagement with the state was viewed as by far the most promising for a peaceful and workable Ugandan political order. Most people were convinced that if political parties were introduced they would once again try to abolish the kingship. Although the NRM were initially reluctant to allow the kingship to be restored, they eventually did so in 1993 in a move that was timed very strategically so as to win them the Buganda vote in the 1994 elections for the constituent assembly. It should be clear by now the rural Baganda haven’t been yearning since colonial times for the blessing of Western liberal democracy, nor have they by any means been the dupes of their various and variously undemocratic post-colonial rulers. They’ve responded to state initiatives according their established patterns and conceptions of authority and political legitimacy, and although these patterns and conceptions are not what we would call democratic, they do include mechanisms of accountability and conceptions of a popular voice to which it is incumbent upon rulers to listen and respond. I am not going to argue that this political culture is intrinsically and unalterably inimical to democracy. In fact, experience with the NRM’s form of controlled democracy is changing people’s orientation towards elections, if not towards political parties. Nor am I going to argue that the NRM’s policies are democratic simply because they correspond to local conceptions. In fact, since about 1994 I’d say that the NRM government has proven a rather poor custodian of its own best democratic innovations. They’ve failed to fully institutionalize the local councils as functioning organs of democratic control at the lower levels. An absence of resources and an absence of further development and experimentation has caused a kind of local institutional sclerosis in a lot of places. In practice the NRM abandoned the broad-based approach to national governance once they had won their first decisive electoral victory in 1994, and they haven’t really developed the theory of no-party governance of as a genuinely open and democratic alternative to multi-party democracy. More recently they seem to have allowed the privatization of a great deal of Uganda’s public sector to turn into a nearly Russian-style scheme for the enrichment of senior NRM insiders. These failures, along with an unpopular and expensive war in the Congo, may well make the referendum to be held on political systems next year into a very real fight, and also as the first real test as whether the NRM’s democratic commitment extends as far as accepting electoral defeat. But the disappointing recent performance of the NRM is less interesting for my purposes here than the lessons that I think can be learned from the rural reception of their earlier initiatives. NRM policies were successful in significant part because they resonated with local popular conceptions and practices, at least in this critical region of Uganda where I did my research, and this despite the fact that the NRM’s reasons for pursuing these policies bore no relationship to the Buganda royalist political ideology towards which the NRM itself was extremely scornful. The lesson from this I would say is that local culture does matter for the reception and the prospects of democratic reform efforts, more fully that institutional democratization takes place in a socio-cultural context which inflects and domesticates its implementation and reorders the priorities of state and political reformers. Local responses and judgements as to the legitimacy and democratic potential of reform efforts cannot be predicted as if from some universal handbook of democratization. One response to this argument might be to ask “Why should we worry about what rural Africans think about democratic reforms patently don’t understand them?” In fact, I would say that most analyses of African democratization do proceed as though rural non-elite Africans don’t really matter. I can answer this question at several levels, but only briefly. One answer would be a matter of principle. Rural Africans constitute a majority of the demos, or the citizenry, who are supposed to be the beneficiaries of democracy. What’s the point of advocating democracy if we don’t really care what most of the people think about it? A second answer would be a matter of analytical pragmatics. Democracy actually makes popular conceptions and perceptions relevant in determining political outcomes. In fact, president Museveni’ssolid victory in the 1996 presidential elections was secured by a huge majority of votes from rural Buganda despite the growing dissatisfaction of Ganda elites and the open support of the Ganda royalist leadership for the opposition. This was an entirely unexpected outcome, even by me who had studied these things, and I think one of the crucial variables in explaining it is precisely the popularity of NRM policies over the previous ten years. A third answer has to do with the potential truth value of local perceptions. These views speak to a certain experience of Uganda’s political history which may not fit well into analytical frameworks of political evolution, but which are nonetheless real. As played out locally, so-called democratic national elections may seem none too democratic or even particularly civil. How does it promote the interests of an impoverished Ugandan peasant to choose between two wealthy members of the urban elite as his or her member of parliament, especially if, as often happens, that MP visits his constituents only at election time and then concentrates on outbidding his rivals in the currency of beer and the distribution of petty cash, and if intimidation and violence are the central means of local vote-getting? Like presidential mandates and parliamentary majorities, these are facts of many democratization efforts in Africa and our analyses should take them into account. Finally, a fourth answer has to do with the character of democratization as a cultural as well as an institutional process, and of democracy in the non-West as a syncretic rather than an evolutionary outcome, but I will return this point after making a few others. The complex and sometimes surprising reception of democratic conceptions and practices in Buganda leads me to wonder about the deeper character of the alleged groundswell of popular support for democratic reforms across so much of the Sub-Saharan African continent in the early 1990s. Did Africans outside the urban elites of such countries as Benin, Mali, Zambia, Malawi, Kenya, and so on really formulate their political aspirations under the heading of democracy, and if so do we really know what they meant by this? Do we know how their conceptions of democracy articulated with other elements of their political imaginarian(?) ideology? Did multi-party elections figure as centrally in their political aspirations as in those of political reformers in the capitols and in the analyses of outside observers? And when elections were staged, did peasants behave in expectable ways as individualized citizen-consumers choosing between competing party platforms? To take only the last of these questions, the few ethnographic accounts that I know of of election conduct and attitudes, one in northern Nigeria by William Miles, one in Senegal by Charles Schaffer, one in Malawi by Harry Eglund, and one in Sierra Leone by Marian Firmi (?) all suggest a considerable density and cultural inflection and domestication in the actual local conduct of African elections, but these accounts are few and far between, and we would need many more answers to questions like these in order to get a real grasp of the social impact and articulation of African democratization. At the very least the bottom-up perspective on democratization suggests that our institutional measures of degrees of democracy may fail to capture important dimensions of the actual social articulation of the political reforms. In so far as there is variability in the political aspirations of different African peoples, the relative weighting of such aspects of democratic reform as national elections on the one hand and more participatory local government on the other, multi-party competition versus ethnic reparticipation, alternation in power at the top of the political structure versus decentralization of power toward the bottom, these may all vary as well. This is not to despair of being able to judge the progress or regress of any given democratic process, but it does suggest that a denser and more complex analysis than the usual checklist of democratic features is needed. Now the one new analytical departure in the past decade or so with regard to understanding the way democratic reforms articulate with the broader social order is actually the revival of an old perspective. This is the civil society paradigm. This paradigm has usefully broadened our perspective on democratization, calling attention to certain dimensions of social transformation that have historically underpinned the shift toward and consolidation of democratic forms of governance in Europe and America. But I want to suggest that rather narrow limits have been set for the capacity of this paradigm to capture the broader cultural context of democratization efforts by the particular construction that has been placed on it when it has been applied to post-colonial Africa. Civil society is a notoriously slippery concept. This is due in considerable part to the complex theoretical genealogy of the term in 18th and 19th century European political philosophy. In its Africanist incarnation recently, however, the concept seems to have been stripped to its empirical bones by defining it as a universal objective category of interest group formation. That is, as the domain of non-state voluntary associations with demographically cross-cutting memberships and egalitarian liberal aims and internal practices. Such associations articulate with the state, according to the theory, in such a way as to both legitimate and strengthen the state in its legitimate functioning and also to check its excesses. Although the empirical specification of civil society as consisting of voluntary associations does help to make the concept more easily applicable in empirical research, and has produced quite valuable empirical research in regards to African associational life, I would argue that it has done so at some cost to the analytical adequacy of the concept. In Uganda, for instance, it’s noteworthy that some of the most interesting and promising developments in associational life and in the positive rearticulation of state and society since 1986 would fall outside of the civil society category as it’s commonly defined in Africanist work. First, the revival of clan associations fits uneasily into this established definition of civil society because it’s kin-based and therefore falls on the private side of a pre-supposed public/private dichotomy. Yet clans have long played a leading role in public and political life in Uganda, and their revival is explicitly conceived in terms of a kind of return to civility after Uganda’s years of brutality and violence. Moreover, the single biggest victory of any non-state pressure group since 1986 was the NRM’s concession on the restoration of the Buganda kingship, and the campaign for this was led precisely by the association of Ganda clanheads. Secondly, the restoration of the kingship would also be an unlikely candidate for inclusion in the civil society category. It’s based on ethnicity which is an ascriptive rather than a voluntary identity, and monarchy is presumably an anti-democratic political form. Yet the restoration of the kingship has probably done more than anything else to reintegrate the crucial Buganda region, once deeply alienated from the state, into the national polity. It’s allowed Baganda to rethink the assumption that loyalty to the king and loyalty to the nation are somehow in contradiction to one another and begin to conceive them as compatible, thus in some sense to voluntarily choose a more civil way to enact this non-voluntary or ascriptive identity. And finally, the local council system would not be classified as a part of civil society because it began as an institutional creation of the state, and civil society is unanimously defined as the non-state associational domain, but it’s a vast overestimation of the capacity of the Ugandan state to think that it can oversee these councils at the village and parish level in any direct way. In fact, their highly variable constitution and functioning on the ground is almost entirely the product of local energies, voluntary energies as it turns out, since the lower level councilors are, or at least were until recently, unpaid. In a sort of unintentional return to both pre-colonial and colonial patterns of local governance I would say the local council system thoroughly blurs the modern institutional boundary between state and society. In each of these cases, then, the failure of the civil society paradigm to capture important developments in the state-society relationship in Uganda can be traced, I think, to unexamined presuppositions built into the institutional framing of the concept in terms of voluntary associations, presuppositions as to the boundaries between public and private, between ascriptive and voluntary allegiances, and between state and society. In other words, the paradigm falls short because it attempts to project a category of Euro-American historical development onto African socio-political life without first asking whether the social, cultural, and structural preconditions for the salience and prominence of that category are in place. What about civil society as it is commonly defined? Has it contributed to Uganda’s democratization? The NRM has considerably opened up the space for voluntary associations in Uganda since 1986 and donor funds have flooded into some of these associations, but although they may prove important to Uganda’s democratization process in the future, several recent empirical studies suggest that they don’t currently have such a capacity. This kind of pessimism doesn’t seem to be restricted to the civil society literature on Uganda. In fact, although the academic application of the civil society paradigm to Africa was initially a move of renewed analytical optimism after the dashed hopes of more state-centered approaches, it shows sign now I think just a decade later of turning into yet another discourse of Afro-pessimism, with less emphasis on local associational energies than on the need for massive infusions of financial assistance to encourage and consolidate a weak voluntary sector. I wonder, has this paradigm arrived at an impasse? One possible way out of such an impasse, I think, might be to return to the origins of the civil society concept. The concept began as a framework for thinking about ways to overcome new socio-political oppositions and fragmentations arising out of the commercial, industrial, and political revolutions of the late 18th century, oppositions such as those between individual and collectivity, between public and private, and, as in contemporary Africa, between state and society. In this sense civil society stands as a normative question rather than as an empirical answer to a question. The question: What social mechanisms can bridge the gap and ease the alienation between state and society? Voluntary associations may be one answer to this question, but they are by no means the only conceivable answer. Let me return briefly in concluding to my broader topic, which is the complex social articulation of African democratization projects and our analytical capacity to capture that process. I’ll conclude with a few brief suggestions as to the direction that I think a fuller study of African democratic reforms and potentials might take. First, democratization, like state formation and nation-making, should be viewed as a cultural transformation, and not simply as a political or institutional one. This cultural dimension of democratization is if anything even more crucial to the understanding of contemporary non-Western democratization than of the original European and American experience. This is because democratization can no longer viably proceed as a gradual step-wise extension of the franchise beginning with the bourgeoisie or property owning and urban class and male gender. The generalization of the concept of democratic citizenship in the intervening 200 or so years is now part of the democratic legacy that has to be accommodated by democratic movements and reforms. It’s therefore not merely a cultural transformation of elites that must be attended to, important though that may be, but also a cultural transformation of entire populations. Secondly, rather viewing the worldwide spread of democratic conceptions and demands as a process of political evolution in a pre-determined trajectory toward a Western endpoint, we should view it as part of the cultural process of globalization, indeed as one of the most significant contemporary areas of such cultural globalization. And here I take globalization to be a dynamic process of syncretism and hybrid formations rather than the uni-directional imposition of Western culture imagined in an earlier discourse of Western cultural hegemony. There is considerable research being conducted now by historians and anthropologists on the complex processes whereby certain key elements of the Western ideological complex have been locally received and reconfigure. Christianity and nationalism have probably come in for the most sustained evaluation and scrutiny, but democracy has been almost entirely neglected from this perspective. Applying a globalization perspective to democratization can also, I think, help us to complicate our sense of the endpoint or telos of democratization and recognize that there are a multitude of possible broadly democratic outcomes, just as there are multiple syncretic forms of Christianity and nationalism. Indeed, this kind of complexity is true even within the West which is so often ideologically homogenized in comparative accounts. Third, the emphases on cultural transformation and on globalization should be combined in such a way as to avoid a radical cultural relativism as applied to political culture. The culturalist argument that non-Western political cultures are absolutely incompatible with Western democratic institutions is a relic of a monolithic and static view of culture which has been largely discarded by anthropologists. Instead any particular political culture needs to be regarded as dynamic and internally complex with elements of both authoritarian and anti-authoritarian ideology. When Western democratic ideas are locally appropriated they’re assimilated to local elements of anti-authoritarianism and they reciprocally act to amplify those elements within the existing political culture. The contemporary world-wide relevance of the specifically Western forms of anti-authoritarian thought does not stem from their intrinsic superiority. Rather it stems from the fact that the modern nation-state to which those concepts pertain has already been globalized in the post-colonial world order. Local mechanisms of accountability and reciprocity between rulers and subjects are often simply inadequate to the scale and material resources of the modern state. This is why they may need a creative add mixture of elements from the Western democratic tradition in order to effectively impose accountability on those who wield the power of the state. My plea for a more culturally informed approach to the study of African democratization is not a call for anthropologists to replace political scientists. On the contrary it is a call for collaboration. Anthropologists have made their own contribution to the neglect of the cultural dimension of democratization and post-colonial state formation by taking a very limited interest in how these macro-level processes intersect with the micro-worlds in which they tend to conduct their research. If the view from the political scientist’s perch often renders opaque or invisible the cultural understandings of the people down below, the anthropologists view from the village often seems to relegate national political events to the distant margins of the lives of ordinary Africans. A fuller picture of the social articulation of democratic forms in Africa will require corrective surgery for both these forms of perspectival myopia. Thank you.