1. This ceremony, described in chapter 6, was the subject of my first publication about Islam among the Dyula (Launay 1977b). Parts of this article have been incorporated into chapter 6, but for the most part I have preferred to reformulate the discussion in somewhat different terms. [BACK]

2. A version of chapter 4 (Launay 1990) is included in that volume. [BACK]

1. For accounts of the history of the idea of "totemism" and of its demise, see Lévi-Strauss 1962 and Kuper 1988. [BACK]

2. Strictly speaking; it is quite conceivable that individuals identify themselves as "totemists." One African friend of mine proudly claimed to be an "animist." However, such cases involve the adoption by individuals, if not whole groups, of labels invented by outsiders; Islam is clearly in another category. [BACK]

3. Such a position is explicitly adopted by Zein 1977; see Asad 1986 and Eickelman 1987 for critiques of it. [BACK]

4. I have been as guilty of this practice as anyone else (Launay 1977b and 1982). By and large, such misuse of language has not been intended as the expression of an explicit theoretical position, but its very ambiguity reflects real conceptual concerns of anthropologists. It should be noted that many "anthropologists of Islam" who have used such phrases in the past have, like myself, publicly or privately recanted. [BACK]

5. I certainly do not wish to imply that "traditional" religions are organic products of local communities, whereas "universal" religions are not. The pitfalls of the organic approach may be as great in the analysis of nonscriptural religions; they are simply less obvious. [BACK]

6. Notable exceptions are Zein 1974 and, most recently, Holy 1991. [BACK]

7. For example, Abner Cohen 1969, Amselle 1977, and Copans 1980, to cite only book-length studies. [BACK]

8. The literature on West African jihad movements and the states they established is voluminous. Monographs on the subject include Last 1967, Hiskett 1973, Quinn 1972, and Robinson 1985. [BACK]

9. Sufis and Sufi orders have also generated a massive corpus of scholarly writings. Again, purely by way of example, one can cite Behrman 1970, O'Brien 1971 and 1975, Coulon 1981, and Copans 1980 on Senegal alone, as well as Martin 1976, Paden 1973, Brenner 1984, and O'Brien and Coulon 1988. [BACK]

10. Such emphasis on these particular themes is already evident in Trimingham's (1962) earlier survey. In many respects, Hiskett's and Clarke's books update Trimingham's information without calling into question the general framework. [BACK]

11. Studies of societies characterized by the Suwarian tradition include Sanneh 1979, Hunter 1976 and 1977, Launay 1982, Ferguson 1973, Levtzion 1968, Wilks 1968 and 1989, Green 1984, and Handloff 1982. [BACK]

12. See, e.g. Beidelman 1982. [BACK]

13. The nature of the Iranian theocratic regime is quite exceptional in Islamic history; see Enayat 1983 and Arjomand 1987. [BACK]

14. I certainly do not wish to imply that all Marxists subscribe to this kind of approach, by any means, though I would argue that it is consistent with the point of view of some Marxists. For a highly sophisticated Marxist approach to religion, one that confronts the problem of religion as "ideology" squarely, see Bloch 1986. [BACK]

1. See Launay 1978 and 1982 for a more detailed description of Dyula trade in Korhogo. [BACK]

2. The issue of children born out of wedlock— nyamogoden— is a complex one. Technically, for instance, a child born to a couple who have eloped, before the marriage has been acknowledged by the woman's family, can be labeled nyamogoden , and could be claimed by the woman's descent groups as a member on these grounds. Of course, not least because the Dyula are Muslims, it is highly insulting to be called a bastard. Whatever "rules" might exist about who actually is a bastard and to which group he or she really belongs are systematically manipulated or violated. See Launay n.d.b. [BACK]

3. See Launay 1979 on the "host/stranger" relationship among the Dyula. [BACK]

4. See Bassett 1985 for a history and evaluation of cotton production in the region. [BACK]

1. See Lovejoy 1980 for a history of kola production and trade in West Africa. [BACK]

2. Binger 1892 provides a detailed, firsthand description of Kong. For modern accounts of Kong's history, see Bernus 1960, Green 1984 and 1986, and Université Nationale de Côte d'Ivoire 1977. [BACK]

3. On the chiefdom of Kadioha, see Launay 1988a and 1988b. [BACK]

4. Bernus 1961, Ouattara 1977. However, Ouattara argues that these oral traditions reflect Dyula influences, and dates both Nanguin's reign and the foundation of Korhogo much earlier. [BACK]

5. Outtara 1977. [BACK]

6. The identity of the Dieli language has been a subject of some controversy; see, e.g., Person 1964: 328 and Glaze 1981: 41. See Launay (n.d.a.) for a critical overview of theories about the Dieli's origin. In any case, their language is unintelligible to any other group in the region. The language is rapidly disappearing; many younger Dieli cannot understand it, much less speak it. [BACK]

7. Louis Binger (1892: 298) noted a similar phenomenon in Kong, whose Muslim population he divided into three categories: (1) literate Muslims, (2) illiterate Muslims who nevertheless followed Koranic precepts reasonably strictly, (3) dolo -(i.e., beer-) drinking Muslims. Binger's beer-drinking Muslims are no doubt analogous to the tun tigi of Korhogo. [BACK]

8. Strictly speaking, the term banmana was not used to refer to all unbelievers, but only to those living in the West African Sahel or the savanna. Forest dwellers such as the Akan or the Guro were called by other names. [BACK]

9. This contrasts sharply with the celebration of Muslim festivities in various kingdoms of northern Ghana, such as Gonja (Goody 1967: 201-2) and Mamprusi (Brown 1975: 95-96, 98), where they were observed by Muslims and non-Muslims alike, and served to express political allegiance to chiefs, rather than membership in the religious community of Islam. [BACK]

10. Cf. Abner Cohen 1971. [BACK]

11. One Dyula informant asserted that the Milaga formerly ate dog flesh, proof positive of their staunch paganism; dog flesh, even more than pork, tends to symbolize forbidden meat ( jufaa ) among the Dyula. Whether or not such an assertion is true, it suggests a means whereby Dyula might symbolically distinguish the "pagan" Milaga from Muslim tun tigi , who, after all, drank beer and offered blood sacrifices to spirits. [BACK]

12. See Goody 1968. [BACK]

13. Ripert, cited in Marty 1922: 152. The story is situated in the Worodugu, a region with which the Korhogo Dyula maintained close contacts, and whose scholars shared with Korhogo the same "Suwarian" tradition (see Wilks 1968: 176-81). [BACK]

14. See Triaud 1974 and Harrison 1988 for discussions of French attitudes and policy toward Islam, and of the context in which such declarations of support were solicited. [BACK]

15. On the notion of "strangers" among the Dyula, see Launay 1979. [BACK]

16. I would tentatively date the introduction of sermons to Koko to the late 1950s or early 1960s. Their introduction is generally attributed to al-Hajj Mustafa "Benkoro" Cisse, who is still active in Koko and is by no means an old man. [BACK]

17. For a fuller discussion of Dyula sermons, see chapter 6. [BACK]

18. In some contexts, the Dieli of Koko continue to stress their separate identity; they have become Dyula for most, but not quite all, intents and purposes. [BACK]

19. See J. and M.-J. Dérive 1986. [BACK]

1. Sanneh 1979 argues for an earlier, thirteenth-century date for al-Hajj Salim Suware. [BACK]

2. The phrase is that of Ivor Wilks (1984), to whom I am heavily indebted for his discussion of al-Hajj Salim and the Suwarian tradition. [BACK]

3. For the career of al-Hajj Mahmud Karantaw, see Levtzion 1968: 147-51, Hiskett 1984: 168-70, and Wilks 1989: 100-103. [BACK]

4. For a history of the "Wahhabi" movement in West Africa, see Kaba 1974 and Amselle 1988; see also Niezen 1990 for a description of a rural "Wahhabi" community in Mali. [BACK]

5. See, e.g., Kaba 1974. Mervyn Hiskett (1984: 290) goes so far as to label them "the religious wing of the Rassemblement Démocratique Africain in West Africa." [BACK]

6. On the role of rote memorization in other traditions of Islamic learning, see Eickelman 1985 and Santerre 1973. [BACK]

7. Reichmuth 1989 describes comparable trends in Islamic education in Nigeria. [BACK]

1. This does not at all imply that ritual only expresses the nature of the moral community. It is senseless to attempt to reduce the "meaning" of ritual to any single dimension. Any analysis, much less any exegesis, is consequently incomplete at best. However, whatever else ritual is "about," it is a social act and thus has social ramifications. The examination of these ramifications, to the exclusion of other possible "meanings,'' is methodologically justifiable. [BACK]

2. Although slavery was abolished in Côte d'Ivoire in 1908, and slaves were allowed to return to their home regions, many individuals remained with their former masters. As late as 1973 I knew a man who was a san jon , a slave who had been captured as a boy and

later was acquired by purchase. The status of worosso , a slave literally "born in the house," that is to say, descended on both sides from slave parents, remains very common. In this, quite special, sense, "slaves" still exist in Korhogo today; see Launay 1977a. [BACK]

3. "Senufo" masks were, in fact, either Tiembara, Fodonon, Fono, or Kule, and so forth. Before the colonial period, masks were unlikely to be conceived as "Senufo" per se, though the Dyula, as a minority, were much more conscious of their ethnic identity as such. [BACK]

4. This was even true of funerals, at least the funerals of elders. The funerals of younger individuals were more uniformly solemn, but they were also far less elaborate. [BACK]

5. Prouteaux 1925 describes the emergence of the lo masks in the old Dyula center of Kong. Lo masks would also dance at the funeral of important elders, mory as well as tun tigi. [BACK]

6. On the chiefdom of Kadioha, see Launay 1988a and 1988b. [BACK]

7. The word saraka has been borrowed by Senufo "unbelievers" to refer to certain sacrifices. However, such offerings would not be defined as saraka by Dyula. On the other hand, there are definitely occasions when "pagans" do offer saraka , notably when following the advice of a Muslim cleric about some personal problem. [BACK]

8. Even nowadays, Dyula may refer to Muslim Senufo as banmana , "pagans." [BACK]

9. Although much attention has been given to the issue of crossing or not crossing arms, it is most clearly the separatism of the Wahhabis that their detractors find most objectionable, and not the posture of prayer per se. One hajji , having observed that most Muslims in Mecca pray with arms crossed, chose on his return to adopt a mode of prayer that, at least superficially, resembled that of the Wahhabis. His behavior was tolerated as an idiosyncrasy. [BACK]

1. See, e.g., Fyzee 1964: 35. [BACK]

2. Hallaq 1984. [BACK]

3. For a detailed description of Dyula weddings, see Launay 1975. [BACK]

4. For a more detailed discussion of the Ivoirian Civil Code and its impact among the Dyula, see Launay 1982: 139-45. [BACK]

5. For reasons I do not understand, these lamb sandwiches seem to form part of the ceremony. The only other sermons I witnessed in which meat was distributed to the audience were during donba , a Muslim calendar holiday. However, donba sermons, unlike all other sermons (including wedding sermons), are an occasion for competitive ostentation in providing food for the audience. [BACK]

6. The entire Saganogo kabila of Koko is an offshoot of the Saganogo of Kong, and consequently its members are all agnates of Mammadou-Labi. This certainly reinforces his links to the Dyula of Koko, though the fact is not relevant to the issue of wedding sermons as far as I know. [BACK]

7. These names are all pseudonyms, except for Mammadou-Labi Saganogo. As he is a nationally known figure, and since his attempt to promulgate the new wedding ceremony was an explicitly public act, I see no reason not to give him the credit for it. [BACK]

1. This distinction, too, is hardly unambiguous; the term mory may also be used to refer specifically to the learned. Among the Mende of Sierra Leone, "morimen" are specialists in written Islamic magic, whereas " karamokos " specialize in the teaching of Arabic (Bledsoe and Robey 1986). Plausibly, the terms mory and karamogo were originally synonymous; through a process of semantic displacement, the distinction has acquired different shades of meaning in different local contexts. [BACK]

2. Both siru and bayani are almost certainly borrowed from Arabic, but their specific derivations are open to question. J. R. Goody (1968: 236) derives the Hausa and Gonja equivalents of siru from the Arabic sihr , "magic" (cf. Eickelman 1985: 67). However, John Hunwick (personal communication) has suggested that these terms might derive from sirr , "secret," and bayan , "clear, evident," a term also associated with public speaking (Holden 1966, cited in Goody 1968: 224). Either derivation of siru is intrinsically plausible. Conceivably, the Dyula term collapses the meanings of both Arabic words. [BACK]

3. I use the term laymen with considerable reservations to refer to individuals who are neither enturbaned scholars nor pursuing an advanced course of study that could eventually qualify them for the turban. It must be stressed that Muslim scholars are not the equivalent of the clergy in Christianity, as the use of the term layman might imply. Indeed, the Dyula have no word or expression to denote nonscholars. [BACK]

4. The few exceptions are all young, unmarried, or very recently married men, who may yet choose to pursue their studies elsewhere if they are serious aspirants to the scholarly profession. The most advanced of these young men was in any case studying with his own father. [BACK]

5. They are always given on a Thursday or Saturday night, technically "Friday" or "Sunday," holy and propitious days; nights are reckoned as parts of the following rather than the preceding day. [BACK]

6. On the similar "ossiffication" of the khutba in early modern Egypt, see Gaffney 1987: 202. [BACK]

7. It is perhaps relevant that this particular scholar happened to be a worosso , a person of slave status with license to joke obscenely (see Launay 1977a). Though the joke was not, strictly speaking, obscene, it might be considered too coarse for a scholar of "free" status. But it must be stressed that such "free" scholars, although they must avoid coarseness or excessive buffoonery lest they appear undignified, are by no means averse to using humor as a rhetorical device. [BACK]

1. This surge of interest in West African Sufi orders has spawned a literature far too vast to cite comprehensively. For an overview of the literature, see Hiskett 1984: 244-60, Levtzion 1986: 12-16, O'Brien 1981, and O'Brien and Coulon 1988. [BACK]

2. Hiskett 1984: 258. [BACK]

3. The term recrudescence is J. Spencer Trimingham's (1962: 155). He uses it with specific reference to the nineteenth century, but recent scholarship is increasingly turning to the eighteenth century for the social and ideological roots of the movements he describes. [BACK]

4. Even more important, although Hiskett does not make this point, the "local patriotism" fostered by the Sufi orders and expressed in terms of allegiance to them may transcend or override traditional lines of segmentation or factional cleavage. Evans-Pritchard's (1949) analysis of the Sanusi order in Cyrenaica stresses precisely this aspect. [BACK]

5. Hiskett stresses the links between the Sufi orders and various jihad movements in detail, but he also points out quite clearly that many Sufi leaders did not subscribe to the ideology of military jihad. [BACK]

6. It is clear that at one time membership in various Sufi orders was not at all exclusive, and that individuals could be initiated into several of them. This is, by and large, no longer possible, and certainly not in West Africa. [BACK]

7. Cohen 1969 has argued that this is the case among the modern "Hausa" of Ibadan. It might be argued that the Mouride order has played a similar role in the development of Wolof ethnicity in Senegal. [BACK]

8. René Otayek (1988: 98) ascribes the absence of tightly institutionalized Sufism and of charismatic Sufi leaders in Burkina Faso to the "Mossi kingdoms' impermeability to outside influences." This kind of argument is an excellent example of the dangers of elaborating a "typical" model and of attempting to explain deviations. As noted, Sufism in Burkina Faso shares many features with Sufism

just across the border in northern Côte d'Ivoire—but the influence of the Mossi kingdoms is decidedly not one of them. [BACK]

9. One exception to this rule is the phrase "A ka di Alla ye" ("It pleases God"). For obvious reasons, this is a special case. Somewhat similarly, persons in positions of authority—elders or parents—may wish to emphasize that they are personally pleased or displeased by the specific actions of others, especially subordinates. [BACK]

10. While it might be labeled "singing" by Westerners, liturgical "chanting," either in Arabic or in Dyula, is never associated with dancing and is never referred to as donkili. It falls rather into the broad category of kalan , along with reading, reciting, studying, teaching, and delivering sermons. Unlike "singing," "chanting" is an exclusively male activity and confers religious merit. [BACK]

11. RDA is in fact a Dieli from Koko. However, for most purposes, the Dieli community of Koko has been completely assimilated to the Dyula. It is interesting, though, that in most dialects of Manding, the term dieli denotes a "caste" of griots. Dyula of Koko use the term dieliba for hereditary griots, to distinguish them from their Dieli neighbors. RDA is often called a "griot" (using the French term), but never a dieliba. Perhaps his hereditary Dieli identity helps make his adoption of griot-like behavior more acceptable to the local Dyula population. [BACK]

12. In Koko, this is usually the Friday mosque, which doubles as a local mosque for the quarter. [BACK]

1. Combs-Schilling 1989 provides an extended analysis of the symbolism of this specific "blood sacrifice" in Morocco. [BACK]

2. Ties specifically with in-laws (including prospective in-laws) are stressed, not during tabaski but rather during Ramadan, where they invariably receive gifts of sugar. The symbolism is suggestive: "blood" ties are acknowledged through reciprocal, but asymmetrical, gifts of meat; affinal ties through unidirectional (husband to wife's parents) gifts of sweets. [BACK]

3. Crudely, this amounts to one-fortieth of an individual's income. However, there are numerous qualifications (the prescriptions according to Maliki law are outlined in Qayrawani 1968: 126-39). Laypeople may need to consult scholars in order to determine the extent of their obligation. [BACK]

4. Cf. Goody 1968. [BACK]

5. In fact, I knew of no one from Koko reduced to begging, at least in a literal sense. However, Korhogo is very close to the borders of the much poorer countries of Mali and Burkina Faso, and conse-

quently attracts a certain number of beggars, including blind ones, from outside. [BACK]

6. Frequently, bride and groom belong to the same group, given stated preferences for in-marriage (Launay 1975 and 1982). [BACK]

1. Cf. Brown 1972: 220: "Altogether, the intellectual events of the last decade of the fourth century are a reminder that controversies emerge not only when people have new ideas, but when they suddenly wake up to realize that they no longer have the old ones." [BACK]

2. There is a vast literature on African sectarian churches. The pioneering monograph was Bengt Sundkler's Bantu Prophets in South Africa (1948). More recent studies include Peel 1968, Fabian 1971, Augé 1975, Jules-Rosette 1975, MacGaffey 1983, and Fields 1985. [BACK]

3. See Mudimbe 1988: 161-75 for a review of the issues involved, from a philosopher's perspective. [BACK]