¹ Porter, Gina, ‘A note on slavery, seclusion and agrarian change in Northern Nigeria’, J. Afr. Hist., XXX (1989), 487–91CrossRefGoogle Scholar; Fisher, Humphrey J., ‘Slavery and seclusion in Northern Nigeria: a further note’, J. Afr. Hist., XXXII (1991), 123–35.CrossRefGoogle Scholar

² M. G. Smith, ‘Introduction’, in Smith, Mary F., Baba of Karo, a Woman of the Muslim Hausa (New Haven, 1981), 22–3Google Scholar; first published in 1954 by Faber & Faber. The passage is regularly cited in discussions of seclusion; for example, Porter cites this passage at a key moment in her argument, ‘Note’, 489; it is cited by Callaway, Barbara in her discussion of rural kulle, Muslim Hausa Women in Nigeria (Syracuse, 1987), 59Google Scholar, and by Strobel, Margaret in her discussion of the adoption of purdah in East African coastal society, Muslim Women in Mombasa, 1890–1975 (New Haven, 1979), 74Google Scholar.

³ Callaway in fact found that male informants in Kano suggested that newly freed men might have supported their wives' movement into seclusion as a sign of their own status; Muslim Hausa Women, 59.

⁴ For an excellent review of the state of the literature on this subject, see Roberts, R. and Miers, S., ‘The end of slavery in Africa’, in Miers, Suzanne and Roberts, Richard (eds.), The End of Slavery in Africa (Madison, 1988), 3–68Google Scholar.

⁵ This work is based upon oral interviews and archival research conducted in Niger and France and funded through a Fulbright-Hays Doctoral Research Abroad Fellowship in Niger and France in 1988–9. Earlier drafts have been substantially reworked as a result of the remarks and questions of David Geggus, Steve Feierman and Joseph Miller. I am, of course, responsible for any errors which remain.

⁶ For a discussion of the role of these ‘plantations’ in the Caliphate, see Lovejoy, Paul E., ‘Plantations in the economy of the Sokoto Caliphate’, J. Afr. Hist., XIX (1978), 341–68CrossRefGoogle Scholar.

⁷ Lovejoy, , ‘Plantations’, 347Google Scholar; Adamu, Mahdi, The Hausa Factor in West African History (London, 1978), 185–6Google Scholar.

⁸ Informants in Maradi associated slavery with warfare (Hajjiya Rabe, 1 Feb. 1989); although the aristocratic class did have farm slaves (Hajjiya Jeka, 14 Feb. 1989), most captives seem to have been absorbed into the administration. Commoner women were often seized at will by the aristocratic warriors, in their own territories as well as in enemy territories (Hajjiya Indo, 25 Oct. 1989; Fatchima, 5 Feb. 1989).

⁹ Thus Baba remarks that, unlike some other slave raiders, the Maradawa gave up slaves for ransom. Smith, Mary, Baba, 47Google Scholar, see also 38–9, 46–7. Landeroin emphasizes that domestic slaves were better integrated into the household than trade slaves but also notes that prior to the French occupation the court dignitaries who had greatest access to slaves took little interest in agriculture or commerce. Landeroin, Captain, ‘Du Tchad au Niger: notice historique’, Documents scientifiques de la mission Tilho, 1906–1909 (3 vols.) (Paris, 1911), i, 518, 536.Google Scholar Périé noted the high value of slaves on the market, especially women; Périé, Jean, ‘Notes historiques sur la région de Maradi (Niger)’, Bulletin de l'IFAN, 1 (1939), 394–5Google Scholar. Djibo Hamani suggests that in Adar a similar phenomenon obtained: slaves were easily ransomed, and the slave population was concentrated among the Tuareg. Some slaves were to be found in the homes of the aristocracy, but in general slaves made up only a small percentage of the sedentary population. Hamani, Djibo, Contribution à l'étude de l'histoire des états hausa: l' Adar précolonial (République du Niger) (Niamey, 1975), 227–8Google Scholar.

¹⁰ Smith, M. G., ‘A Hausa kingdom: Maradi under Dan Baskore 1854–1875’, in Forde, D. (ed.), West African Kingdoms in the Nineteenth Century (London, 1967), 118Google Scholar; Landeroin, , ‘Notice historique’, 536.Google Scholar

¹¹ I am here echoing Polly Hill's discussion of the range of slave use patterns in Nigeria; see Polly Hill, ‘Comparative West African farm slavery systems’, in Willis, J. R. (ed.), Slaves and Slavery in Muslim Africa (2 vols.) (Ottowa, 1985), ii, 45.Google Scholar Slave use patterns in Maradi probably resembled those in Borno; see Fisher's citation of Nachtigal, , ‘Slavery and seclusion’, 125.Google Scholar

¹² Guy Nicolas reports that among the non-aristocratic farmers of the valley there still exist joking relations between the lineages of former slaves and their former masters; among the ‘dynastic’ urban class the evidence of former slave ownership even among commoners persists in the memory of female ancestors who had been taken into the homes of urban dwellers as concubines. Nicolas, Guy, Dynamique sociale et appréhension du monde au sein d'une société hausa (Paris, 1975), 77, 179, 213.Google Scholar

¹³ Henri Raulin argues that the Zerma used slaves in agriculture to a greater degree than the Hausa of Niger, who were less averse to performing physical labor; Techniques et bases socio-économiques des sociétés rurales du Niger occidental et central (Niamey, 1964), 98–9.Google Scholar André Salifou demonstrates that slaves played a significant role in the administration, military, mining, trade and agriculture of nineteenth-century Damagaram; Le Damagaram ou le Sultanat de Zinder au XIXè siècle (Niamey, 1971).Google Scholar Captain Landeroin also noted that there seemed to be more slaves in the Kanuri area of Damagaram than in the Hausa-speaking areas of the Niger territory; ‘Notice historique’, 522. For a full discussion of slavery in Damagaram, see Roberta Ann Dunbar, ‘Slavery and the evolution of nineteenth-century Damagaram’, in Miers, S. and Kopytoff, I.(eds.), Slavery in Africa: Historical and Anthropological Perspectives (Madison, 1977), 155–77Google Scholar.

¹⁴ Women in the Zerma-speaking west may help at planting but do not grow their own millet; they may produce some condiments or vegetables; Diarra, Fatoumata-Agnes, Femmes africaines en devenir, les femmes zarma du Niger (Paris, 1971), 104Google Scholar. Similarly women in the Hausa-Kanuri region of Damagaram do not hoe but may help at planting and harvest and may have small garden plots; however, Margaret Saunders suggests that women may have hoed in the recent past. Saunders, M., ‘Marriage and divorce in a Muslim Hausa town (Mirria, Niger)’ (Ph.D. thesis, Indiana University, 1978), 7–8Google Scholar.

¹⁵ For detailed discussions of the agricultural technologies in use in Niger see Raulin, , Techniques, 98–9Google Scholar, and Raulin, H., ‘Techniques agraires et instruments aratoires au sud du Sahara’, Cahiers ORSTOM, XX (1984), 339–58Google Scholar. For the Maradi region see Raynaut, Claude, ‘Outils agricoles de la région de Maradi (Niger)’, Cahiers ORSTOM, XX (1984), 505–36Google Scholar.

¹⁶ Mervyn Hiskett, for example, argues from evidence in the mid-nineteenth century and from the poetry of the Fulani reformers that pre-jihād society practiced plantation slavery, despite citing Polly Hill's objections to the generalization that plantation farming was widespread in Hausaland; The Development of Islam in West Africa (London, 1984), 97–100.Google Scholar He uses the Kano Chronicle to argue for the existence of slave villages in pre-jihād Hausaland, but this source on its own is not very compelling evidence as it reflects post-jihād inventions: ‘Enslavement, slavery and attitudes towards the legally enslaveable in Hausa Islamic literature’, in Willis, (ed.), Slaves and Slavery in Muslim Africa, 107.Google Scholar

¹⁷ The word rimji, which is used in Hausa to refer to slave village/plantations, is Fulani in origin, set in contrast to the ruma, or the Fulani master's cattle encampment, as in the proverb: sai ruga ta kwana lafiya, rimji ke kwana lafiya (Only when the master sleeps well does the slave sleep well); see Abraham, R. C., Dictionary of the Hausa Language, (London, 1962), 84, 742Google Scholar. Jan Hogendorn cites a slave song which reconfirms the association of rimji slavery with the Fulani; ‘The economics of slave use on two “plantations” in the Zaria Emirate of the Sokoto Caliphate’, Int. J. Afr. Hist. Studies, X (1977), 382.Google Scholar Polly Hill emphasizes the difference between the ‘Hausa’ form of farm slavery, in which private farmers own and work alongside a small number of slaves, and the more ‘Fulani’ system of large plantations owned by absentee aristocrat and merchant farmers and shows that the vocabulary of the rimji form comes primarily from Fulfulde; ‘Slavery systems’, 36, 40.

¹⁸ Fisher, , ‘Slavery and seclusion’, 132.Google Scholar

¹⁹ Fisher, , ‘Slavery and seclusion’, 132–3.Google Scholar

²⁰ Today in Maradi small rural farmers speak of farming as their ‘heritage’ (gado) and may remark with some pride that all they know is farming, while larger urban farmers see farming as only part of a diversified range of income generating activities. For a comparative discussion of Arna and Katsinawa understandings of farming see Nicolas, , Dynamique sociale, 279–99, 413–18.Google Scholar

²¹ Robertson, C. and Klein, M., ‘Women's importance in African slave systems’, in Robertson, Claire C. and Klein, Martin A. (eds.), Women and Slavery in Africa (Madison, 1983), 3–6Google Scholar. Roberts, and Miers, , ‘End of Slavery’, 21Google Scholar. Female slaves were used in production and could also serve important reproductive functions.

²² Landeroin, , ‘Notice historique’, 518.Google Scholar

²³ Archives d'Outre Mer (Aix-en-Provence, France), microfilm of AOF documents (hereafter designated AOM) 1191 K15 Letter from Lt. Gov. du Soudan Français to the Gouverneur Général de l'AOF, 15 Sept. 1899. The French were not alone in fearing catastrophe if emancipation were to proceed too quickly. Lugard and his successors in Northern Nigeria used legal and tax reforms to eliminate slavery gradually rather than eradicate it immediately; see Hogendorn, Jan and Lovejoy, Paul, ‘The reform of slavery in early colonial Northern Nigeria’, in Miers, and Roberts, (eds.), The End of Slavery in Africa, 391–414Google Scholar, as well as other works in the same volume.

²⁴ AOM 1191 K15 Circular from Délegué Ponty to Commandants de Cercle, 1 Feb. 1901.

²⁵ AOM 1191 K16 ‘Rapport sur l'esclavage: enquête’, from Gouverneur Général de l'AOF to the Ministre des Colonies [1903].

²⁶ AOM 1192 K17 ‘Enquête sur l'esclavage en AOF’, l'Administrateur Poulet, 1905.

²⁷ Rural wives were expected to perform all the arduous tasks of a slave, and issues of control of labor probably centered less on slave versus free status than upon men's control of women as wives. Thus on the eve of colonial rule in the region Landeroin would report the response of a local dignitary to the suggestion that mills would one day reduce the labor of women; ‘La femme qui travaille n'a pas le temps ni le désir de tromper son époux!’; ‘Notice Historique’, 474 (face).

²⁸ See Roberts, Richard, ‘The end of slavery in the French Soudan, 1905–1914’, in Miers, and Roberts, (eds.), The End of Slavery in Africa, 282–307Google Scholar, and Hogendorn, and Lovejoy, , ‘Reform’, 396.Google Scholar

²⁹ The conceptual and functional kinship of marriage and enslavement has been remarked upon by others and arises in part because men of limited means could marry by obtaining a slave as wife and in part because captive women and their offspring were often integrated into households as the wives and children of the owner. See for example, Kopytoff, Igor and Miers, Suzanne, ‘African “slavery” as an institution of marginality’, in Miers, and Kopytoff, (eds.), Slavery in Africa, 65–7.Google Scholar In Borno slaves served as wives to men who lacked the social and monetary means to obtain a full wife; Fisher, , ‘Slavery and seclusion’, 128.Google Scholar

³⁰ Nicolas, Guy, Don rituel et échange marchand dans une société sahelienne (Paris, 1986)Google Scholar.

³¹ For the words baiko and baiwa, see Abraham, , Dictionary, 59–60.Google Scholar

³² Abraham, , Dictionary, 947.Google Scholar

³³ ‘Coutumes Haoussa et Peul (Cercle de Maradi) 1933’, Coutumier Juridique de l'A.O.F. (Paris, 1939), iii, 281–3.Google Scholar

³⁴ ‘Coutumes Haoussa et Peul’, 263.

³⁵ Landeroin, , ‘Notice historique’, 515.Google Scholar

³⁶ Archives Nationales du Niger (hereafter ANN), 14 Jan. 1912, ‘Monographie du Lieutenant Villomé, 1913’.

³⁷ ‘Coutumes Haoussa et Peul’, 292.

³⁸ Complicity between the colonial administration and the male aristocracy concerning female slavery occurred in northern Nigeria as well. See Lovejoy, Paul, ‘Concubinage and the status of women slaves in early colonial Northern Nigeria’, J. Afr. Hist., XXIX (1988), 245–66CrossRefGoogle Scholar.

³⁹ The words for ‘co-wife’ and ‘jealous rivalry’ are the same—ʾkishiya— and Hausa folklore depicts co-wives as irrational and violent and women generally as naturally divisive and cunning. Stories from B. L. Edgar's collection of Hausa Folktales echo these beliefs about women: co-wives regularly attack one another irrationally, often with magic (i, nos. 10, 30); they are a constant source of trouble between men (i, nos. 40, 67, 148); the quality of jealousy in a wife is worse than stealing or unbridled sexuality (i, no. 120). Edgar, B. L., Litafi na Tatsuniyoyi na Hausa (Edinburgh, 1924), vol. i.Google Scholar

⁴⁰ Usman ʾdan Fodio's followers, who had no quarrel with slavery itself, cited this ranking of wives as evidence of the non-Islamic nature of the Hausa kingdoms. Hiskett, Mervyn, ‘Kitab al Farq’, Bulletin of SOAS, XXII (1960), 567, 578.Google Scholar However, circumstances since the jihād have not tended to eliminate it either in northern Nigeria or in Niger.

⁴¹ Hajjiya Jeka, 12 Feb. 1989.

⁴² Where none of my oldest informants mentioned any problems concerning serial polygyny, women married from about 1945 on have often described marriages which ended because their husbands took on new wives and divorced previous wives to remain within the limits prescribed by Islam. In rural areas serial polygamy following the more familiar pattern of successive marriages to only one woman occasionally occurs.

⁴³ ‘Women's Worth and Wedding Gift Exchange’ (paper presented at the Annual Meeting of the African Studies Association in Seattle, Nov. 1992).

⁴⁴ One of the features of life in the new town which particularly struck women informants was the ready availability of water; in the old town good water was to be found only in three wells outside the city. See David, Philippe, Maradi: I'ancien état et l'ancienne ville (Niamey, 1964), 144.Google Scholar

⁴⁵ Nicolas, , Dynamique sociale, 181.Google Scholar

⁴⁶ Saunders, , ‘Women's role’, 70.Google Scholar

⁴⁷ See Callaway, , Muslim Hausa Women, 55–68.Google Scholar