¹ ‘The Modern Prince’, Selections from the Prison Notebooks, edited and translated by Hoare, Quinton and Smith, Geoffrey N. (New York, 1971), 200.Google Scholar Many thanks to Joseph C. Miller for his rigorous critique of an early draft. Thanks also to David Andersonid, Steve Feierman, Karen T. Hansen, James Kern, Pier Larson, James Oakes, Steve J. Stern, Jan Vansina, and Ivor Wilks. Research was supported in part by a Fulbright Grant and by the Social Science Research Council and the American Council of Learned Societies.

² Hobsbawm, E. J., Primitive Rebels (New York, 1959), 2.Google Scholar

³ Thompson, E. P., ‘Eighteenth-century English society: class struggle without class?’ Journal of Social History, III (1978), 133–65.CrossRefGoogle Scholar

⁴ For plebeians: Thompson, E. P., ‘The moral economy of the English crowd in the eighteenth century’, Past and Present, l (1971), 76–136CrossRefGoogle Scholar, and ‘Patrician society, plebeian culture’, Journal of Social History, VII (1974), 382–405.Google Scholar Some agrarian examples: Womack, John, Zapata and the Mexican Revolution (New York, 1968)Google Scholar; Brenner, Robert, ‘Agrarian class structure and economic development in pre-industrial Europe’, Past and Present, LXX (1976), 30–75CrossRefGoogle Scholar; Mallon, Florencia, The Defense of Community in Peru's Central Highlands (Princeton, 1983)Google Scholar; Kanogo, Tabitha, Squatters and the Roots of Mau Mau (London, 1987).Google Scholar

⁵ In a classic description of this state of affairs Stuart Schwartz showed how sugar planters in colonial Brazil calculated the costs of reproduction via the trans-Atlantic slave trade to be cheaper than allowing their slaves the conditions in which sexual reproduction would be possible. For a concise statement, see ‘Colonial Brazil, c. 1580–c. 1750: plantations and peripheries’, in Bethell, Leslie (ed.), Cambridge History of Latin America (7 vols.) (New York, 1984), ii, 423–53.Google Scholar Some authors, chiefly Claude Meillassoux, have deemed this a ‘slave mode of production’, but because of the many other ways in which slave labor has been employed, such terminology creates more confusion than clarity. I prefer the term ‘plantation slavery’. (Many modern plantation economies using non-slave labor operate in similar manner, with the costs of reproduction being borne in the remote peasant villages which extrude labor migrants.)

⁶ Or dependence on warfare: Meillassoux, Claude, Anthropologie de l'esclavage: le ventre defer et d'argent (Paris, 1986).Google Scholar

⁷ This struggle is most evident in the substantial literature on Caribbean maroon communities, but it is also present in studies of the cultures and communities of those still enslaved. Examples include Craton, Michael, ‘The passion to exist: slave rebellions in the British West Indies’, Journal of Caribbean History, XIII (1980), 1–20Google Scholar; Beckles, Hilary and Watson, Karl, ‘Social protest and labor bargaining: the changing nature of slaves’ responses to plantation life in eighteenth-century Barbados’, Slavery and Abolition, VIII (1987), 272–93CrossRefGoogle Scholar; Beckles, Hilary, ‘From land to sea: runaway Barbados slaves and servants’, and Anthony McFarlane, ‘Cimarrones and palenques: runaways and resistance in colonial Colombia’, both articles in Heumann, Gad (ed.), Out of the House of Bondage: Runaways, Resistance and Marronage in Africa and the New World (London, 1986)Google Scholar; Price, Richard (ed.), Maroon Societies (Baltimore, 1979)Google Scholar; Mintz, Sidney and Price, Richard, An Anthropological Approach to the Afro-American Past: a Caribbean Perspective (Philadelphia, 1976).Google Scholar

⁸ Miers, Suzanne and Kopytoff, Igor, ‘African “slavery” as an institution of marginality’, in Miers, and Kopytoff, (eds.), Slavery in Africa (Madison, 1977), 3–81.Google ScholarFinley, Moses, ‘Slavery’, International Encyclopedia of the Social Sciences (17 vols.) (New York, 1968), xiv, 307–13.Google Scholar

⁹ Miers and Kopytoff are often accused of suggesting the existence of a single ‘African’ ideology of slavery. In fact, they take pains to distance themselves from such an idea; the inverted commas with which they enclose the world ‘slavery’ are meant not to deny that such a thing as slavery ever existed in Africa (another of their critics’ frequent misinterpretations) but to highlight that what is subsumed under that single category in fact corresponded to many different specific social institutions.

¹⁰ On pages 5–6 of Slavery and Social Death (Cambridge, Mass., 1982)Google Scholar, Patterson writes that the slave was denied the ability to construct a meaningful heritage linking himself to those around him. ‘He had a past, to be sure. But a past is not a heritage. Everything has a history, including sticks and stones’. Slaves were ‘not allowed’ to elaborate a heritage or a sense of community. Presumably, then, by denying slaves permission to think about their universe, the master reduced them to little more than sticks or stones. Of course, admits Patterson, slaves managed in spite of this to forge ‘informal’ community structures. ‘A large number of works have demonstrated that slaves…had strong social ties among themselves. The important point, however, is that these relationships were never recognized as legitimate…’. Thus, Patterson seems to accept the central myth of the classic ideology of slavery: only the master's will mattered. Although (unlike the slave masters) he recognizes the sporadic presence of the slaves’ countervailing will, he considers it to have been of no sociological significance. These analytical shortcomings are especially glaring in Patterson's erudite discussion of slave religion, which he maintains was no different than the religion of the masters. Compare his discussion of the religion of the Old South (in chapter 2) with those of Eugene Genovese, Roll Jordan Roll (New York, 1972), and Stuckey, Sterling, Slave Culture (Oxford, 1987), ch. I.Google Scholar

¹¹ This analogy emerges most strikingly in an article Lovejoy wrote for a Festschrift to Aptheker: ‘Fugitive slaves: resistance to slavery in the Sokoto Caliphate’, in Okihiro, Gary (ed.), In Resistance: Studies in African, Caribbean and Afro-American History (Amherst, 1986), 71–95.Google Scholar It can also be seen in some of his other writings: ‘Problems of slave control in the Sokoto Caliphate’, in Lovejoy, (ed.), Africans in Bondage (Madison, 1986), 235–72Google Scholar; ‘Slavery in the Sokoto Caliphate’, in Lovejoy, (ed.), The Ideology of Slavery in Africa (Beverly Hills, 1981), 201–40Google Scholar; and Transformations in Slavery (Cambridge, 1983).Google Scholar

¹² For a recent (and more subtle) description of this process, see Miller, Joseph C., Way of Death (Madison, 1988)Google Scholar, part One. Lovejoy actually writes of the emergence of a ‘slave mode of production’; I avoid this terminology for reasons already mentioned.

¹³ Against Miers and Kopytoff's assertion that the notion of freedom was a distinctly Western one (an argument elaborated by Patterson), Lovejoy is implacable; yet he must rely on European missionary catechists as a source for slaves’ ideas of freedom and on the conditions induced by colonial conquest for the slaves’ most successful attempts at rebellion and escape. See, inter alia, Transformations, 247 (from which the quote is taken), 276–7.

¹⁴ Patterson, by contrast, considers the quest for an understanding of rebellious slave consciousness invalid, even dishonest (Social Death, especially page II). Lovejoy was assisted by the work of many other scholars, perhaps most notably Cooper, Frederick, Plantation Slavery on the East Coast of Africa (New Haven, 1977).Google Scholar

¹⁵ Ironically, this resembles the view adopted by Patterson in an important early article on Jamaica, first published in 1970: ‘Slavery and slave revolts: a sociohistorical analysis of the first Maroon War, 1665–1740’, in Price, (ed.), Maroon Societies, 246–92.Google Scholar Jamaica was a ‘pseudo-society’, wrote Patterson, in which the masters enjoyed no ideological hegemony over their slaves. Hence, Patterson reifies slave class-consciousness, which he characterizes as a ‘single-minded’ hatred for the master and all his institutions. There is really no contradiction between this position and that taken in his later comparative study. Both see the hegemonic ideology of slavery as a seamless fabric, an absolute monolith to be either accepted or rejected whole. In the early work he stresses the rejection of the ideology; in the later he stresses its power to dominate. The reification of slave class-consciousness was a common tendency in writing on maroons and slave revolts in the 1970s, as is evident in the Price volume. An important exception was Stuart Schwartz, who described the remarkably ambivalent spectacle of Brazilian maroons bargaining with their masters about the terms under which they would be willing to return to slavery: ‘Resistance and accommodation in eighteenth-century Brazil: the slaves’ view of slavery’, Hispanic American Historical Review, LVII (1977), 69–81.Google Scholar

¹⁶ One should in fairness emphasize that Miers and Kopytoff are by no means guilty of all the functionalist heresies of which they stand accused. On pages 22–4, for example, they explicitly acknowledge that movement along the full length of the ‘slavery-to-kinship continuum’ was not universal and that even in a given society chattel slaves could exist alongside slaves who had become near-full members of their master's kin group. Their error lies in the implication that the norm within African systems of slavery was to traverse the full continuum from outsider to kin; that ‘What gives African “slavery” its particular stamp…is the existence of this “slavery”-to-kinship continuum’ (p. 24). Subsequent pages are given to an elegant elaboration of this process, thus strengthening the impression that it is this, which serves above all to reduce social conflict, that lies at the heart of African institutions of slavery.

¹⁷ Gramscl's understanding of popular consciousness emerges from a broad reading of the Selections from the Prison Notebooks. For the concept of contradictory popular consciousness, the notebook on ‘The Study of Philosophy’ (pages 321ff.) is especially important.

¹⁸ Patterson's elaboration of the theme of the ‘reduction of marginality’ is especially relevant and useful in his comprehensive discussion of Muslim manumission and institutions of freedman clientele, or wala; Social Death, 241–7.

¹⁹ For a description of ‘absorptive’ slavery on the early colonial North American frontier, and its transformation into a rigid plantation system, see Wood, Peter, Black Majority (New York, 1974).Google Scholar

²⁰ The terms used in this paragraph might be compared to those used by James L. Watson, who refers specifically to the ‘open’ and ‘closed’ boundaries of the kinship structures of the non-slave community. Unlike Watson, I reject the notion of any ‘system’ being inherently ‘open’ or ‘closed’. I merely intend to focus attention on social boundaries as an issue of conflict between slaves and masters, the outcome of which helps determine the shape of the system at any given moment. ‘Slavery as an institution: open and closed systems’, in Watson, (ed.), Asian and African Systems of Slavery (Oxford, 1980).Google Scholar

²¹ For the Caribbean, see Craton, ‘Passion’; Gaspar, David Barry, ‘Slavery, amelioration, and Sunday markets in Antigua, 1823–1831’, Slavery and Abolition, IX (1988), 1–28.CrossRefGoogle Scholar Similar themes lie at the heart of Genovese, Roll. For the Swahili examples: Glassman, , ‘Social rebellion and Swahili culture: the response to German conquest of the northern Mrima’ (Ph.D. thesis, University of Wisconsin-Madison, 1988)Google Scholar; Lienhardt, Peter, ‘The mosque college of Lamu and its social background’, Tanganyika Notes and Records, LIII (1959), 233–41Google Scholar; el Zein, A. H. M., The Sacred Meadows (Evanston, 1974).Google Scholar

²² C. L. R. James long ago pointed out the direct connections between maroon activity and outright slave rebellion in the Caribbean: The Black Jacobins (1938; reprint New York, 1963).Google Scholar See also Price (ed.), Maroon Societies. For African examples, see Cassanelli, Lee V., ‘Social construction on the Somali frontier: Bantu former slave communities in the nineteenth century’, in Kopytoff, Igor (ed.), The African Frontier (Bloomington, 1987)Google Scholar; Isaacman, Allen, ‘Ex-slaves, transfrontiersmen and the slave trade: the Chikunda of the Zambesi Valley’, in Lovejoy, (ed.), Africans in Bondage, 273–309Google Scholar; Clarence-Smith, W. G., ‘Runaway slaves and social bandits in southern Angola’, in Heumann, (ed.), Out of the House, 23–33Google Scholar; Glassman, , ‘The runaway slave in coastal resistance to Zanzibar: the case of the Witu Sultanate’ (M.A. thesis, University of Wisconsin-Madison, 1983).Google Scholar

²³ The Prison Notebooks contain many passages in which Gramsci examines popular uprisings (mostly agrarian) inspired by contradictory rebellious ideologies such as religious millennialism. See ‘The Modern Prince’, 196–200. For Genovese on hegemony and contradictory consciousness, Roll.

²⁴ This is not to suggest, as Miers and Kopytoff are accused of doing, that slavery was simply an extension of African systems of kinship. There was a qualitative difference between slavery and patriarchal domination within the lineage, but the ideological language used to express these very different social ties had a great deal in common. This point is made forcefully in Meillassoux, Anthropologic, especially pages 9–15. See also Miller, Joseph C., ‘Lineages, ideology, and the history of slavery in western central Africa’, in Lovejoy, (ed.), The Ideology of Slavery in Africa (Beverley Hills, 1981), 40–71Google Scholar; Wright, Marcia, ‘Women in peril’, African Social Research, XX (1975), 800–19Google Scholar; and Lerner, Gerda, ‘Women and slavery’, Slavery and Abolition, IV (1983), 173–98.CrossRefGoogle Scholar Lerner's argument is presented on a scale as universal as Patterson's; as she points out, Patterson consistently fails to recognize the links between ideologies of patriarchy and slavery. In chapter 4 of ‘Social rebellion’, I show that women were as ‘socially dead’ in patriarchal Swahili institutions as were slaves.

²⁵ For the Swahill's closely related neighbors, the Zaramo, see Holmwood, quoted in Hutchinson, Edward, ‘Progress of the Victoria Nyanza Expedition of the Church Missionary Society’, Proceedings of the Royal Geographical Society, XXI (1877), 502.Google Scholar Holmwood's observations came at a time when Pangani sugar slavery was reaching its peak.

²⁶ The Swahili example is best illustrated in Glassman, ‘Social rebellion’. Other examples include Wood, Black Majority; Isaacman, ‘Ex-slaves’; Beckles, Clarence-Smith and McFarlane in Heuman (ed.), Out of the House.

²⁷ Most notably by Scott, James, in whose influential study, The Moral Economy of the Peasant (New Haven, 1976)Google Scholar, the moral economy seems as elegant, monolithic, and incapable of change as do any of Patterson's models of the ideologies of slave societies. Scott's view has changed considerably since then: Weapons of the Weak (New Haven, 1985)Google Scholar.

²⁸ Patterson's extensive discussion of the peculium is in chapter 7 of Social Death. For counter-examples, see especially the articles already cited by McFarlane on colonial Colombia and Gaspar on the struggle of Antiguan slaves to preserve their rights to the Sunday market; also Dirks, Robert, The Black Saturnalia (Gainesville, 1987).Google Scholar (In this last work, the wheat of the consciousness of slave resistance must be winnowed out of the chaff of Dirks’ reductionist ecological analysis.) There is much evidence that Caribbean maroons sought to participate fully in the colonial commercial economy: see for example Sheridan, Richard, ‘The Maroons of Jamaica, 1730–1830: livelihood, demography and health’, in Heuman, (ed.), Out of the House, 152–72Google Scholar; Kopytoff, Barbara, ‘The early political development of Jamaican Maroon societies’, William and Mary Quarterly, XXXV (1978), 291Google Scholar and ‘Jamaican Maroon political organization’, Social and Economic Studies, XXV (1976), 90Google Scholar; Price, Richard, ‘Introduction’, Maroon Societies, 12–13.Google Scholar For East African examples different from that presented here, see Glassman, ‘Runaway slave’, and Isaacman, ‘Ex-slaves’.

²⁹ ‘The Modern Prince’, 200.

³⁰ Sacleux, Charles, Dictionnaire Swahili-Franfais (Paris, 1939–1941), 618Google Scholar; Krapf, Ludwig, A Dictionary of the Swahili Language (London, 1882), 257.Google Scholar

³¹ Cooper, Plantation Slavery; also Glassman, , ‘Social rebellion’, 152–3, 166–70.Google Scholar Much of the following description of Mrima slavery is drawn from ‘Social rebellion’, ch. 3. I here use the Swahili term for ‘slave’ in order to emphasize that it can denote many different categories of subordinate, including personal dependents who would not fulfil most common definitions of ‘slave’. Henceforth it should be assumed that when I use the word ‘slaves’, I am referring in fact to watumwa.

³² This distinction between a defining slave ‘status’ or ‘estate’ (état) on the one hand, and the hugely varied ‘conditions’ which slaves enjoyed or suffered on the other, is derived from Meillassoux, Anthropologic. Meillassoux also emphasizes the contradiction between the absence of a legal recognition of the slave's personhood and the fact that it is the slave's very humanity that makes him valuable. Focussing on this contradiction as the source of struggle is my own elaboration.

³³ We will leave aside those household slaves whose duties were predominantly domestic, sexual and consultative.

³⁴ For shamba and ijara slavery, see Glassman, ‘Social rebellion’, ch. I, and pages 154–7.

³⁵ Fischer, G., ‘Einige Worte über den augenblicklichen Stand der Sklaverei in Ostafrika’, Zeitschrift der Gesellschaft für Erdkunde zu Berlin, XVII (1882), 75Google Scholar; Mackenzie, D., ‘A report on slavery and the slave-trade in Zanzibar, Pemba and the mainland’, 6 May 1895, Anti-Slavery papers, Rhodes House Library, Oxford (Mss Brit. Emp. S.22 G.3), 30–1.Google Scholar

³⁶ ‘Social rebellion’, ch. 2 (on caravan labor), and pages 150–61. Some additional examples of conflict between trading slaves and their masters over issues of credit, etc., can be found in Velten, Carl, Desturi za Wasuaheli (Göttingen, 1903), 264, 315–16Google Scholar; Mathews to Euan-Smith, 5 July 1888, Zanzibar National Archives (henceforth ZNA), AA2/47/364; Fischer, G. A., Mehr Licht im dunklen Weltteil (Hamburg, 1885), 84–5Google Scholar; Toeppen, Kurt, ‘Eigene Beobachtungen und Erkundigungen in den Deutschen Schutzgebieten Ostafrikas’, Deutsche-Kolonial-Zeitung (Berlin) (henceforth DKZ) (1886), 521–2Google Scholar, and ‘Aus Sansibar’, DKZ (1887), 558.Google Scholar

³⁷ Mtoro bin Mwinyi Bakari, The Customs of the Swahili People, edited and translated by J. W. T. Allen (Berkeley, 1981), 172. Mtoro's account (a translation of a portion of Velten, Desturi) was written at the turn of the century, before slavery was abolished on the Mrima. Glassman, ‘Social rebellion’, 161–4.

³⁸ The term is used in the literature on runaway slaves to denote fugitives whose escape from their masters was not intended to be total or permanent. It stands in contrast to grands matrons, who typically constructed and defended autonomous communities beyond the frontiers of the slaveholding society.

³⁹ A few vibarua made an attractive small investment for poorer persons: such investors were frequently slaves themselves, and a slave who so managed to accumulate wealth might take on more than one wife. ‘Social rebellion’, 160–1; Quaas, E., ‘Die Szurl's, die Kull's und die Sclaven in Zanzibar’, Zeitschrift für Allgemeine Erdkunde, neue Folge, IX (Berlin, 1860), 443–4.Google Scholar

⁴⁰ For an incident in the Pangani hinterland in which freeborn women attempted to enlist as porters but were prohibited from doing so by their menfolk, see von Höhnel, Ludwig, Discovery of Lakes Rudolf and Stefanie (2 vols.) (London, 1894), i, 49–51.Google Scholar

⁴¹ Quaas, , ‘Die Szurl's’, 458–9.Google Scholar

⁴² Some female slaves, of course, escaped field labor by becoming concubines. The following sources explain the sexual imbalance of field slaves by referring to males’ employment as porters, mafundi, and the like, while females were left to cultivate: Leue, A., ‘Die Sklaverei in Deutsch-Ostafrika’, DKZ (1901), 238–9Google Scholar; New, Charles, Life, Wanderings and Labours in Eastern Africa (2nd ed., London, 1874), 62–4Google Scholar; Baumann, Oscar, Usambara und seine Nachbargebiete (Berlin, 1891; translation by MrsGodfredsen, M. A., n.d., Pangani Valley Research Project, University of Dar es Salaam), 13–14.Google Scholar

⁴³ The sultan's official, backed by Stanley, prevailed. Bennett, N. R. (ed.), Stanley's Despatches to the New York Herald (Boston, 1970), 190–1.Google Scholar

⁴⁴ Questioned about a female slave of his who had been abused, he replied that she had been flogged ‘not by me but by her husband…, also my slave’ the cause of the flogging was ‘a domestic affair of their own and a trouble similar to my own’. (Emphasis added.) Statement of Diwan Mambo Sasa, 12 March 1877, Public Records Office, London (henceforth PRO), FO 84/1484, 254–5. My speculations rest on what I believe to be a basic facet of Swahili culture. If there were any single ‘fundamental status distinction’ in Swahili society, constant throughout the nineteenth (and twentieth) century, it was not between master and slave, but between male and female. This status distinction was fundamental to all of social organization and cut through all other distinctions of social stratum. See ‘Social rebellion’, 248–78.

⁴⁵ Kirk, 14 July 1877, PRO, FO 84/1485, 327–8; also FO 84/1515, passim (1878); FO 84/1575, 330–1 and passim (1880). For a particularly striking example, Farler, J. P., The Work of Christ in Central Africa (London, 1878), 23–4.Google Scholar

⁴⁶ Glassman, ‘Social rebellion’, ch. 4; slave religion is also discussed on pages 175–85. For the Sufi orders on the Mrima, see also Nimtz, August, Islam and Politics in East Africa (Minneapolis, 1980)Google Scholar, and Martin, B. G., Muslim Brotherhoods in Nineteenth Century Africa (Cambridge, 1976), ch. 6.Google Scholar Similar trends on the northern Swahili coast are discussed in Lienhardt, ‘Mosque college’, and el Zein, Sacred Meadows.

⁴⁷ ‘Social rebellion’, ch. 5.

⁴⁸ Plantation Slavery, 219, 223–6; for a critique and counter-examples, see Glassman, ‘Social rebellion’, 182 (footnote), and chs. 4 and 5.

⁴⁹ This error is more pronounced in the generalizations which Lovejoy draws from Cooper's work in Transformations. For Lovejoy, the only choice for rebellious slaves was between plantation slavery and freedom in the Western sense.

⁵⁰ For examples of the mzalia's disdain for the mshenzi, see Fischer, Mehr Licht, 80–1; Toeppen in DKZ (1887), 557.

⁵¹ Plantation Slavery, 219; see also 263–4. Cooper stresses that descent determined who could and could not be regarded as a mwunguiana, without acknowledging that descent was often a set of ideals to be manipulated and rules to be broken. In ‘Social Rebellion’, ch. 4, I show how these rules were regularly broken in ways which allowed outsiders to marry into the freeborn community and permitted men of low status to marry up. Such marriages, as well as the construction of fictive kinship, assisted persons of slave descent to escape the stigma of their birth.

⁵² For this usage of mwungwana, see ‘Social rebellion’, ch. 2. For mswahili, see Schmidt, A. R., ‘Deutsch-Witu-Land’, Globus, LIV (Braunschweig, 1888), 133Google Scholar; Beech, M. W. H., ‘Slavery on the east coast of Africa’, Journal of the African Society, XV (1916), 146Google Scholar; Thomson, Joseph, To the Central African Lakes and Back (2 vols.) (London, 1881), i, 91–2.Google Scholar These and other sources are discussed in ‘Social rebellion’, 172–3.

⁵³ Absorption into the kin groups of the freeborn was neither impossible nor insignificant and is discussed in ‘Social rebellion’, ch. 4. Informants on the Swahili coast, both in Kenya and Tanzania, commonly refer to the institutions of slavery as essential to the process by which Africans of upcountry origin have become Swahili. For general descriptions of this process, see Blais, J., ‘Les anciens esclaves à Zanzibar’, Anthropos, X/XI (1915–1916), 504–11Google Scholar; also the many essays by James de Vere Allen, including ‘Town and country in Swahili culture’, in German UNESCO Commission, Symposium Leo Frobenius (Cologne, 1974)Google Scholar; ‘The Swahili world of Mtoro bin Mwinyi Bakari’, in Mtoro, Customs. The process of slaves and other outsiders ‘becoming Swahili’ is a basic theme in the analysis of nineteenth century Swahili culture presented in Glassman, ‘Social rebellion’.

⁵⁴ Additionally: Hopkins, A. G., An Economic History of West Africa (New York, 1972)Google Scholar; Sheriff, Slaves, Spices.

⁵⁵ Cooper, , Plantation Slavery, 79Google Scholar and passim.

⁵⁶ Plantation Slavery, 98–9.

⁵⁷ They acted as a class, in opposition to other local social groups, during the severe political crises which gripped Pangani in the course of German conquest in the late 1880s and early 1890s. See ‘Social rebellion’, ch. 7.

⁵⁸ The Shirazi chief Mambosasa, the only Swahili-speaking sugar planter of whom we have knowledge, is a cultural exception who proves this rule of class excellently: in the political crises of the 1880s and 90s, Mambosasa sided with the Omani planter class against his fellow Shirazi patricians.

⁵⁹ Tellingly, Omani planters are remembered for having refused to teach religion to their slaves, in contradiction of the teachings of the Prophet. Thus, the Ibadhi sect remained restricted to the small Omani caste, and most slaves subscribed to the Sunni rites of the coastal majority.

⁶⁰ The plantation sector dominated Panganl's local economy as it did at no other point on the Mrima. One of the main causes for this difference lay in the structure of the long-distance trade to the interior from the various Mrima towns: whereas Arab settlers at Bagamoyo and Dar es Salaam were interested chiefly in the caravan trade, those at Pangani were excluded from it. See the discussion in ‘Social rebellion’, especially 194–5, also Tables 1.3, 1.4 and 3.2. Important sources for Mrima trade figures include ZNA, AL2/107/255–56; PRO, FO 84/1415/266; Kaerger, Karl, Tangaland und die Kolonisation Deutsch-Ostafrikas (Berlin, 1892).Google Scholar

⁶¹ Oral testimony, 1985–6: Akida Karimu, Mkwaja; Bomu Juma Kirimo, Saadani (Bagamoyo District); Hatibu Salim, Mwera; Chande Maftaha, Saadani; Abdallah Mahine, Bweni (Pangani District); Hashim Abdallah, Bweni; Juma Omari, Mkwaja. Baumann also contrasts Mauya slavery with that existing elsewhere on the Mrima: Usambara, 34.

⁶² For the economic history of Pangani sugar production, see ‘Social rebellion’, 185 ff. Important sources include Kaerger, Tangaland, 18–23; idem., ‘Das Zuckersyndikat’, DKZ (1896), 245; Meinecke, Gustav, Aus dem Lande der Suaheli (Berlin, 1895), 56–8, 111–13, 119–24Google Scholar; and oral testimony, 1985–6, as cited above, plus Ali Waziri, Pangani; Ndembo Maburuki, Pangani. For Caribbean sugar, see Schwartz, ‘Colonial Brazil’; Knight, Franklin W., Slave Society in Cuba during the Nineteenth Century (Madison, 1970)Google Scholar; and the concise survey and annotated bibliography in Klein, Herbert S., African Slavery in Latin America and the Caribbean (Oxford, 1986).Google Scholar

⁶³ There were precise historical reasons that these three towns shared Panganl's intensive version of shamba slavery. Dar es Salaam's plantation economy was dominated by the holdings of the Sultan of Zanzibar, and hence the coercive powers of the Zanzibari state were present there to an unusual degree; Malindi was the site of grain plantations intensively developed by a planter class as exclusive as that of Pangani (see Cooper, Plantation Slavery); and Kilwa was the great entrepôt of the East and Central African slave trade. For shamba slavery at Mauya: Oral testimony, 1985–6: Ali Waziri, Pangani; Hashim Abdallah, Bweni (Pangani); Hatibu Salim, Mwera; Idi Mwinyikombo, Mwembeni (Pangani); Juma Omari, Mkwaja; Ndembo Maburuki, Pangani. See also Weidner, Fritz, Die Haussklaverei in Ostafrika (Jena, 1915), 18–19Google Scholar; Baker, E. C., Report on Social and Economic Conditions in the Tanga Province (Dar es Salaam, 1934), 36.Google Scholar

⁶⁴ See the highly suggestive comments contrasting landlords employing the shamba system with those employing ijara, in Beech, ‘Slavery’, 147–8.

⁶⁵ It should be noted, however, that few of the captives resulting from the kiva were absorbed directly by Mauya planters: they would be re-exported to Pemba or to the Kenya coast, thus replacing the imports from Kilwa who, having come from a more distant homeland, would be preferred for local use. Glassman, ‘Social rebellion’, 202–4. See also Feierman, Steven, The Shambaa Kingdom (Madison, 1974)Google Scholar; Curtin, Philip, Feierman, S. et al. , African History (Boston, 1978), ch. 13.Google Scholar

⁶⁶ Oral testimony for this and the previous four paragraphs includes (1985–6): Akida Karimu, Mkwaja; Abdallah Mahine, Bweni; Hashim Abdallah, Bweni; Juma Omari, Mkwaja; Heri bin Isa and Nusura bin Isa, Makorora (Mkwaja); Juma Omari, Pangani. Also: Baumann, Usambara, 34; von Behr, H. F., Kriegsbilder aus dem Araberaufstand in Deutsch-Ostajrika (Leipzig, 1891), 216.Google Scholar Meinecke's proposals for the establishment of a European sugar mill at Mauya include estimated wages for child laborers, presumably vibarua to be hired from local slave-owners: Aus dem Lande, 142.

⁶⁷ ‘Social rebellion’, chs. 1 and 2.

⁶⁸ This desire to participate more fully in the colonial economy of the Caribbean began even before the maroons had fled from the plantations: a recurrent issue of slave resistance was the right to participate in market festivities, at which slaves would sell the surplus of their provision grounds. See sources cited in note 28, above.

⁶⁹ For East and Central African examples other than the Swahili coast, see Cassanelli, ‘Social construction’; Isaacman, , ‘Ex-slaves’, and The Tradition of Resistance in Mozambique (Berkeley, 1976)Google Scholar; Clarence-Smith, ‘Runaway slaves’.

⁷⁰ Glassman, ‘Runaway slaves’; Travers to Krauel, 25 Aug. 1885, Zentrale Staatsarchiv, Potsdam, Germany (henceforth ZStA), RKA 394, 20–1; Fitzgerald, W. W. A., Travels in the Coastlands of British East Africa (London, 1898), 121–4.Google Scholar

⁷¹ Burton, Richard, The Lake Regions of Central Africa (New York, 1860), 86, 519CrossRefGoogle Scholar, and Zanzibar, City, Island and Coast (2 vols.) (London, 1872), ii, 150–1Google Scholar; Krapf, Church Missionary Society Archives (microfilm, Memorial Library, University of Wisconsin-Madison: henceforth CMS), CA5/016/173 (15 July 1848), CA5/016/177 (29 March 1852) and CA5/016/179; New, Life, 49; Guillain, C., Documents sur l'histoire, la géographie et le commerce de l'Afrique Orientate (3 vols., in 2) (Paris, 1856–1858), iii, 263–4Google Scholar; Behr, Kriegsbilder, 104, 132, 172–3. For both this and the following paragraph, see Glassman, , ‘Social rebellion’, 220–3.Google Scholar

⁷² For a particularly revealing example, see Fischer, , Mehr Licht, 59–61.Google Scholar Père Etienne Baur of the Catholic missions of Bagamoyo and Mandera wrote that those who wished to become mission adherents (frequently single mothers seeking protection and hopefully a powerful husband) would often open with the entreaty, ‘Blanc, achète-moi!’ His colleague, Alexandre Le Roy, describes labor relations at the Bagamoyo mission that were virtually the same as shamba slavery. Baur, and Le Roy, , A Trovers le Zanguebar (Tours, 1886), 107, 341–2.Google Scholar

⁷³ See the accounts of the conflict between Farler and the village of Makumba, in files concerning the Magila, Umba and Mkuzi missions, in Universities’ Mission to Central Africa archives (microfilm, Center for Research Libraries, Chicago: henceforth UMCA), especially Boxes AI(iv)A, AI(v)B, and AI(vi)a; also PRO, FO 84/1599, 11–42, 74–81. For the more explosive and hence better-known scandals at the CMS missions, see Bennett, Norman, Arab versus European (New York, 1986), 68–75, 95–8, 177–80Google Scholar; Morton, Rodger F., ‘Slaves, fugitives, and freedmen on the Kenya coast’ (Ph.D. thesis, Syracuse University, 1976)Google Scholar; and Githige, R. M., ‘The issue of slavery: relations between the CMS and the state on the East African coast prior to 1895’, Journal of Religion in Africa, XVI (1986), 209–25.Google Scholar For the missionaries as ‘big men’ on the African model, see also Giblin, James, ‘Famine, authority, and the impact of foreign capital in Handeni District, Tanzania’ (Ph.D. thesis, University of Wisconsin-Madison, 1986).Google Scholar UMCA missionaries in Malawi found themselves in similar predicaments concerning slavery and patronage: White, Landeg, Magomero (Cambridge, 1987).Google Scholar

⁷⁴ These periods were mid-century near Mombasa, and in the 70s and 80s near Malindi and on the Mrima. Cooper, , Plantation Slavery, 208–9Google Scholar; Iliffe, John, A Modern History of Tanganyika (Cambridge, 1979), 73.CrossRefGoogle Scholar

⁷⁵ ‘Precis of Zanzibar news to 5 July 1873’, enclosure in Kirk to Political Resident, Aden, 4 July 1873, ZNA, AA 1/12/350.

⁷⁶ Kirk to FO, 29 Aug. 1873, PRO, FO 84/1375, 245–9.

⁷⁷ Reichard, Paul, Deutsch-Ostafrika, das Land und seine Bewohner (Leipzig, 1892), 120–1.CrossRefGoogle Scholar Weidner, in fact, categorically denies the existence of Swahili debt slavery (Haussklaverei). There was such a thing as pawnage, but this was most often practiced against non-Muslims of the hinterland; see New, Life, 494–5; ‘Bondei customs’, Pangani District Book, vol. 3, Tanzania National Archives, Dar es Salaam (henceforth TNA), MF 9, 30; Velten, Carl, Safari za Wasuaheli (Göttingen, 1901), 215.Google Scholar

⁷⁸ European observers commonly failed to recognize that autonomous porter-traders were in fact slaves; this was often an act of denial since European explorers motivated by the ideology of abolition would hardly wish to recognize that their very missions depended on slave labor.

⁷⁹ Heri bin Isa and Nusura bin Isa wa Amrani, Makorora, February 1986. There was a dispute between the two informants as to whether the founders of Makorora were in fact slaves or free laborers; Nusura deferred to the authority of the elder Heri on the issue. This dispute itself lends support to my analysis.

⁸⁰ Juma Omari, Mkwaja, February 1986. There is no gender distinction in Swahili personal pronouns; in the translated passage I have chosen the feminine ‘she’ because Mzee Juma's account of the brutality of Mauya slavery, like all the others, stresses the hardships borne by women.

⁸¹ Hobsbawm, Primitive Rebels.

⁸² Glassman, ‘Social rebellion’, chs. 4 and 5.

⁸³ The preceding description is based largely on the testimony of Juma Omari, Mkwaja; Heri Isa and Nusura Isa wa Amrani, Makorora; and Ndembo Maburuki, Pangani. Because there is no formal tradition of oral historiography in coastal culture, the oral testimony is quite scant and must be treated with extreme caution.

⁸⁴ Kirk to FO, 8–10 Dec. 1873, PRO, FO 84/1376, 167–8; oral testimony, as above. For Mohammed bin Hamed, see also FO 84/1390, 23.

⁸⁵ See figures in ‘Social rebellion’, Table 3.1.

⁸⁶ Behr, Kriegsbilder, 216–17.

⁸⁷ It is significant that this unromantic detail, which is confirmed in Kirk's report quoted above, is preserved not in the oral testimony of the rebels’ descendants but in that of the son of their fellow slave-overseer Maburuki, whose refusal to join the rebellion gave rise to enmity between the two families, which has continued for over a century (Ndembo Maburuki, Pangani).

⁸⁸ Juma Omari, Mkwaja.

⁸⁹ Behr, Kriegsbilder, 216–17.

⁹⁰ Juma Omari, Mkwaja; Heri Isa wa Amrani, Makorora. The language of both informants was identical. Mzee Heri claims to have seen the parchment, which was once preserved by his family and may still be in their possession; it is unclear whether it still exists.

⁹¹ Sacleux, Dictionnaire. Krapf, who was perhaps more fully imbued with the abolitionist ideology than was Père Sacleux, notes this usage in his Dictionary but does not specify it as part of his definition as Sacleux does.

⁹² In the case of a slave manumitted by his master, the two remained tied by the bonds of wala, with the slave and his descendants remaining subordinate clients of the master and his. See the discussion of wala in Patterson, Social Death; for the Swahili case, Glassman, ‘Social Rebellion’, ch. 3.

⁹³ The anecdote is widely recalled in the area. The contradictions are all the more striking when one considers that in general TANU was eagerly accepted in this region specifically because it was seen as challenging the aristocrats and Arabs who had hitherto dominated coastal society; if anything, the descendants of slave rebels could have been expected to welcome TANU warmly. See Iliffe, , Modern History, 528–9.Google Scholar For an additional nineteenth-century source on Makorora, see Wissmann, 20 Nov. 1889, ZStA, RKA 743, 32–3. Behr refers to Makorora as ‘Kikokwe’, and subsequent historians, including Iliffe (p. 73). have repeated this error. The village of Kikokwe is located on the southern arm of land forming Pangani Bay, directly opposite Pangani Town, and it is not possible that a maroon community could have been founded there; the village has its own oral history, having nothing to do with watoro. Some of the other contemporary accounts give Makorora's name as ‘Kikora’ and ‘Kikowa’, and it is likely that Behr confused these names with that of Kikokwe, where a German cotton plantation was founded in the years immediately following colonial conquest.

⁹⁴ This is of course more pithy in Swahili: ‘Mshenzi mpe nguo mpya ili asahau kwake’. Taylor, W. E., African Aphorisms, or Saws from Swahili-land (London, 1924), no. 313.Google Scholar

⁹⁵ Taylor, African Aphorisms, no. 351; ‘Shairi la mtumwa’, in Velten, Carl, Prosa und Poesie der Suaheli (Berlin, 1907), 401–3.Google Scholar

⁹⁶ Gramscl's concept of ideological hegemony is often misunderstood to mean absolute ideological domination or ‘false consciousness’; Scott, Weapons of the Weak, is a recent example. See Femia, Joseph, ‘Hegemony and consciousness in the thought of Antonio Gramsci’, Political Studies, XXIII (1975), 32.Google Scholar Conversely, Professor James Oakes points out to me that the common notion of ‘false consciousness’ is a crude distortion of the subtle writings on the topic by Georg Lukács.