Research for this article was funded in part by the Human Sciences Research Council. I would like to thank Kathryn Oberdeck, Derwin Munroe, and two anonymous readers for helpful comments on earlier drafts.

¹ Young, M. C., “Zaire: Is There a State?,” Canadian Journal of African Studies, vol. 18:1 (1984), 80Google Scholar.

² See especially Chabal, P., ed., Political Domination in Africa: The Limits of State Power (Cambridge: Cambridge University Press, 1986CrossRefGoogle Scholar); Rothchild, D. and Chazan, N., eds., The Precarious Balance: State and Society in Africa (Boulder: Westview Press, 1988Google Scholar).

³ See especially J. McGaffey, “Economic Disengagement and Class Formation in Zaire,” in D. Rothchild and N. Chazan, eds., The Precarious Balance, 171–88; Bratton, M., “Beyond the State: Civil Society and Associational Life in Africa,” World Politics, 41:3 (1989), 407CrossRefGoogle Scholar–30. In much historical research the political realm has given way to the cultural realm in analysing social power relations. See, for instance, Hobsbawm, E. and Ranger, T., eds., The Invention of Tradition (Cambridge: Cambridge University Press, 1983Google Scholar); Vail, L., ed., The Creation of Tribalism in Southern Africa (Berkeley: University of California Press, 1989Google Scholar).

⁴ Guyer, J. (in “Representation without Taxation: An Essay on Democracy in Rural Nigeria, 1952–1990,” African Studies Review, 35:1 [1992], 41–79Google Scholar), makes this point from a rather different point of view but reaches much the same conclusions.

⁵ State-centered explanations became prevalent when the analytical shortcomings of modernization theory and underdevelopment theory became increasingly clear: in particular, their failure to account adequately for the impact of apparently autonomous state actions on processes of politics and development in emerging African polities. For compelling critiques of the systemic determinisms of modernization theory and underdevelopment theory, see Lonsdale, J., “States and Social Processes in Africa,” African Studies Review, 24:2–3 (1981Google Scholar), and Cooper, F., “Peasants, Capitalists and Historians: Review Article,” Journal of Southern African Studies, 7:2 (1982),285Google Scholar–9.

⁶ Sklar, R., “The Nature of Class Domination in Africa,” Journal of Modern African Studies (1979), 531CrossRefGoogle Scholar–52; Berry, S., Fathers Work for Their Sons (Berkeley: University of California Press, 1985Google Scholar); Joseph, R., “Class, State and Prebendal Politics in Nigeria,” in Kasfir, N., ed., State and Class in Africa, 21–37 (London: Cass and Co., 1983Google Scholar); Leys, C., “Capital Accumulation, Class Formation and Dependency: The Significance of the Kenyan Case, 241–66,” Socialist Register (1978Google Scholar).

⁷ Jackson, R. and Rosberg, C., Personal Rule in Black Africa: Prince, Autocrat, Prophet, Tyrant (Berkeley: University of California Press, 1982)Google Scholar; Callaghy, T., The State—Society Struggle: Zaire in Comparative Perspective (New York: Columbia University Press, 1985)Google Scholar; R. Higgott, “The State in Africa: Some Thoughts on the Future Drawn from the Past,” in Shaw, T. and Aluko, O., eds., Africa Projected: From Recession to Renaissance by the Year 2000?, 12–39 (London: MacMillan, 1985)CrossRefGoogle Scholar.

⁸ Migdal, J., Strong Societies and Weak States: State-Society Relations and State Capabilities in the Third World (Princeton: Princeton University Press, 1988)Google Scholar; Young, C., Ideology and Development in Africa (New Haven: Yale University Press, 1982)Google Scholar; Duara, P., “State Involution in Comparative Perspective,” in Breman, J. and Mundle, S., eds., Rural Transformation in Asia, 161–91 (Oxford: Oxford University Press, 1991)Google Scholar.

⁹ N. Chazan, “Patterns of State-Society Incorporation and Disengagement in Africa,” in D. Rothchild and N. Chazan, eds., The Precarious Balance, 121.

¹⁰ Thompson, E. P., Customs in Common (New York: The New Press, 1991), 2Google Scholar, ch. 3 and passim; Scott, J., The Moral Economy of the Peasant: Subsistence and Rebellion in Southeast Asia (New Haven: Yale University Press, 1976Google Scholar).

¹¹ In this sense, I use Thompson's and Scott's arguments as providing a starting point rather than as offering methodological models. Both would reject the structuralist elements of my argument, and both would be uncomfortable with my emphasis on language in making social relations.

¹² The tension between practices as discrete actions and the neo-structuralist concept of practice raises the problem of historical agency. If one bears in mind this important distinction, it is apparent that the process I describe is political, characterised by clashing claims and conflicting interests, and never determinate. I address this tension more fully below, at the level of state theory, and elsewhere, at the level of empirical analysis.

¹³ This follows Weber's basic definition (in “Politics as Vocation,” From Max Weber. Essays in Sociology, H. H. Gerth and C. Wright Mills, trans, and eds. [New York: Oxford University Press, 1946]) of the state as “a human community that (successfully) claims the monopoly of the legitimate use of physical force within a given territory” (p. 78, added emphasis in original). See also Jessop, B., “Capitalism and Democracy; The Best Political Shell?,” in Littlejohn, G., ed., Power and the State (New York: St. Martin's Press, 1978), 11Google Scholar, for a Marxist perspective.

¹⁴ Young, Ideology and Development, 73–74. Kitching, G., Development and Underdevelopment in Historical Perspective (London: Routledge, 1989), 4Google Scholar, points out that nationalism is assumed to be the basis for development thinking.

¹⁵ Regimes may be able to secure effective forms of social and political reproduction either by presenting the state as a kind of Hobbesian protector (much colonial administrative thinking was based on this political tradition) or by extending the state's capacity for coercion and control into the very core of social relationships (as in totalitarian or authoritarian traditions). Usually, state ideologies reflect some combination, forged in real political struggles.

¹⁶ The concept of the universalized state separate from (and dominating) civil society is developed in Hegel, G. F. W., Hegel's Philosophy of Right, Knox, T. M., trans. (London: Oxford University Press, 1952), especially 122CrossRefGoogle Scholar–60. For the concept of the state as idea in Africa, see Higgott, “The State in Africa,” 32. I do not endorse Hegel's teleology of the universal rational state emerging from the division of labour and the demands of war. As Marx perceived, states are inseparable from modes of domination, although Marx's view that superseding social and political domination meant getting rid of the state is scarcely more compelling in the world today. The importance of the Hegelian concept of state universality is his argument that the link between the universality of the state and the particularity of civil society is law, which establishes the “right as law” (see especially pp. 134–39). Here Hegel directs us to the political importance of struggles to define the relationship between right, knowledge, and law (and property) in molding political identities (group and individual).

¹⁷ See Hegel, Philosophy of Right, 155–60; Habermas, J., Legitimation Crisis, McCarthy, T., trans. (London: Heinemann, 1976Google Scholar), pt. I.

¹⁸ The most dramatic example of such a compromise is the comprehensive systemic crisis of sociopolitical systems in the Soviet Union and eastern Europe in recent years. Such crises demonstrate the intimate yet complex relationships between the social allocations demanded by the organisation of material production and distribution in society and those demanded by sociopolitical organisation and distribution. This is the burden of Habermas's, “legitimation crisis” (in Legitimation Crisis, McCarthy, T., trans. [London: Heinemann, 1976Google Scholar]). To hold and control the institutions of the state does mean to wield state power (Skocpol, T., States and Social Revolutions: A Comparative Analysis of France, Russia, and China [Cambridge: Cambridge University Press, 1979], 156CrossRefGoogle Scholar) but does not necessarily mean to exercise comprehensive domination over civil society. Where basic determinants of social continuity and change (such as control of labour at the point of peasant production or control of kinship relations in rural communities) lie outside of the state's domain, these must be brought within the ambit of state authority if the hegemony and revenue base of the state, on the one hand, and the political dominance and access to opportunities for accumulation of elites, on the other hand, are to be secured. See Cooper, , “Peasants, Capitalists, and Historians”; Allen, C. and Williams, G., eds., The Sociology of Development Societies: Sub-Saharan Africa (London: MacMillan, 1982), 176Google Scholar–9.

¹⁹ J.-F. Bayart “Civil Society in Africa,” in Chabal, ed., Political Domination in Africa, 117– 20.

²⁰ See Migdal, Strong Societies and Weak States, 10–15; Young, Ideology and Development, ch. 1 and passim.

²¹ For brief discussions, see Bates, R., Beyond the Miracle of the Market: The Political Economy of Agrarian Development in Kenya(Berkeley: University of California Press, 1989), 3Google Scholar; Breman, J. and Mundle, S., eds.,Rural Transformation in Asia (Oxford: Oxford University Press, 1991Google Scholar), xi–xii; O'Brien, D.Cruise, Dunn, J., and Rathbone, R., eds., Contemporary West African States (Cambridge: Cambridge University Press, 1989), 3Google Scholar.

²² J. Lonsdale, “Political Accountability in African History,” in Chabal, ed., Political Domination in Africa, 128; see also Lonsdale, “States and Social Processes,” 204. This moral calculus constitutes a crucial distinction between the nature of pre-modern and modern states. Some writers have observed that African states seem to embody elements of both pre-modern politics and modern states and, indeed, that the “preconceptions and categories of institutionalised public politics” ought to be abandoned in analysing modern African politics in favour of pre-modem “practices of palace politics”; cf. Jackson and Rosberg, Personal Rule in Black Africa, 1–12. However, the relations between state and society in pre-modern states were mediated, institutionally and ideologically, by a range of structured feudal (class) relationships very different from those mediating the relationship between the modern state form and society.

²³ In this respect, it may be useful, following Howe, H. and Ottoway, M. (in “State Power Consolidation in Mozambique,” in Keller, E. J. and Rothchild, D., eds., Afro-Marxist Regimes [Boulder, CO: Lynne Rienner Publishers, 1987]Google Scholar) to distinguish between a secure regime and a weak state.

²⁴ For a useful introduction to debates about the state that reveals this presupposition, see Skocpol, T., “Bringing the State Back In: Strategies of Analysis in Current Research,” in Evans, P., Rueschemeyer, D., and Skocpol, T., eds., Bringing the State Back In (Cambridge: Cambridge University Press, 1985Google Scholar).

²⁵ Mitchell, T., “The Return of the State” (Paper presented to conference on “Power,” University of Michigan, Ann Arbor, Michigan, 1992), 18Google Scholar.

²⁶ The use and interpretation of the term state are not easily separated, partly—though not entirely—because of the central conceptual position of the state in Western political theory. See Dunn, Western Political Theory, chs. 1—3.

²⁷ Here I follow Stepan's, A. argument (in The State and Society: Peru in Comparative Perspective [Princeton: Princeton University Press, 1978Google Scholar], xii–xiii) that the state has greater continuity than the government. For a telling example, see Greenberg, S., Legitimating the Illegitimate: State, Markets and Resistance in South Africa (Berkeley: University of California Press, 1987Google Scholar).

²⁸ Even in cases where citizenship is not formally extended to all, it is those to whom citizenship is extended whose social position defines the public realm. In general, the definitions of private and public have come down to us in terms of the res publica definitions embodied in Roman law and incorporated into the rechtstaat that provides and regulates a system of law, of taxation, and of administration. See Habermas, J., The Structural Transformation of the Public Sphere, Burger, T., trans. (Cambridge, MA: MIT Press, 1989Google Scholar), ch. 1.

²⁹ Jessop, B., State Theory. Putting Capitalist States in their Place (University Park, PA: The Pennsylvania State University Press, 1990), 360Google Scholar.

³⁰ Quoted in Jeffries, R., “Ghana: The Political Economy of Personal Rule,” in O'Brien, D. B. C., Dunn, J., and Rathbone, R., eds., Contemporary West African States (Cambridge: Cambridge University Press, 1989), 87Google Scholar.

³¹ Here I differ specifically with Skocpol (in States and Social Revolutions, 29), who views the state as “a set of administrative, policing, and military organisations headed by an executive authority” that “extracts resources from society and deploys these to create and support coercive and administrative organisations.” For Skocpol, the administrative and coercive organisations are the basis for state power. This institutional conception of the state is partial; there is ultimately a difference between the form of state power and the nature of state power.

³² Gramsci, A., Selections From the Prison Notebooks, Hoare, Q. and Smith, G. K., ed. and trans. (New York: International Publishers, 1971), 167Google Scholar–8.

³³ The normative content of this concept is of course not settled, nor is it simply a problem of politics and history. It invokes one of the central debates in western political thought–the ontological status of individuals.

³⁴ Frequently the claim to knowledge is associated with a claim to empirical verification in the tradition of positivist science.

³⁵ Zolberg pithily, A. makes this point in Creating Political Order. The Party-States of West Africa (Chicago: Rand McNally, 1966), 44Google Scholar: “Political thought is usually not formulated by professional philosophers or other contemplative men, but rather by practical politicians. It is not expressed in the form of carefully wrought treatises but rather in the form of addresses to announce, to explain and justify particular choices of policies.” For a provocative discussion of commissions of enquiry as a vehicle for public transcripts that convey state impartiality by disseminating access to knowledge, see Ashforth, A., “Reckoning Schemes of Legitimation: On Commissions of Enquiry as Power/Knowledge Forms,” Journal of Historical Sociology 3:1 (1990), 1–22CrossRefGoogle Scholar.

³⁶ Jackson and Rosberg (in Personal Rule in Black Africa) rightly draw attention to the importance of patronage and clientilism. However, their conclusion that politics has thus become de-institutionalised is unpersuasive, since state institutions provide the loci and channels for patronage. All states try to regulate society, and it is through the formal institutions of political incorporation and administration that this occurs.

³⁷ See Hughes, A., “The Nation-State in Black Africa,” in Tivey, L., ed., The Nation-State: The Formation of Modern Politics (New York: St. Martins Press, 1981Google Scholar) for a good general discussion of this point.

³⁸ See MacPherson, C. B., The Political Theory of Possessive Individualism. Hobbes to Locke (Oxford: Oxford University Press, 1962Google Scholar); Habermas, J., Communication and the Evolution of Society, McCarthy, T., trans. (Boston: Beacon Press, 1979Google Scholar), chs. 3 and 5.

³⁹ Nevertheless, as Phillips, A. shows (in The Enigma of Colonialism. British Policy in West Africa [London: James Currey, 1989Google Scholar]), the relationship between the colonial state and metropolitan capital was always uneasy.

⁴⁰ For a useful, if broad, typology, see Amin, S., “Underdevelopment and Dependence in Black Africa; Origins and Contemporary Forms,” Journal of Modern African Studies, 10:4 (1972), 503CrossRefGoogle Scholar–24.

⁴¹ In particular, states were—and have remained—charged with ensuring that labour markets operated with some efficiency to secure a flow of labour. See Yudelman, D., The Emergence of Modern South Africa: State, Capital, and the Incorporation of Organised Labour on the South African Gold Fields, 1902–1939 (Cape Town: David Philip, 1984Google Scholar); Phillips, The Enigma of Colonialism. On the politics of market regulation, see Bates, R., Markets and States in Tropical Africa: The Political Basis of Agricultural Policies (Berkeley: University of California Press, 1981Google Scholar).

⁴² This has been termed “primary” or “primitive” accumulation, and it is still widely discernible in Africa: Arrighi, G., “Labour Supplies in Historical Perspective: A Study of the Proletarianisation of the African Peasantry in Rhodesia,” Journal of Development Studies, 6:3 (1970),214Google Scholar; Bayart, “Civil Society in Africa,” 115–6.

⁴³ Mann, K. and Roberts, R., eds., Law in Colonial Africa (London: James Currey, Ltd., 1991Google Scholar); Phillips, Enigma of Colonialism.

⁴⁴ This is of course a general historical process; see Polanyi, K., The Great Transformation: The Political and Economic Origins of Our Time (Boston: Beacon Press, 1944Google Scholar); Moore, B., Social Origins of Democracy and Dictatorship: Lord and Peasant in the Making of the Modern World (Harmondsworth: Penguin, 1966Google Scholar).

⁴⁵ For arguments about the nature and peculiarities of an “African mode of production,” see Coquery-Vidrovitch, C., “Research on an African Mode of Production,” in Gutkind, P. and Waterman, P., eds., African Social Studies: A Radical Reader (New York: Monthly Review Press, 1978Google Scholar); Bates, R., Essays on the Political Economy of Rural Africa (Cambridge, UK: Cambridge University Press, 1983CrossRefGoogle Scholar), chs. 1–2.

⁴⁶ See, for instance, Ranger, T., “Tradition and Travesty: Chiefs and the Administration in Makoni District, Zimbabwe, 1960–1980,” Africa, 52:3 (1982), 20–41Google Scholar; L. Vail, Creation of Tribalism; Mann and Roberts, Law in Colonial Africa. For a strict structuralist argument about the inevitability of such transformation, see Wolpe, H., “Capitalism and Cheap Labour Power in South Africa: From Segregation to Apartheid,” Economy and Society, 1:2 (1972Google Scholar).

⁴⁷ For particularly good accounts that draw on the moral economy perspective, see Beinart, W. and Bundy, C., Hidden Struggles in Rural South Africa. Politics and Popular Movements in the Transkei and Eastern Cape 1890–1930 (Johannesburg: Ravan Press, 1987)Google Scholar and Ranger, T., Peasant Consciousness and Guerilla War in Zimbabwe. A Comparative Study (London: James Currey, 1985)Google Scholar. For a particularly good discussion from a rational choice perspective, see Bates, Beyond the Miracle of the Market.

⁴⁸ See, for instance, Bundy, C., The Rise and Fall of the South African Peasantry (Berkeley: University of California Press, 1979)Google Scholar; Palmer, R. H. and Parsons, N., eds., The Roots of Rural Poverty in Central and Southern Africa (London: Heinemann, 1977)Google Scholar; Palmer, R. H., Land and Racial Domination in Rhodesia (London: Heinemann, 1977)Google Scholar. See also Hart, K., The Political Economy of West African Agriculture (Cambridge: Cambridge University Press, 1982)CrossRefGoogle Scholar and Bates, Markets and States in Tropical Africa.

⁴⁹ On the readiness of subordinate groups in the community to expand their own options under changing economic conditions, see Marks, S. and Atmore, A., Economy and Society in Pre-Industrial South Africa (London: Longman, 1980)Google Scholar.

⁵⁰ The term partiality refers to more than simply the political base of the state. It entails also the consistent bias of the state towards particular social groups. See Greenberg, Legitimating the Illegitimate, ch. 2.

⁵¹ See Saul, J., “The State in Post-Colonial Societies: Tanzania,” Socialist Register (1974), 352–67Google Scholar; Higgot, “The State in Africa,” 23–32.

⁵² See, for instance, G. Williams, “Taking the Part of Peasants: Rural Development in Nigeria and Tanzania,” in , P. C. W.Gutkind, and Wallerstein, I., eds., Political Economy of Contemporary Africa (Beverly Hills, CA: Sage, 1976)Google Scholar; Bates, Markets and States in Tropical Africa, chs. 5–7; Leys, “Capital Accumulation, Class Formation, and Dependency,” 241–66.

⁵³ T. Callaghy, “Politics and Vision in Africa: The Interplay of Domination, Equality and Liberty,” in Chabal, ed., Political Domination in Africa, 31–32.

⁵⁴ See Barkan, J. and Okumu, J., eds., Politics and Public Policy in Kenya and Tanzania (New York: Praeger, 1979)Google Scholar; Tordoff, W., Politics in Zambia (Manchester: Manchester University Press, 1974)Google Scholar; Callaghy, T., The State-Society Struggle: Zaire in Comparative Perspective (New York: Columbia University Press, 1984)Google Scholar.

⁵⁵ I am grateful to Jane Guyer and Subir Sinha for discussions on this point. It is of course also important to note that peasants’ market behaviour is sometimes inseparable from intra-household struggles over control of the production process. Where negotiations over market and state are inseparable, it is arguable that they are in part also negotiations over the supply and definition of public goods.

⁵⁶ See, for instance, Phillips, The Enigma of Colonialism, 161–3.

⁵⁷ Lonsdale, “States and Social Processes,” 195.

⁵⁸ It is for this reason that one cannot explain the peasantry or its prospects by analysis at the point of production alone; cf. Cooper, “Peasants, Capitalists, and Historians,” 284–314. The point of exchange is also of central importance, for it is here that the state seeks to regulate the productive life of peasant communities and to integrate them into the broader imperatives of control, domination, and development at the level of the social order.

⁵⁹ See, for instance, R. Otayek, “Burkina Faso: Between Feeble State and Total State, the Swing Continues,” in D. B. C. O'Brien, J. Dunn, R. Rathbone, eds., Contemporary West African States, 13–30; Bayart, “Civil Society in Africa,” 109–25; Dunn, J. and Robertson, A. F., Dependence and Opportunity: Political Change in Brong Ahafo (Cambridge: Cambridge University Press, 1973)Google Scholar; Sklar, “The Nature of Class Domination in Africa,” 531–52.

⁶⁰ While political domination never goes uncontested, such contestation is not necessarily revolutionary. As Scott, J. (in Weapons of the Weak: The Everyday Forms of Peasant Resistance [New Haven: Yale University Press, 1985], 255)Google Scholar has argued with respect to forms of “routine resistance” among peasant communities: “Very little of this activity … poses a fundamental threat to the basic structure of agrarian inequalities, either materially or symbolically. What it does represent, however, is a constant process of testing and renegotiation of productive relations between classes.”

⁶¹ See, for instance, Mandala, E., Work and Control in a Peasant Economy (Madison: The University of Wisconsin Press, 1990)Google Scholar; Carney, J. and Watts, M., “Manufacturing Dissent: Work, Gender and the Politics of Meaning in a Peasant Society,” Africa, 60:2 (1990), 207–41CrossRefGoogle Scholar.

⁶² This is particularly difficult where political allegiances are heavily influenced by ethnic considerations. Indeed, ethnicity is often mobilised as a form of resistance to the inculcation of this authority.

⁶³ Plamenatz, J., “Two Types of Nationalism,” in E. Kamenka, ed., Nationalism: The Nature and Evolution of an Idea (Canberra: Australian National University Press, 1976), especially 34–36Google Scholar.

⁶⁴ The most recent indicator of this acceptance is, perhaps, the statement by Daniel arap Moi of Kenya that Kenya is “two hundred years behind the west.” Uganda's Yoweri Museveni has made similar pronouncements.

⁶⁵ For a theoretical discussion of organic statism, see Callaghy, The State-Society Struggle, 14–17. On African populism, see Arrighi, G. and Saul, J., Essays on the Political Economy of Africa (New York: Monthly Review Press, 1973)Google Scholar; Kitching, Development and Underdevelopment. On one-party stateism, see A. Zolberg, Creating Political Order.

⁶⁶ See, for instance, Joseph, “Class, State, and Prebendal Politics,” 21–38; Berry, Fathers Work for Their Sons; Hayward, F., ed., Elections in Independent Africa (Boulder, CO: Westview Press, 1988)Google Scholar.

⁶⁷ In some cases, decentralisation was also a strategy for managing potential political threats from rival power groups. See, for instance, Barkan, J. and Chege, M., “Decentralizing the State: District Focus and the Politics of Reallocation in Kenya,” Journal of Modern African Studies (1989)CrossRefGoogle Scholar, and generally Wunsch, J. and Olowu, D., eds., The Failure of the Centralized State: Institutions and Self-Governance in Africa (Boulder, CO: Westview Press, 1990)Google Scholar.

⁶⁸ Oyatek, “Burkina Faso: Between Feeble State and Total State,” 14.

⁶⁹ A. Odhiambo, “Democracy and the Ideology of Order in Kenya,” in W. Oyugi, A. Odhiambo,Chege, M., and Gitonga, A., Democratic Theory and Practice in Africa (London: James Currey, 1988), 116–7Google Scholar. See also Zolberg, Creating Political Order, 75.

⁷⁰ Zolberg (in Creating Political Order, 63–64) stresses the admiration of African leaders for planning and “rational control of the economy.” In the post-colonial context, however, planning, funding, and the control of knowledge were closely related. Post-independence development required planing and funding—funding frequently depended on planning, which is a technical process heavily influenced by ex-colonial or imperial technocrats. Consequently, the terms of the discourse were already significantly entrenched and strengthened by the ways in which money was tied to the acceptability of plans to the funding institutions. Thus, foreign interests continued to influence significantly the realm of authoritative knowledge. Even where control of development discourses was contested by nationalist governments, such control did not include the targets of policy.

⁷¹ The problem was exacerbated by the fact that African Governments did not have full control over their constitutions at independence and by the inertia generated by the need to keep the system operating. Also, powerful groups from the pre-independence period retained the capacity to put pressure on state functionaries.

⁷² As Corpierre has pointed out, every deployment of ethnic identities manipulates and changes those identities—as soon as the ethnic party attains state power, it loses (at least in part) its ethnic character. This is dramatically demonstrated in South Africa by the rise of Afrikanerdom but also most recently by the politics of Inkatha and in Kenya by the Moi government's manipulation of ethnicity for political and ideological purposes. In this sense, the invention of traditions reflects an effort to transform tribalism in order to structure citizenship. See Vail, The Creation of Tribalism, 2–16 and passim.

⁷³ Hyden, G., Beyond Ujamaa in Tanzania: Underdevelopment and an Uncaptured Peasantry (Berkeley: University of California Press, 1980)Google Scholar; Hirschmann, A. (in Exit, Voice, and Loyalty: Responses to Decline in Firms, Organizations, and States [Cambridge: Harvard University Press, 1970], 21–25)Google Scholar defines the “exit option” as an analytical concept.

⁷⁴ See Bates, Markets and States in Tropical Africa, 8 and passim.

⁷⁵ Kitching, G., Class and Economic Change in Kenya. The Making of an African Petite-Bourgeoisie (New Haven: Yale University Press, 1980)Google Scholar; Hill, P., The Migrant Cocoa Farmers of Southern Ghana: A Study in Rural Capitalism (Cambridge: Cambridge University Press, 1970)Google Scholar; Palmer and Parsons, Roots of Rural Poverty, especially pt. 2.

⁷⁶ See, for example, Morris, M., “The Development of Capitalism in South African Agriculture,” Economy and Society, 5:3 (1976), 292–363CrossRefGoogle Scholar; Bernstein, H., “Notes on Capital and Peasantry,” Review of African Political Economy, no. 10 (1977), 60–73CrossRefGoogle Scholar; Cooper, “Peasants, Capitalists, and Historians,” 284–314. Degrees of incorporation do vary, but where peasants do not produce for, and sell in, the local market it is frequently because their productive base is too small. Often peasant communities that remain altogether outside of the produce market are dependent, therefore, upon the labour market in the capitalist sector for their viability and survival. See, for example,Lan, D., Guns and Rain: Guerrillas and Spirit Mediums in Zimbabwe (Harare: Zimbabwe Publishing House, 1985)Google Scholar, ch. 2.

⁷⁷ See, for instance, Shipton, P., Bitter Money: The Classification of Forbidden Commodities among the Luo of Kenya (Washington: American Ethnological Society, 1989)Google Scholar; Feierman, S., Peasant Intellectuals. Anthropology and History in Tanzania (Madison, WI: The University of Wisconsin Press, 1990)Google Scholar.

⁷⁸ Lonsdale, “States and Social Processes,” 179.

⁷⁹ Cooper has argued (in “Peasants, Capitalists, and Historians,” 314) rightly that, to understand the actions of peasants over time, they must be regarded in the light of a historical process of transforming relations of power at the point of peasant production. However, while concentrating analysis on the control of labour at the immediate point of production may elucidate the immediate nature of peasant struggles and strategies, it presents them as essentially reactive to outside pressures and neglects that these struggles themselves shape the demands made on peasant communities. This is particularly well illustrated by Carney and Watts, "Manufacturing Dissent,” 207–41. See also Scott, Weapons of the Weak.

⁸⁰ A good example is provided by government policies towards rural women. In the early twentieth century, rural administrators in Zimbabwe were able to establish political alliances with local patriarchs who feared that new social and economic opportunities associated with colonialism had loosened their control over women. In later years, the state targeted women as a constituency for building rural stability and for stimulating economic activity. Women participated in state initiatives for their own reasons. After independence, the government pursued a strategy of rural integration which stressed (and linked) women's rights and women's productivity.

⁸¹ Colonial history is well marked by miscalculations of colonial administrators and policy makers of peasant objectives, partly because these objectives were often generated in reaction to colonial policies. Utilitarian arguments for stringent individual maximation as the root of peasant decisions are also vulnerable to such miscalculation. See, for instance, Popkin, S., The Rational Peasant (Berkeley: University of California Press, 1979), 2Google Scholar.

⁸² A striking example is the discussion of South Africa in Greenberg, Legitimating the Illegitimate, especially chs. 5–7.

⁸³ See Gramsci, Selections from the Prison Notebooks, 168.

⁸⁴ As Bourdieu points out (inOutline of a Theory of Practice [Cambridge: Cambridge University Press, 1977], 29–30)Google Scholar, although social actions occur according to social rules because they are only intelligible to the participants within the context of those rules, they cannot be determined by those rules because there are invariably more ways than one to respond to a proposition.

⁸⁵ Tilly, C., From Mobilization to Revolution (Reading, MA: Addison-Wesley, 1978), 133–8Google Scholar.

⁸⁶ A consciousness of resistance within communities is, of course, not necessarily a class consciousness in the sense of creating self-conscious class cohesion. Marx, of course, was very scathing about the potential of peasants for class cohesion, describing them in The Eighteenth Brumaire as having unity in the same way as a sack of potatoes. The potential disjuncture between resistance and class cohesion is well illustrated in C. van Onselen, “Worker Consciousness in Black Miners: Southern Rhodesia, 1900–1920,” in P. Cohen, Gutkind, P. C. W., and Brazier, P., eds., Peasants and Proletarians: The Struggles of Third World Workers (New York: Monthly Review Press, 1979)Google Scholar. Van Onselen argues that workers resort to desertion not instead of, but as a form of, combination. Although this reflects a strategy of negotiation open to workers, it does not necessarily reflect a proletarian class consciousness. See also van Onselen, C., Chibaro: African Mine Labour in Southern Rhodesia, 1900–1933 (London: Pluto, 1976), 243–4Google Scholar for a sensitivity to this point.

⁸⁷ In African countries, where the ideology of development initiatives tends, partly for historical reasons, to be highly technocratic and directive, if not directly coercive, peasants frequently treat with considerable suspicion advantages that are obvious to development workers and agencies. Why technocratic language persists in rural policy is, therefore, an important political question and one that is eliciting increasing interest among scholars. See, for example, Drinkwater, M., “Technical Development and Peasant Impoverishment: Land Use Policy in Zimbabwe's Midlands Province,” Journal of Southern African Studies, 15:2 (1989), 287–305CrossRefGoogle Scholar.

⁸⁸ Bourdieu, Outline of a Theory of Practice, 29–30.

⁸⁹ Greenberg, Legitimating the Illegitimate, 149. It is for this reason that the mediating structures of civil society are so important, for this panoply of social rules, relations, and institutions provides the material and conceptual framework according to which people's lives are ordered and a sense of limits established.

⁹⁰ It is partly because the path of process is not determinate that appeals to the innate nature of a class or to the imperatives of history in the structural development of a mode of production simpliciter do not explain political outcomes. Theories of modes of production which are “in contestation” or “in articulation” are particularly susceptible to this criticism. Such modes of articulation or contestion are themselves both outcomes and processes of actual political struggle. For a good illustration, see Legassick's argument (in “South Africa: Capital Accumulation and Violence,” Economy and Society, 3:3 [1974], 253–91Google Scholar) about the creation of African reserves in early twentieth-century South Africa.

⁹¹ Tordoff, Politics in Zambia; Barkan and Okumu, Politics and Public Policy.

⁹² Here I take as a starting point Alasdair Maclntyre's argument that people and communities understand and constitute themselves as parts of socially local histories and traditions (in After Virtue. A Study in Moral Theory [Dame, IN: University of Notre Dame Press, 1971], 119, 122, 142–53)Google Scholar. I do not regard such traditions as fully determining as Maclntyre seems to. Maclntyre develops a concept of narative histories in which people live out their lives. These histories also provide a kind of “causal history” of their actions (ch. 15). This is the argument that I wish to adapt. However, I make two important distinctions. First, in using the phrase “language of memory,” I do not take the position that social actions are similar to or reducible to linguistic actions. The struggles that I describe here are material struggles—neither ideologies nor conventions overdetermine them. Second, I distance myself from the view that narrative communities in which people are embedded provide the only route to virtue. As I. Shapiro points out (in “Realism in the Study of the History of Ideas,” History of Political Thought, 3:3 [1982], 554)Google Scholar “political languages are embedded in the real world and instrumental in its reproduction.”

⁹³ Anderson, B., Imagined Communities. Reflections on the Origins and Spread of Nationalism (London: Verso, 1989)Google Scholar.

⁹⁴ For an evocative scenario whereby such a process might take place, see Skinner, Q., “Some Problems in the Analysis of Political Thought and Action,” Political Theory, 2:3 (1974), 294Google Scholar.

⁹⁵ In some cases, such as Senegal, religious affiliation has played an important role. For the changing relationship of Senegalese peasants to the Marabouts, see C. Coulon and D. B. C. O'Brien, “Senegal,” in D. B. C. O'Brien etal., eds., Contemporary West African States, 152–63.

⁹⁶ Herbst, J., State Politics in Zimbabwe (Harare: University of Zimbabwe Publications, 1990), 261Google Scholar and ch. 11 generally.

⁹⁷ J. Dunn, “Social Theory, Social Action and Political Theory,” in Dunn, J., Rethinking Modern Political Theory (Cambridge: Cambridge University Press, 1985), 137–8Google Scholar.

⁹⁸ Among modernization theorists, see especially Huntington, S., Political Order in Changing Societies (New Haven: Yale University Press, 1968)Google Scholar, chs. 3 and 6. On Marxist-Leninist development paths, see Young's argument (in Ideology and Development, ch. 2) that “the ultimate verdict on the Afro-Marxist pathway” would probably depend on the political and economic development of Mozambique. By 1990, the FRELIMO government in Mozambique, teetering on the brink of political and economic collapse, had disavowed Marxism-Leninism and the one-party state and was struggling to rebuild popular support. For the Zimbabwe state, watching its neighbour and close ideological friend closely, a clear danger was signalled: Connections between the state and rural society might become increasingly tenuous over time, and the political effects might be devastating.

⁹⁹ In this respect Murphree's reference (in “Communities as Institutions for Resource Management” [Paper presented to the National Conference on Environment and Development, Maputo, 1991]) to communities as institutions is provocative.