West African anthropology carries the burden of a chasm between what is considered traditional or authentically African and Islamic. This reveals itself in ignoring Islam, misrecognizing the cultural legacy of Islam in today's lives beyond self-professed Muslims, and exaggerating the contrast between the "pre" and the "post" in recent cases of conversion. A more balanced, historically informed understanding of contemporary Africa requires greater awareness of the central role of Mediterranean links and the canvas of meaning deposited by Islam. Sections on mobility and literacy provide a rapid survey of these themes, which are emblematic of what social anthropologists ought to bring to the forefront of their vision of West Africa, though typically they do not. A section on politics explores the framework for the impact of Islam, and a final section, on mimesis, discusses some cultural processes still at work. We need to reimagine West Africa, both to reach a new cosmopolitanism to transcend the we--they contrast, and to allow anthropology to make more significant contributions to the study of contemporary Africa.
In West African anthropology, a glass wall separates the small room reserved for Islam from the great hall exhibiting the ethnographic splendors of sub-Saharan cultures. Showcased in the small room are classical pioneers in Senegal (Copans 1980; Cruise O'Brien 1971), a younger generation of scholars on Mali (Soares 2005), and milestones of northern Nigeria, going back to the venerable name of S. F. Nadel (1946). There is no question of the quality of what is found there; the problem is how thoroughly the glass wall divides the two sides.
Islam is relevant, one would think from the rest of the anthropological literature, only for the specialist of this religion, or for those interested in recent developments in limited areas of West Africa. What is genuinely African has nothing to do with it. The anthropologist interested in shrines, ancestors, cosmogony, or agriculture does not need to know much about it. It is as if the broader academic division of labor between the orientalist tradition, specializing in scriptural civilizations, and the ethnological tradition, specializing in "native life," is reproduced within anthropology itself.
Anthropological accounts of the colonial period exude the language of "first encounter." In typical anthropology, one finds at most a nod at the baseline of post-sixteenth-century Atlantic trade, but little recognition of earlier links with the Mediterranean basin, Europe, and West Asia.
In this respect, the work of social anthropologists of West Africa has profoundly diverged from the work of historians, who have increasingly focused on the Islamic component of the past in this region. (1) I think the lack of interest in past cultural flows between West Africa and the neighboring regions is an unrecognized perpetuation of patterns of thinking introduced in the colonial period and distort our understanding. These patterns have limited the contribution that anthropologists could make to the understanding of modern West Africa, and have hindered them from participating in African intellectuals' effort to overhaul colonial stereotypes. Given the weight that anthropology has had in all types of social and humanistic studies of West Africa, the issue has relevance beyond the borders of one discipline. I suggest here that we resituate West Africa in our imagination, by putting Islam near the center of our training and thinking--as a major ingredient of West Africa's historical heritage, and not as a "foreign" incursion.
For several centuries, Islam served as a template and vehicle for the exchange of many traits between West Africa and the Mediterranean, Europe, and the Middle East. The canvas of meanings thus created shaped people and milieus well beyond the ranks of confessed Muslims. This cultural content resurfaces today, sometimes stripped of its overt religious significance, on all sides of local social boundaries. We cannot fully make sense of the ongoing conversations in West Africa, between Muslims and non-Muslims, among Muslims, or even among Christians, without considering this substratum of wider historical connections. The rapid and thorough Islamization of many parts of the region, under colonial administrations unfriendly to Islam, is a topic on which earlier historical contacts would throw considerable light.
The scant attention paid to Islam in the anthropology of West Africa can be illustrated from literature on the Yoruba, the largest for any ethnic category in Africa. The Yoruba chapter in a widely used reader states in its religion section that it will omit Islam and Christianity, although it reports that the devotees of traditional cults were only 6 and 20 percent in two areas the author studied (Lloyd 1965:574). The omission is distorting also for those who are not Muslim. A novice in Nigerian studies would be surprised to discover certain information buried in the literature on the Yoruba: that the If a oracle derives its name from al-fa'l (Arabic for "good omen," also "fortune"), that it shares its basic procedure with a type of astronomical geomancy that existed in Italy and France until at least the seventeenth century, and that Orunmila is possibly a Yoruba vocalization of Arabic al-raml "divination by sand" (Morton-Williams 1966:407). The sharing of esoteric knowledge between Muslim and non-Muslim specialists can account for such adoption in Mali and in Nigeria, and for its further indigenization among Fon-speakers of the Republic of Benin (Brenner 2000:153-162). The Ifa oracle became at some point a focus for resistance to Islam, suggesting how awareness of Islamic cultural connections complicates and deepens our understanding of West African developments.
Continuing indifference to Islam in Yoruba anthropology can be illustrated with Apter's highly acclaimed contribution on Yoruba religion (1992). I have used Apter's book as reading in a class. A group of undergraduates had the task of preparing, from outside sources, a small report on the Yoruba. They encountered much discussion of Yoruba Muslims in the resources that they consulted (encyclopedias, websites, etc.), at such variance from the picture of their anthropological reading as to make them uncertain whether they had hit upon the right literature. Let us hope the inclusion of an erudite chapter on Islam in Peel's historical anthropology (2000) heralds a new attitude.
In French ethnography, the Dogon have a comparable status. The Dogon and the Manding-Bamana of Mali were the favored subjects of Marcel Griaule and his followers. Griaule and Germaine Dieterlen were fond of drawing parallels between their findings and ancient Egyptian and Greek thought, but assiduously avoided any reference to obvious connections to the Islamic centers in Bandiagara (Brenner 1984; Tamari 2001). According to van Beek (2004:57), they also overlooked the impact of contemporary Christian missionary activity on their Dogon interlocutors, thereby missing the opportunity of exploring how it resonated with this heritage.
Islam and traditional Africa are conceptually kept apart, even in settings that are now predominantly Muslim. Berliner's recent study in Guinea, set among "six thousand rice cultivators ... (e)thnographically speaking ... practically unknown," investigates what children of first-generation observant Muslims say about ritual practices that predate "the introduction of Islam" (2005:580. The word introduction here is a gloss for conversion, in the 1950s; before then, the people were "great sculptors, ritual specialists, and drinkers of palm wine," with a custom involving "invisible entities, ritual houses, sacred groves, initiations, secrets, and ritual prescriptions" (2005:579). How worthy, indeed, of the great hall!--until we remember that the mid-Atlantic coast people of West Africa, of which these few thousand are part, have included Muslims, and centers of Islamic political power, since the thirteenth century, with well-documented trade and collaboration among non-Muslims, Muslims, and Europeans since the seventeenth century. A different kind of study on historical "memory" is conceivable in this area, on the echo of past interactions, fluctuating borders, and oppositions in this seemingly abandoned and yet remembered custom, on how the pagan was constructed by the presence of the Muslim and the extent to which this helps understand the way today's self-confessed neo-Muslims imagine their non-Muslim forebears. Such a study would presuppose a historical continuum, tearing down the firewall between the "pre" and the "post" of the recent episode of conversion.
The tight separation of Islam and pre-Islam owes much to colonial circumstances: the need to justify military occupation, and later the concerns of political control, made European observers reluctant to recognize West Africa's connections to neighboring world areas. It was comforting to think of the native subjects as bursting into the stage of world history with European exploration and colonial domination (a way of thinking established in the encounter with Native Americans and Pacific Islanders), rather than under categories that made them old adversaries. Islam is not amenable to the sort of reshaping involved in Griaule's romantic metaphysics of the Dogon; it represents a totally different kind of Other. Muslims, too close and too well known to Europe, served as cosmic foil against which Europeans defined themselves as world civilization (Eaton1993:2). The extirpation of Islam was for European observers a way of distancing the image of West Africa from the unsavory historical contest between east and west, and relocating it in more fragrant exotic nature. As well, in humility and self-criticism, we should wonder if a professional stake was not present. Removal of the subject of study from history made the discovery cardinally alien and all the more spectacular.
We can contrast the ethnographic representation of West Africa starting in the end of the nineteenth century with European representations from the period before colonization. Renaissance pictorial representations of West Africans may have been fanciful, but made no clear distinction between the Black and the Moor, both presented in vaguely Islamic accoutrements (Mark 1974). This was consonant with the fact that most of what Europe knew about West Africa derived from Islamic sources or from Mediterranean sources in the orbit of Islam. Conceptions had not yet changed at the dawn of the period of European explorations, for example, when Rene Caillie traversed, from south to north, the area in southeastern Mali that lies now near the border with Burkina Faso, armed with an approximate Arabic learned among the Brakna Moors and pretending to be a young man of Egyptian origin (Caillie  1979).
It would be credulous to attribute the transformation we are contemplating to the influence of one or two persons, but key individuals who served as precursors and signposts for this sea change in European imagination of West Africa can be identified. One of them was Louis Faidherbe, the mid-nineteenth-century governor of the embryonic colony of Senegal, mostly known as the person who set the stage for the aggressive policy of expansion to the interior, which came to fruition a generation after him. At the same time, Faidherbe produced considerable ethnographic and linguistic work, and participated in policymaking, providing the foundation for a new way of conceiving sub-Saharan Africa, which in due time became colonial ideology and a fountainhead for scholarship (Amselle 1996:117-150; Manning 1998:12-14).
Faidherbe's pioneering linguistic classification, combined with nineteenth-century racial ideas, underlay the novel conception that the societies of the West African savanna were distinct from those north of the Senegal River, and their specificity lay in that each linguistic community was profoundly original and separate from any other. Islam and trans-Saharan trade had a pernicious effect on this juxtaposition of ethnic groups. Amselle places Faidherbe at the head of a major lineage of thought, including the explorer Gustave Binger, writer-administrators such as W. Ponty, L. H. Lyautey, Paul Marty, and Maurice Delafosse, and the politique des races philosophy of administration that blended in French West Africa direct and indirect rule in its own peculiar manner. The conceptual separation of tribal black Africa from Mediterranean Muslim Africa had a practical corollary in the concern of British and French governments with Islam as military threat. They tried to cut off their West African Muslim subjects from Arabs and other Muslims, and to slow down the rate of conversion (for the French colonial world, see Conklin 1997; Harrison 1988; Saul and Royer 2001; Triaud 2000:177).
One way to reorient this inherited image of West Africa is to bring back into our field of vision the region's historical ties to the Mediterranean world. Through the Middle East and southern Europe, West Africa connected with the rest of Eurasia. The historian McNeill (1963) introduced the concept of Eurasian ecumene, joining the Mediterranean rim with Greater India, China, and the Malay Archipelago on the one hand, and northern Europe on the other. More properly, we should call it the Afro-Eurasian ecumene. In this vast expanse, layers of transformative influences bound together the destinies of millions of people. More recently, Philip Curtin coined the term Atlantic Community for a similar pattern of planetary connections, which started to take shape in the sixteenth century. Before the Atlantic Community was a Mediterranean Community, which had lasted much longer, and continued in tandem up until the onset of colonialism, at the turn of the twentieth century. What came to be recognized as good style in anthropology discourages the study of the traces and effects of this interaction, and they gradually sank into oblivion. Christopher Steiner (1994:127-128) describes how nineteenth-century European trade beads have become valued collector's items: tourists, dealers, and expatriates enthusiastically purchase them in Africa, as if they were ethnic art. I think many a field notebook filled with West African discoveries includes snippets that are the ethnographic equivalent of tourists' wonder at the encounter of millefiori beads in the stall-booth of a West African marketplace.
We should be careful to distinguish what is proposed here from diffusionism. The vocabulary of diffusion seems to tie cultural elements irrevocably to an alleged place of origin, keeping them forever foreign to the receivers. The view I offer is instead one of historical exchange, absorption, adaptation, and developments in tandem. Diffusion has been opposed to agency, because it seems to be underlain by an epidemiological model, as if the proclivity to spread is inherent in the cultural item itself. To choose, adopt, assimilate, or imitate require on the contrary exercise of agency. For this reason, I am somewhat uncomfortable with the more recent image (of structuralist inspiration) of postmodern Africa as home to endlessly ecstatic bricolage and multiple modernities, as Young (2003:2) puts it. Appadurai (1996) offered the suggestive expression "production of locality," to point out that the global forces of economic integration and culture flows of our age do not seem to be moving us to a terminus of world homogeneity. In past centuries, despite different modalities of contact and economic process, economic articulation and the flow of culture obviously did take place, at varying rhythms, as did the "production of locality." The coalescence of features results not in simple aggregation, but in original patterns of thought and behavior, which cannot be traced back to any of the parent forms. In premodern times, as at present, the availability of foreign elements was a resource on which locals could seize, and sometimes resulted in fateful turns in local trajectories. My conclusion is that for all the communities and places of the past, we need to keep in sight that they also participated in a world-time, which has an "active presence" in their present.
For some West Africans, Islam has been for centuries the basis of a universal identity, as Christianity, mathematics, science, or Greek philosophy are considered the foundations of European identity. Many Africans who do not identify with Islam live in a cultural web including Islamic and Mediterranenan elements experienced as part and parcel of their being. The focus should be not on the former foreignness of each element, but on the existence of historical links, which make West Africa like any other place on earth. Consider the seven-day week, which is part of the reckoning of time in many areas of the West African savanna. Combined with a shorter market week (five-day in Senufo and Juula areas, three-day in Moose areas), which generates a longer cycle (of, say, 35 or 21 days), it shapes local ritual and agricultural calendars, and may no longer be associated with Islam. But the day-names used across the region are local vocalizations of their Arabic equivalents--in Juula: Kari 'Sunday', Ntene, Talata, Araba, Alamisa, Jejuma, and Sibiri. (2) Many West African languages adopted these names not directly from Arabic, but from a neighboring African language, as they did for a vast number of words of Arabic origin. To stress "borrowing" in these cases would reveal that we place West Africa in a special space. The same seven-day week reached Europe with Christianity, but we do not think in this instance of borrowing. We further know that the seven-day week did not originate with Islam or Christianity, and can be traced to Mesopotamia. It is an adoption in the Arab world, and in Jewish and Christian calendars. It has the same status in West Africa and Euro-America. In contrast, to think of how we share these elements can be liberating, by helping free our minds from a few layers of exoticizing filter.
In premodern times, the dry expanse of the Sahara slowed down contacts between North and West Africa, but only in the manner of a porous membrane. It is no small detail for European history that much of the gold circulating in Western Europe before the discovery of the Americas was supplied by the West African trade. Its counterpart was the intellectual processes associated with Islam in West Africa--my present focus.
Islam arrived in sub-Saharan West Africa more than a thousand years ago, following the westernmost of its three major historical pathways between the Mediterranean coast and the interior: the trade routes leading from Sijilmasa in southern Morocco to the central Niger, the shortest crossing of the inhospitable environment of the desert. In the central Niger River area, Islam became acclimatized, generated its own distinctive scholarly conventions and institutions, and spread toward east and west in the grasslands. Through the movement of ideas and people, and through conversions, Islam then disseminated southward, through the upper Senegal and Gambia valleys. When it reached the rainforest, it spread eastward again, through a corridor along the forest and savanna borderline (Levtzion 1968:10-18).
Thus, in its earliest and most important phase of expansion, Islam moved in West Africa mostly from west to east, and in two parallel lines of progression. In this development, it preserved features that were brought together in the Western end of the Maghrib, including adherence to the Maliki madhab and the kufi script. It was only in the early seventeenth century that a more central route of connection to the Middle East opened up, starting from Tripoli and reaching Hausa country in what is now northern Nigeria, by way of long, arduous stages through Murzuk, Ghat, Agades, Katsina, and Kano. (New research on lapidary inscriptions may qualify this view; see Moraes Farias 2003:303.) With this connection, influences from Egypt and the Middle East started flowing to West Africa directly, constituting a recognizably more recent layer of Islamization and Mediterranean influence. In the Volta region (Burkina-Ghana), Islamic influences from west and east, Manding and Hausa, met and blended.
In what follows, I focus on four themes from the history of West Africa. Two of these are mobility and literacy, which seem emblematic of what anthropologists need to remember to modify their inherited disposition toward West Africa. The third is a detour on politics, highlighting cultural processes other than opposition or conversion. The fourth, mimetic appropriation, complements the discussion on politics. The whole represents an incursion into the historical and ethnographic literature, perforce idiosyncratic; it aims to add resolve onto the desire for a new perspective, which I hope is already felt by many in the scholarly community, rather than to convince the reluctant by the force of syllogism.
The high propensity of Muslims, throughout centuries, to move to distant lands in numbers that are startling from the perspective of premodern Europe is a standard observation in the specialist literature. Not too long ago, two anthropologists edited a beautiful volume on this topic (Eickelman and Piscatori 1990). This is also true for West Africa, although the mobility of both Muslim and non-Muslim has been downplayed in common perceptions. Since the stories of the "discovery" of the African interior by European explorers imply by omission stationary natives, serving as backdrop to another image, a multitude of stable cultures, which the task of anthropology is to describe one at a time, it is worthwhile to dedicate a few paragraphs to travel in precolonial West Africa.
Any discussion of traveling Muslims has to start with the hajj, the pilgrimage to Mecca. Malian ruler Mansa Musa's 1324-1325 pilgrimage to the holy city of Islam has been described in numerous popular venues, and is therefore familiar, but without the larger cultural context, it appears as a singular event. Musa was one of six Malian and Songhay rulers who performed the holy act, all within the span of three centuries. In Islamic history this is remarkable, because Muslim rulers were less inclined than their subjects to perform the hajj. On his way to Arabia, Musa amazed the population of Cairo with the size of his retinue and the opulence of his lifestyle. Anecdotes of his passage eventually found their way to European sources; the most informative Arab account of the visit, written by al-Umari, who went to Cairo twelve years after the event, can be found in Levtzion and Hopkins (1981:268-273).
Pilgrimage to Mecca by West African Muslims continued in subsequent centuries, their numbers oscillating with economic and political conditions, and most of these were people of more modest means: scholars and pious men, not only from the towns of the southern Sahara or the Sahel, but also from Senegambia, the Volta region, and the part of the southern savanna included now in Cote d'Ivoire, Ghana, Togo, and Benin. In the seventeenth and eighteenth centuries, from the Bandama River to the Volta valleys, towns that had an important Islamic presence, such as Kong and Bouna, organized caravans of pilgrims (Boutillier 1993:347). The biographies of many eighteenth- and nineteenth-century West African Muslims who were important enough to be written about include an episode of hajj. A belt of villages and kinship groups of West African origin stretches from Lake Chad to the Red Sea, owing its origin to permanently settled pilgrims en route to or from Arabia, a testimony to this centuries-long traffic (Stewart 1990:201). At present, relative to population West Africa sends as high a rate of pilgrims to Mecca as any Islamized region in the world, and irrespective of income differences (Bianchi 2006: 61-63).
In the days before the modern means of travel, a journey to the holy lands was naturally quite different from the hajj package deals of today. Al- Naqar (1972:92-113) describes the routes the travelers followed and how these developed over the centuries; Birks 1978 focuses on the twentieth century. From West Africa, it took five to eight years to accomplish the journey, with large detours to lands that might appear today totally out of the way. It involved frequent and extended stopovers, partly in order to earn the resources that would allow one to continue the journey, but also to take advantage of the teaching of famous scholars. The hajj was a life-turning event, but could also be undertaken to crown a locally eminent scholastic career. The stay in Hijaz (West Arabia) itself usually involved, after the initial hajj, visits to various cities. It was commonly augmented with a minor pilgrimage to Jerusalem (ziyara), which houses the second-holiest shrine for Muslims. In the Volta region not so many years ago, I met elderly Marka Muslims who desired a pilgrimage to Jerusalem more dearly than a pilgrimage to Mecca.
Most pilgrims who went overland passed through the great cosmopolitan center of Cairo. In 1817, Hutchinson encountered in Kumasi a Muslim from Jenne who had apparently been an eyewitness at Nelson's victory over the French fleet in 1798 off the Egyptian coast (Goody 1968:200). Some others targeted a closer Mediterranean seaport, from which they sailed part of the way, before joining an overland caravan to Arabia. Morocco, Istanbul, or the ports of the Levant were common transit points for them, as Clapperton ( 1966:206) discovered in Sokoto. Underestimating the cosmopolitan connections of West Africans could have unforeseen consequences for the nineteenth-century European explorers. In the last leg of his trip before reaching Timbuktu, Heinrich Barth pretended to be a Muslim from Istanbul to conceal that he was Christian, but his cover was blown when a man tried to engage a conversation with him in Turkish at the gate of Timbuktu (Barth 1857-1858:4:404-405). Grosz-Ngate (2006) discusses other examples of this erudite explorer's encounter with well-traveled coevals, including a man who addressed him in Greek, and one who could tell stories about the English from his stay in Bombay.
The wanderlust ignited by the performance of the hajj finds its paragon in the person of ibn Battuta, who in 1325 set out of his native Tangier for Mecca, but continued to travel, to accomplish twenty-nine years later visits to lands lying today in forty-four countries, for an estimated total of 73,000 miles through land and sea (Dunn 1986:3), making him the greatest known traveler of premodern times.
Pilgrims brought back from the hajj world news, fresh knowledge, contacts, objects, and occasionally persons. Mansa Musa returned from his pilgrimage in the company of jurists and scholars, and an Andalusian poet and architect, Abu Ishaq al-Sahili, who built him a palace and the great mosque of Timbuktu (Levtzion 2000:68-69). At the end of the fifteenth century, the Songhay ruler Muhammad Askia met in Cairo, on his way to Mecca, the scholar al Suyuti, who introduced him to the Abbasid caliph (Levtzion 2000:70), and when he returned to his country, Muhammad Askia established an advanced school in Gao.
Religious innovations and interpretations diffused from the Hijaz, which continued to be a center for the transmission of ideas, also from one West African to another, well into the twentieth century. In Medina and Riyadh, several African teachers held schools, which served as hospices particularly popular with sub-Saharan pilgrims (Harrison 1988:197; Stewart 1990:201). In 1978, during fieldwork in southern Moose country, Burkina Faso, I met an elderly religious leader of Bisa origin, Mahamudu, who had gone to hajj with a passport issued for him in Cameroon in 1930. He had spent in Arabia eight years to continue his advanced education under a teacher originally from Masina, who introduced him to the way of Tijaniyya (Saul 1984). Upon his return Mahamudu became a well-known teacher, stimulated numerous conversions, and founded a new rural Tijaniyya community around him, which at the time of the investigation was in full swing and growing under younger leaders. Tijaniyya, which had originated in North Africa, became in the nineteenth century the most important Sufi order in West Africa, largely aided by the hajj connections to the Hijaz. It is now a major intellectual export of West Africa: in Turkey in the 1950s, Tijaniyya adherents assumed the vanguard in opposing the Europeanizing state; in the United States, a small African-American community maintains close ties to the Senegalese brotherhoods (McCloud 1995:93).
The hajj has a profound impact on the sense of the self. In addition to facilitating the circulation of ideas, it sharpens the feeling for global community among Muslims. The focus of the pilgrimage, the cube-shaped building called Kaaba, was built, according to Muslim tradition, by the Prophet Abraham. When the Prophet Muhammad returned from his exile in Medina, he entered this buildingto remove its tribal idols and restore its original purity. The Kaaba stands for God's covenant with the Muslim community. In pilgrimage, Muslims live intensely a return to the sources of Islam with a sensory experience of worldwide kindred. They become "more mindful of God's indifference to their differences" (Bianchi 2006: 45). The hajj rituals end with the Feast of Sacrifice (Id al-Adha), commemorating God's command to Abraham to sacrifice his son. In West African French inherited from colonial times, this festival carries the remarkable name Tabaski--a marvelous glimpse at the underground rivers that connect Afro-Eurasian cultures, because it seems to be a Berber derivation from the Hebrew-Greek Pasch "Easter." On the same day, Muslims across the world typically sacrifice a lamb, joining in imagination the pilgrims who reenact the story of Abraham and refreshing their sense of belonging to a single community and a universal tradition. Even in the daily routine of an observant Muslim, turning toward the Kaaba in prayer reorients and resignifies local space by making it part of a global expanse, a long distance from the fragmented tribal world of ethnography.
For many non-Muslim West Africans, Mecca became a religious icon. Oduduwa, the mythical father of all Yoruba people, for example, is said to have arrived from Mecca (Chazan and Abitbol 1988:31). References to Mecca in the jo societies of Manding- and Senufo-speaking areas of Mali and Burkina Faso will be discussed below.
Travel created a two-way traffic. The canvas of Islamic civilization encouraged not only Africans to travel abroad, but also outsiders to travel to West Africa. After Mansa Musa returned from his hajj, the Alexandrian merchant Saraj al-Din followed to claim a debt from a loan that he had extended him. Traders, scholars, and adventurers arrived in the great Sahelian cities of West Africa from the Middle East, Spain, and North Africa, including renegade Christians, as in the case of Judar Pasha's army, which occupied Timbuktu for the Moroccan sultan in 1591. The Sahelian cities, in turn, maintained trade and scholarly relations with localities in the southern savanna (Stewart 1976). In the early nineteenth century, the first British missionaries visiting Lagos met traders from Tunis and Constantinople (Peel 2000:191). In 1822, Englishmen in Cape Coast received an elderly merchant hailing from Astrakhan, who had crossed West Africa starting at Tripoli and passing through Ghat, Agades, Katsina, Kano, Timbuktu, Jenne, Kong, Bouna, and Salaga, to end up as guest of the Asantehene in Kumasi (Wilks 1967c). West of Kano, he had seen two other Europeans of undetermined nationality, who had had a boat accident on the Niger. Visitors were not always transitory; some of them settled down and left descendants in the area, as evinced in proud claims heard today in places scattered between the Sahel and the rainforest.
Within West Africa, long-distance trade and migration mixed people, explaining the wide dispersion of Manding-, Hausa-, and Jerma-speaking communities. These movements were not slow, amorphous drifts of populations blind to their fate and trajectory, but planned actions of flesh-and-blood persons, undertaken with clear motives, which can be recovered from memory. Migrant communities in West Africa predate European colonialism. In 1888, the French explorer Gustave Binger encountered around the town of Wahabu, Burkina Faso, numerous immigrants from Kaarta and Bakel, areas more than 800 miles to the west, who had arrived so recently that they gathered around his men to ask about news of the relatives and friends that they had left behind (Binger  1980:1:421).
Pursuit of knowledge was an important motivation for travel. Muslims placed a high value on the peripatetic tradition of education, and aspiring scholars went from one renowned master to another. Travel for scholarship resulted in diffuse networks of acquaintances, friends, and soulmates, and accounted for the rapid diffusion of news and innovations. The biography of Abu Bakr al-Siddiq, a boy born in the eighteenth century to a clerical family, illustrates this mobility (Wilks 1967a). He was born in Timbuktu, but educated in Jenne. His mother came from Katsina (about a thousand miles east of Jenne), where his father conducted trade trips. Before reaching the age of nine, he had traveled with his tutor more than 600 miles. Then he spent a africatodAy IslaM and West afrIcan anthropology 1 year in Kong, and went to Bouna for further education. In all these places, he had relatives and in-laws. Colonial investigator Paul Marty remarked humorously that belonging to Islam was like belonging to a touring club (Goody 1968:240), which included some very young members. We will return to Abu Bakr's biography in a moment.
Sub-Saharan Africa in general is often treated as the quintessence of orality, that is as fundamentally preliterate, nonliterate, illiterate, or whatever. What has been said so far already suggests that this image is inaccurate. Before colonial and postcolonial schooling started taking effect, West Africa had a type of literacy that did not conform to our current norms, but the comparison is anachronistic. The ideal of universal primary education, or the objectives of contemporary adult literacy programs, are based on a feeling that can be called "interiority," which, according to Ong (1982), resulted from solitary writing, reading, and then print technology, fully achieved in Europe during the nineteenth century. In premodern settings, the interaction between orality and literacy was of a different nature. Goody and Watt (1968) provide an extended discussion of this issue with emphasis on sub-Saharan Africa and Asia. Goody's pioneering writings on the significance of Islamic writing in West African culture history (1968, 1971) stimulated further anthropological work, but some of this displays the curious slant of focusing on the empty half of the glass. Bledsoe's discussion on the uses of literacy among the Mende of Sierra Leone, for example, stresses how they reinterpreted and adapted it to their own cultural environment (1986:204). Culture appears then as a self-contained organism, capable of absorbing only elements that it can completely digest without itself changing, and this in a setting that includes not only Arabic writing for at least three centuries and English for more than one, but also the local invention of a whole new alphabet. Would it not be more appropriate to write about literate and nonliterate persons, rather than cultures or societies, and the different uses to which people put literacy, which may evolve in a lifetime, without speaking of centuries?
What is most commonly known about writing in West Africa is the scholarship carried out in Sahelian cities, the first solid account of which was provided by Heinrich Barth (1857-1858:4: ch. 66). Timbuktu authors became better known in the colonial period, thanks to the translation efforts of scholars such as Octave Houdas and Maurice Delafosse (for an update, see Hunwick 1999). Other important centers of literacy existed in West Africa, such as the fabled Gana, and on the Niger valley Ja (Dia), Gao, and Kabora, and commercial diasporas extended out of them after the eleventh century (Levtzion and Pouwels 2000:3; Reichmuth 2000:426). More recently, the Arabic collection of the Institute of African Studies at the University of Ghana that resulted from manuscript-collection drives since the 1950s (Levtzion 1968: xviii-xix) and studies by historians in northern Nigeria gave us a better sense of the centers of writing in West Africa and its role in political and economic history.
In precolonial West Africa, literacy was the specialty of a minute minority and of the type that Goody dubbed "restricted." This qualifier can be understood to refer to the number of literate persons or to the uses of literacy--concepts that raise different issues. The number of people who could read and write, while very low in the general population, was high within particular communities, and this minority was scattered everywhere in the savanna. It provided specific services to many people, frequently including humble villagers. Restricted literacy is not limited to West Africa, but also characteristic of many great "civilizations," which are not generally thought of as "preliterate," because literacy can influence social and economic processes and affect common attitudes without being widespread or all-purpose. This last point was brought home to me thinking on an observation I often made among Bobo farmers in Burkina Faso. Many elderly people I interacted with in the 1980s had not been to any kind of school, but kept a carefully wrapped package of documents--ancient tax records, expired identification cards, letters, old army discharge papers, prescriptions--that may have belonged to a grandfather or great-uncle; or a man would produce, from the bottom of a trunk, a tattered notebook, in which his father had a student scribble notes on the location of his farm in particular years. They hoped, it seemed, such records would serve in the future. Occasionally too, I noticed a worried glance at my notebook. This reverence for papers has perhaps been intensified by the mysteries and misuses suffered in a century of dealings with government bureaucracies, but I am convinced that it is also rooted in precolonial associations. I interpret in the same light the regret, frequently reported in the West Africanist literature, that traditions have been lost because the ancestors did not know how to write them down. Its counterpart is the aide-memoire pages on family traditions, previously in Arabic, now mostly in schoolboy French, that many house heads jealously preserve, even if they are unable to read them. The Bobo core vocabulary includes words for 'read'(kala) and 'write' (saba), which ultimately derive from Arabic but were most probably adopted from Juula.
It is ironic that literacy has such a low profile in anthropologists' imagination of West Africa because it had such an important role in the religious history of the region. Greenberg (1946:10) remarked that in West Africa, the implantation of Islam and its influence on native ritual practices did not take place through intensive contacts following from the massive influx of a foreign population (as, for example, it happened in North Africa): they resulted from the spread of literacy and books. The native scholarly class adapted, and adapted to, what they found in the written and printed sources at their disposal. From them and modest clerks, the use of Arabic writing passed to communities that did not care to have members who had mastered the technique, but valued it as fetishized elements in shrines and masks, as the ubiquitous charms or writing-water potions.
An important conglomerate of clerical groups in West Africa stemmed from the city of Ja, on the middle Niger. They were originally Soninke-speakers but they acquired their historical identity in the thirteenth century, when they migrated to the Upper Senegal valley and mixed with Manding-speakers. They created a tradition that, by dispersion and emulation, was to have tremendous impact in a vast swath of the savanna along the northern edge of the rainforest. From there, small groups of them arrived, known as Juula, to the Volta valleys, the area of greatest interest to me. This tradition combined literacy and Islamic scholarship, piety, political neutrality, life as small enclaves among non-Muslims, and the provision of services to them. It rested on an economy of farming and herding, conducted mostly by slave labor and supplemented with trade. Under their influence, learning centers arose in far-flung corners. They introduced Islam and later strengthened it even as far east as Kano (Levtzion 1968: 15, Sanneh 1979:28-31). Their presence brought literacy to places far from the hubs of mercantile activity or political authority; with them, literacy assumed a sort of central-place logic: while scholars of major reputation gravitated to famous centers (which added to their eclat), clerics of humbler learning filled up the areas in between, providing services to populations with fewer resources.
It is therefore important to shake the simplified view that Islam and literacy diffused from the Sahara contact zone in the north to the south in an uncomplicated manner, or that "black" Africa met Islam late and superficially as a foreign religion. Influences went in all directions. The view from the Western Volta region is that Muslim groups arrived first from the south (Malinke-Juula), and then from the north and west (Maraka), and then from the east, but following a southern route--Zaberma, mixed with Hausa elements that strengthened Islamic learning. An overlapping Islamic impact in the nineteenth century derived from Sheku Amadu's religious hegemony over the Masina, and later from the Tukulor followers of the jihad of Al Hajj Umar Tal. Early in the colonial period, Marty argued that the Muslims of Cote d'Ivoire were "more orthodox even than in North Africa" (Harrisson 1988:127).
Literacy was rehearsed on wooden tablets, like the ones much used today in traditional Qur'anic education. Writing paper was scarce and expensive, and the impermanence of the palimpsest was considered safer against the risk of defilement. One of the best illustrations of the continuing relevance of trans-Saharan trade with the Mediterranean, even after the Atlantic trade gained in importance, is that paper in the interior of West Africa continued to be supplied through this channel until quite late. In the eighteenth century, most paper available in Ashanti, for example, was not of north European make, but of Italian origin (Wilks, Levtzion, and Haight 1986:203).
Among the clerical Juula, as among settled Fulbe Muslims, most men and a significant number of women learned to read. In the nineteenth century, Binger ( 1980) found that in Kong and Bondoukou, few persons were unable to read Arabic. The same was true in Salaga (Wilks 1968:167). Even Kumase, a precolonial West African capital that the popular mind does not associate with Islam, had a well-established Muslim community in the eighteenth century. In 1820, the British consul in Kumase, Joseph Dupuis, could converse with its Muslims in Arabic. Political men, merchants, and dignitaries who could not write employed the services of a Muslim cleric. The Asantehene Osei Tutu Kwame, like his counterparts in the Sudanic zone, kept correspondence with scholars and other political figures (Wilks 2000:105). Ouagadougou--that other capital that colonial intrigue painted with remarkable longevity as a pagan stronghold--had by the end of the eighteenth century not only a Muslim community, but even a Muslim king, Naaba Dulugu (Skinner 1966:177). The late nineteenth century French explorers depicted the naabas of Ouagadougou completely under the sway of their Muslim counselors (Levtzion 1968: 163-170). Like their relatives and counterparts in Dagomba, they had an official imam in their court, and sent their sons to Qur'anic schools, where in the course of the nineteenth century several of them learned more than basic reading.
Muslims used Arabic writing to keep personal records, notes of birth, accounts, or to make travel itineraries, place orders for goods, and organize caravans (Goody 1968:209-210)--services that they also provided to others. Nineteenth-century accounts of European explorers make clear how letters of introduction from one religious or military leader to another insured safe passage and served for diplomacy (Binger  1980:1:331-332). These practices were crucial in zones such as the western Volta, where long-distance unstable alliances formed the basis of regional political life. The standard opening formula in the letters of the clerks expressed clearly the communicative role: "Praise to God, who gave us the paper as messenger and the pen as tongue."
Letters were also important in large-scale commerce. Specialists of precolonial West African economic life know that the highest-value goods, such as horses, slaves, and bales of kola nuts, were sold not in the markets, but out of major merchants' homes or warehouses, through personal connections. Much of it happened across large distances, suppliers with important quantities in stock sending letters of offer to potential buyers, who were few in number. Thus, we know that during the French colonial conquest of the western Volta, Widi of Barani, an ally of the French, tried to sell one lot of horses to the French enemy Samori--which caused a shock when the French seized the pertinent Arabic correspondence written in Widi's hand (Saul and Royer 2001:66).
With Islamic education came the love and possession of books. In the sixteenth century, Leo Africanus wrote that in Timbuktu, manuscripts brought from North Africa "sold for more money than any other merchandise" (quoted in Goody 1968:219). Libraries were not limited to the Sahel: in the beginning of the colonial period in northern Ghana, even small, unimpressive villages had people who possessed three or four books, mostly printed; and in Cote d'Ivoire, every cleric had a library, a few of which included up to two hundred works (Marty 1922:274-275).
Reading can be mastered at varying levels of skill, and the ability to compose is yet another matter. That Timbuktu and the cities of the Sahel had scholars producing books is commonly known, but at the other end of the region, in northern Ghana too, one found in addition to readers and copyists Arabic authors (see the list prepared by Goody and Wilks in Goody 1968:241-261 and Wilks, Levtzion, and Haight 1986). Although most of these works reveal poor Arabic lexicon and syntax, an accomplished scholar, such Al Hajj Umar, also lived in Salaga; he was a specialist of pre-Islamic Arabic poetry (which he translated into Hausa with an art that deeply impressed R. S. Rattray) and the author of a book on epistolary styles, later printed in Cairo (Goody 1968:220, 242).
Colonial observers dismissed West African Qur'anic schools because these at the primary level relied heavily on copying, unison repetition, and rote learning, and did not systematically teach Arabic language and grammar until an advanced stage, which few students reached; most local languages were not committed to script. This pedagogic tradition accords poorly with the Rousseauian spirit of our schools, and students provided a considerable amount of physical work in the master's farm and household, but we should reserve judgment as to teaching efficacy. As in other premodern settings, literacy was only one component in a broader education, which relied heavily on verbal rhetoric and memory. A central part of religious training in West Africa was reading aloud and, in local languages, closely translating, interpreting, and discussing passages from the Qur'an and other texts. These practices transformed the expressive possibilities in those languages, sharpened the sense of some of their words in a technical direction, facilitated finding and establishing translation equivalents between them, and, most importantly, made stories, themes, and modes of reasoning of Islamic origin available to non-Arabic speakers and non-Muslims--points documented for the contemporary period in the detailed observations of Tamari (1996, 2002).
Unexpected evidence for the proficiency of precolonial West African Islamic education is coming to light in pioneering research on Muslim slaves brought to America. A manual on the medical treatment of slaves from 1811 stated that many of those who came from Senegal could "converse in the Arabic language, . . . and some are sufficiently instructed even to write it" (quoted in Gomez 1994:695). We know the names of only a few dozen African Muslims who arrived in early North America, but some of these cases are highly instructive. Ayuba Suleiman Diallo was a Fula, born around 1700 and captured about the age of thirty somewhere south of the Gambia River. In Maryland, where he was brought, he composed in Arabic a plea for liberty, which so impressed a Royal African Company officer that Diallo was returned to his native country before two years had passed (Curtin 1967; Gomez 1994:690). Diallo's letters reveal a correct Maghribi script but hesitant Arabic, filled with stock phrases and quotes from the Qur'an--an education indicative of a man belonging to a clerical group but not destined to be a scholar. Umar ibn Said, another Pular-speaker, taken as an adult from his native Futa Toro (Senegal) later in the eighteenth century, had had more advanced education. His writings include several letters in Arabic, Qur'anic passages that he wrote from memory, sections from an Arabic Bible that was given to him in captivity, and the only extant autobiography in Arabic by a former slave in the United States, all penned in a beautiful handwriting (Austin 1997:129; Gomez 1994:688-689). His contemporary, Lamin Kaba, a Jakhanke scholar from Futa Jallon of mixed Sarakole and Maninka parentage, ended up in North Georgia. He left a description of his education in West Africa under male and female teachers, including an influential aunt "who was much more learned than himself," and dictated the names of thirty books, which had constituted his advanced curriculum. His story illustrates the hazards that attended scholarship; he had been captured while on an extended trip along the Atlantic coast, hundreds of miles away from home, trying to purchase paper (Austin 1994:115-125; Gomez 1994:691). Bilali, another Fula from Futa Jallon, was a plantation driver in Sapelo Island, off the coast of Georgia, and spiritual leader of the African Muslims who lived on the island (Austin 1997:85-111; Gomez 1994:695-699). It seems that he was able to teach Pular, and perhaps some Arabic, to his numerous children. He left behind a manuscript including rules about prayer, ritual ablution, assertions of the faith, and some enigmatic pages, which may have been produced in response to the needs of the local Muslim community or may have originated in Africa (Martin 1994).
The most compelling evidence is provided by Abu Bakr al-Siddiq, whom we have already met. He was captured and sold into slavery in Bouna when he was barely fourteen, and spent most of the rest of his life in servitude in Jamaica (Wilks 1967a). Three decades later, he met a magistrate who happened to know some Arabic and "was not a little surprised to see [Arabic] written with some neatness and great rapidity by a negro slave" (Renouard 1836:101). Abu Bakr carried out correspondence in Arabic with other Africans in the New World, and kept the accounts of his employer in English spelt in Arabic characters. He was manumitted and taken to England, where he made an impression with his knowledge of the Qur'an and his ability to chant the call to prayer "with the exact pronunciation, intonation, and rhythm that is used from Cairo to Constantinople, and from Belgrade to Delhi" (Renouard 1836:109-110). He wrote an autobiography in Arabic, and supplied a long list of placenames in West Africa, connecting them to one another. Austin (1997:21) remarks that these places are located in an area larger than the portion of the United States between the Mississippi River and the Atlantic Ocean, certainly more than what most of our fourteen-year-olds can manage.
An Excursus into Politics
The spectacular nineteenth-century political movements led by forceful Muslim leaders in West Africa--Usman dan Fodio in Sokoto, Seku Amadu in Masina (inner delta of the Niger), and Al-Hajj Umar Tall (Robinson 2000)--and the complex developments of the colonial period partly ensuing from this history have encouraged a simplified political view, according to which conflict generally involved Muslims on one side and non-Muslims on the other. This view is fundamentally wrong, for the longest stretch of West African history, and for most of its surface area, including zones said to be controlled by jihad leaders, up to the years of colonial conquest wars, that is the turn of the twentieth century (besides the fact that the violent reform movements also targeted primarily other self-professed Muslims). By reinforcing anthropologists' tendency to see Islam as a foreign incursion upon "traditionalists," this view inhibits an understanding of the cultural dynamics that made Islam such an important presence for Muslim and non-Muslim alike. Therefore, I sketch here some elements of political organization that were much more common in the region and render intelligible the interpenetration of ritual practices and discursive content associated initially with different communities.
Muslim dispersion in West Africa was generally premised on the existence of multiple nodes of power, most of which did not rest on territorial sovereignty. This is a political situation that strains our imagination and our analytic vocabulary, because it is so unlike the notions of the political world we inhabit. In much of the savanna, self-administering village groups of limited political scale maintained long-distance solidarities with powerful merchants, warlords, and other distant villages, often against their own neighbors (for a specific example, see Saul 1998). Translocal alliances resulted in unstable temporary balances within this heterogeneous political mix, capable of producing duel-like eruptions between the big players, but more typical were the economics of small-scale raid and banditry, favoring the well connected over the isolated. In this milieu, Muslims were generally neither rulers nor ruled: they constituted autonomous nuclei, perceived by others primarily as technical specialists, ready to offer their services, and valued and respected for that reason--another way of articulating the observation that Jahanke or Juula clerical families were "politically neutral." They often fitted in a place as perennial outsiders, the "guests" of local hosts, who themselves would be neither Muslim nor anti-Muslim, but simply pluralist. This historical situation is well described for northern Cote d'Ivoire by Launay (1992), who contrasts it to Muslim identity based on personal norms such as piety and observance, which has become common in the last few decades.
The unstable boundary among Muslims, "pagan allies," adherents, opponents, and hosts, is illustrated in the foundation traditions of the ancient town of Ja, already mentioned as the origin of the Jahanke-Juula phylum. The town had an unsurpassed reputation for Islamic learning, but its local traditions reveal it as a site of accommodation and cohabitation. It was established by four groups: two of these were Marka, but of distinct origins, the third consisted of Bozo fishermen, and the last was a warrior group going by the name Jawara (Sakai 1990). The Marka were Muslims, but the Jawaras' ancestor, who had arrived on horseback holding a flintlock gun, said: "I do not perform the ritual prayers." The four groups made a blood pact, which still serves as a kind of charter for the town. The first three were offered, in sequence, the rulership but they refused, and it fell upon the Jawara. The Jawara sought out clerical families, who took up residence in the town, but members of the clerical families say that the Jawara organized attacks, abused their rule, and were cruel, yet Ja never ruled any other place, and no memories of its fighting, or being subjugated to, the famed Gana or Songhay empires, exist. These attacks must have been, then, small-scale raids on nearby places. Although the town welcomed clerics, the fighters kept their assortment of power objects (boliw) housed in a structure in the middle of the town, and maintained a sacred grove incorporating a shrine. They made use of the amulets that the clerics confected, but they would not commit to Islam, because for them "any submission meant humiliation." The "king" of the fighters originally owed his occult strength to a jar of magic potion, but one of the imams broke it and put an end to that (Sakai 1990:234). Thus, the town remained an "entanglement of alliances," of groups that jealously guarded the memory of their origins, or, as the author puts it, "interdependent and interdistant."
This account is nearly identical with many others we find further south in the savanna, in places that do not have as high a reputation for Islamic learning. The interplay among keepers of local shrines, Muslim clerics, and an invited or self-imposed group of war specialists is common. In many districts and towns of the southern savanna under Malinke-Juula influence, we find an oligarchic administrative structure with periodic assemblies of dignitaries called jamaana, a likely derivative of Arabic jama'a, and Person (1968-1975:1:69) associates the practice with the democratic institutions of the Berbers of the Maghrib. The meaning of the words jamaana or kafu, and the social practices to which they referred, were an early casualty of colonial government, and the debris served to buttress the canton administration.
What is meant by oligarchy in this context can be rendered more precisely with the term heterarchy: the coexistence of various social units, which are not of the same nature, but among which a strict equivalence is recognized--a big man's house here, a village confederation there, farther up a man-of-arms of great renown living behind his fortified walls with his slave army, influential cleric-scholars, merchants turned military entrepreneurs, and we can throw in, if we wish to take account of late nineteenth-century realities, a French column made up of a few European officers with a rank and file made up of African tirailleurs, followed by a long trail of wives, mistresses, and hangers-on, carrying babies, equipment, cooking utensils, and food. All these groups, which interact, clash, and at times merge, are stratified internally; they appear like microcosms of the larger society that engulfs them. Our conventional tools of political analysis do not make it easy to theorize about this social pattern: not quite a democracy, nor really anarchy, but also the antipode of centralization and monarchy. I do not think that it was unique to West Africa, and as we develop a proper vocabulary, we shall discover its kindred in other places and ages, including perhaps in modern situations, but in West Africa, it served as an incubator for Islam and the legacy of cultural fluidity.
When Muslim clerics moved in this social landscape from the role of sidecast to that of main protagonists, political divisions were still not necessarily all drawn along religious lines. In the Muhun (Black Volta) valley in the mid-nineteenth century, Mahamudu Karantao started a movement described as jihad, but many of his partisans were non-Muslim Ko villagers. His enemies, in turn, were supported by the famous Muslim clerics of the towns of Da and Jinakongo (Saul and Royer 2001:53-60). A center of opposition to him was a Marka league, guided by a set of prestigious fighter families holding personal and public shrines with a reputation that matched their fighting skills, yet the early colonial descendents of these people carried names such as Yisu, Isaka, Davuda, which betray Muslim inspiration. In 1915, members of this league became leaders of a war against the colonial government, and some of them went to agitate in the neighboring province of Kudugu. Without compunction, they pretended to be Muslims, to impress their newfound allies. Meanwhile, the Karantao supported the colonial government, yet some non-Muslim villages fighting against the colonial government appealed to them to obtain war charms, because the Karantao were their old allies. These intricacies are crucial to understand this war, but none of it will make sense if one assumes that Muslims constituted a unified political (and ipso facto cultural) front, that they were all on one side or the other side: every side to every conflict had its own Muslims.
Where public authority was constituted around a center defined by a ruler, Muslim presence was felt in the shared social space, irrespective of how the politically powerful presented themselves in religious terms. We understand this better if we leave behind the premise that religion proceeds from faith. Peel (2000:195-205) presents a thought-provoking discussion of how, among nineteenth-century Yoruba, participation in public ritual, the use of private services, or even adopting the clothing style, can bring people closer to identifying with Muslims, without effecting a change in their inner state. A leap into religious conviction is, however, a possible outcome, impelling a break with the past. In any case, joint or overlapping praxis should be kept distinct from common faith, but not underestimated as a mode of behavior with its own social force. In Kumasi, Ouagadougou, or Bouna, leaders used the divination powers and war charms of Muslim clerics, as did other war specialists in West Africa. They participated in Muslim holidays, such as Id al-Adha, Id al-Fitr, Ashura, and the birthday of the prophet, turning them into public festivities for all (Boutillier 1993:350; Skinner 1966; Wilks 1966). Conversely, imams participated in initiation ceremonies in Senegambia (Mark 1992:140), or provided protection for trade caravans (Sanneh 1979:189).
Although Muslim clerics' thaumaturgic abilities were a major reason of their allure, diplomacy, letter-writing, and counseling services also made them desirable. They administered oaths and ratified important agreements. For example, in most places the administration of oaths that involved the absorption of daga was known as a Muslim specialty. Daga are balls of a stiff dough made with millet flour and water and left to rest overnight for a slight fermentation. They are normally consumed as a thick liquid after the balls are mashed in water. When prepared for an oath, a Muslim specialist kneaded the dough while reciting prayers over it, and the parties took the gruel in front of witnesses, with a copy of the Qur'an and other Islamic paraphernalia adding to the solemnity of the occasion. The administering cleric was sometimes the only Muslim in the scene, somewhat like a European notary public certifying a contract, the only member of the legal profession in a roomful of participants, with the important difference that in the dege oath, the spiritual sanctions warranted by the cleric (not the state) were the principal force making the agreement binding.
The prayers, charms, and other clerical services were prohibitively costly. A manuscript in Hausa reports that when the Zaberma leader Babatu asked his malams to help him in his projected wars (1880s), they demanded a payment of "one hundred head of cattle, one hundred sheep, one hundred goats, one million and one thousand cowries, and one hundred slaves" (Pilaszewicz 1992:72)--certainly not a price many political actors could afford! When the denizens of farming villages appeared in the battle-field guarded only with ancestor and shrine medicines, against adversaries endowed with Islamic furnishings, it was because they could not avail themselves of the latter's state-of-the-art mystical technology, rather than a preference for "resisting" foreign cultural intrusion. They had to fall back on homegrown equivalents, one reason lying behind the mimetic adoptions, which are the topic of the next section.
In West Africa, adherence or conversion to Islam has not been a unidirectional process--perhaps the principal way in which it historically differs from world areas more solidly associated with Islam in our mind. The central shrine of one of the Zara houses of Bobo-Dioulasso, who are mostly practicing Muslims now but were not so before 1925, is said to have been built over the manuscript Qur'an of their ancestor who also bears a Muslim name. Levtzion (1968: 142-144) and Bravmann (1974: 128) report similar cases in northern Ghana and western Cote d'Ivoire. From the village of Nemena (Kossi province, Burkina Faso), I possess a video recording of a mock performance of Muslim ablution and prayer during a funeral. Those who performed it, Bwamu-speakers, were members of families that believe that their ancestors were Muslims, and maintain this memory as a distinctive part of their identity, although otherwise they are part of a mass of non-Muslim farmers. Because of this eventuality of past adherence to Islam, Goody (1968:216) maintained that no group in West Africa could be exhaustively studied without taking into account intellectual links to Islam, and through it to the Graeco-Roman and Semitic civilizations of the Mediterranean.
One implication is that we can err by projecting to the past oppositions or negotiations of religious belonging as currently observed. This is what Shaw (2002:82-88) presents in another example from the mid-Atlantic coast. People keep debating the Islamic or non-Islamic origin of certain practices, in which their own collective identities now hang. From a historical perspective, both sides of the divide are saturated with Islamic elements. Shaw calls the local discussions a two-hundred-year-old "argument"; thus, she brings attention to the forest of framing historical experience, which we should not lose sight of for the trees of the boundary-setting discourse, this one contingent upon current conditions. There is no pagan essence that is set ontologically against Muslim essence.
Another example of "Islam on both sides" is provided by the traditions of "Mande" in the upper Niger. The story of Sun-Jara (or Sun-Jata) and the septennial renewal ceremonies of the sanctuary of Kamablon in southern Mali are the subject of many publications (Dieterlen 1955, 1959; Ganay 1995; Meillassoux 1968), but among the pioneers, only an African researcher (Niane1975 ) mentioned the Islamic connections, perhaps even origin, of this sanctuary. The summit of absurdity is reached when in the course of a story about a certain "Arama" (which, according to a footnote, is also pronounced Adama) and his companion "Awa," the original couple of the world, we are provided with Malinke etymologies for these names (Ganay 1995:155, 159). How much of this was willful dissimulation, one unavoidably asks oneself, because is there a European who would not recognize Adam and Eve? The Kamablon ceremonies start and end with declamations from the Qur'an, and their centerpiece, the Sun-Jara recitation, is shot through and through with references to Mecca and the Qur'an, as was pointed out in the pioneering article of Conrad (1985). One could try to sort out what in these traditions is "West Sudanic" or "Manding"--in the sense of "extra-Islamic"--but that would miss another significant point, the cultural context in which they took their present shape. Can we reconstruct a pre-Islamic traditional baseline out of them? Jansen's (2002) critical history of this field refreshingly stresses the recent layer of Islamic influences. Enough information suggests further that the Islamic elements in the Kamablon complex are constitutive, not superficial accretions. It would be trite to attempt weeding them out in the hope of arriving at an ur-culture, as were the nineteenth century philological efforts in Europe to recover pre-Christian cultural monuments out of the fragments recorded among the peasantry.
In this section, I elaborate on conflict with Muslims as one possible path to Islamic influence. Gibbal observed that among the followers of the Ghimbala cult, of northeast Mali, water-spirits appear as a counterbalance to Islam, but the practice is infused with conscious references to Islam. Every ceremony starts with bi 'ism Allah and ends with la ilaha illa Allah, sacrifices are made facing the direction of Mecca, many genii are said to have come from Medina, and so on (Gibbal 1994:138). The devotees of bori in the Mawri community of Niger studied by Masquelier feel threatened by the advances of Islam, but they forged techniques "borrowing mostly from colonial and Islamic repertoires of signifiers--for instance, building mosques for their spirits, wearing French army uniforms, or positioning their bodies as if they were sitting at desks" (2001:292, 296). A pity not much is made of this observation. The study would have gained a different historical dimension if it had explored the two sides of the debate--asking, for example, why this mimesis now does not satisfy the young, who are converting in large numbers to observant Islam--instead of ratifying the testimony of a small number of spirit mediums nostalgic about the way things were thirty years ago and setting up what the reviewer of the book in this journal called a "starkly Manichean opposition of bori to Islam" (Roberts 2003: 138).
Mimesis is an expressive act, but often enough it draws its energy from the desire to make one's own the elan of a powerful other. Kramer (1993) interprets the adoption of European and Islamic elements in African ritual as an inversion that disvalues the imitated. When feelings remain inchoate, this may well be one dimension of the phenomenon, but we should not disregard the awe and fascination that tend to go along with emulation. From what Bayart (1993) called "extraversion" follows a quest for techniques that promise efficacy over the world, uninhibited by prejudice. Shaw and Stewart (1994) draw attention to antisyncretism, the attitude of erecting a hard boundary against another religion. We should still make a distinction between people's attitude and the cultural content they are wielding. The content may be more variable than the boundary itself. (Think of Hanukkah as alternative Christmas!) Antisyncretism may be no obstacle to the desire to enrich the reservoir of the self by enthusiastic copying of elements detached from what lies beyond the boundary. Thus, imaginative assortments come into existence with items emulated from Islamic, Christian, or medical scientific communities, without the original meanings attached, yet unanticipated effects are not precluded. The mind is not a closed system and, as copraxis, imitative praxis too can lead the way to unsuspected possibilities. Skinner (1966:183) provides the telling example of a non-Muslim diviner who, following the logic of his own procedures, prescribes for a patient conversion to Islam.
First let us consider single traits, like the triangle in ornamental design, which was assumed because it evokes Muslim protective power, or the fez, the headdress of the Ottoman administrative reform, which making its way down from Fezzan through the Hausa corridor became among the Dagomba and Mamprusi the badge of chiefly office, as turbans, embroidered robes, and baggy trousers puckered at the waist were earlier. The fez then entered the ritual paraphernalia of Dagara farmers (Goody 1962:70). Currently in Ghana, the robes, the baggy trousers, and the fez are redefined as traditional, and politicized to assert ethnic belonging as opposed to Akan and coastal identity.
A more complex case is the development of entire para-Islamic ritual complexes. In my opinion, the jo societies of the Bamana-Senufo area are of this sort. The profusion of Islamic references in the famous Komo-Kono societies, the assertion that they have been brought from Mecca, the fact that their shrine houses are called misiri "mosque," and the parallels to and inversions of Islam saturating their ritual practices have perplexed observers (Henry 1910:130-134). "The founder of African magic," the famous Makanta Jigi, brought his esoteric knowledge from Mecca. In one version of the legend, he decides to travel to Mecca as penance for incest. There the Prophet introduces him to the ways of Islam. Later, Jigi discovers the chief of the magicians of Mecca, and lured by the latter's ritual riches, repudiates his allegiance to the Prophet. He returns to Mali with the goods provided by the magician, loaded onto forty-seven camels (Traore 1947). The name of this hero in fact means literally "Jigi who went to Mecca."
Moraes Farias (1989) interprets these Meccan origin legends as charts by non-Muslims for cohabitation and religious plurality. Taken together with the practices of the jo societies, which the Muslims invariably loathed, it is hard to avoid the conclusion, however, that at some level of discourse they also had antagonistic intent. Some of these jo societies can be seen as anti-Islams, built with the very elements of perceived Islam, by emulation and inversion (Saul 1997:17-19).
The links between the spectacular masquerade traditions of West Africa and Islam bring up a different issue. Bravmann (1974) stated first that West African masking traditions could be compatible with Islam, and the idea was elaborated by Mark (1992) and Prussin (1986). Masks are adorned with Muslim prayers (Mark 1992:132, 138; Tamari 2001:100), or triangles and amulets, and masquerades are associated with Muslim celebrations (Bravmann 1974:166-177, 1983:69; Launay 1992:111). Green (1987) noted that the word do is associated with masking in many places in northern Cote d'Ivoire and in Burkina Faso, corresponding to the zone of activity of sonangi warrior groups, who patronized Islamic clerics without being exclusively followers of Islam. Do exists in association with masking in the Black Volta region, among farming groups that are in majority decidedly non-Muslim, described in greatest detail for the Bobo and the Bwa (Capron 1957; Le Moal 1980), but found also among Tussian, Tyefo, and other peoples.
In Bobo-Dioulasso, some actors I interviewed asserted not only that masking is compatible with Islam, but that local Muslims invented masking and diffused certain styles of masks to places where they had not previously existed (Saul 1997). The literal historical accuracy of such assertions is beside the point, but what are we to make of them in a broader interpretive framework? I turn now to Morocco. I discover masking traditions that have numerous similarities to those encountered in West Africa: in morphology (disguise materials attached to the naked body, making mask and carrier inseparable), in function (appropriation by youth groups), and in spirit (secrecy of preparations, playfulness of the performance, ritual efficacy). In Morocco, these masks are specifically associated with Islamic holidays (Hammoudi 1993); in the western Volta, only those of Islamic groups are, and the others follow the seasonal agricultural calendar.
An earlier generation of French scholars had seen in the North African masquerades "pagan survivals," or even an older religion that had gone underground and still competes with Islam. This interpretation sounds like the reverse of the one made for sub-Saharan Africa: in the North African case, Islam was declared central, and the folkloric elements that seemed incongruent were declared foreign (antique paganism); in the sub-Saharan African case, a sort of paganism (animism, fetishism) is declared central, whereas the Islamic elements are ignored or declared foreign. Hammoudi remarks that discussing the masks in Morocco primarily as survival from pre-Islamic times distorts their present meaning, their intrinsic tie to Muslim holidays. For me, the unsuspected affinity between West African masking traditions and those of North Africa opens a new historical vista. The Moroccan festivals are indubitably part of the set of calendar celebrations of peasant groups found all around the Mediterranean rim countries, such as the carnival tradition. These speculations do not take away anything from the dramatic achievement of West African mask celebrations; my purpose is simply to bring them down from the ethereal skies of exotic weirdness to the historical home base of our common humanity.
Are these similarities between North Africa and Western Sudan to be attributed to adoption in times within the framework discussed here (and mediated by the actions of Muslims)? or should we think of them as belonging to an earlier cultural matrix, in which both regions participated (antedating Islam)? Note that not all features that Islam and West African conceptions share are likely to have resulted from contact in the period under review--not, for example, the jinn spirits in Islam, nor the otiose high god of West African cosmogonies. The issue can perhaps be clarified with more substantive research, but it is not vital at all for the argument I present here. Either way, separating West Africa from the circum-Mediterranean area has distorted our understanding of both North and West Africa.
West Africa's historical individuality is lost in the mental map of scattered ethnic labels, which in anthropology still keeps us under its sway in subtle but unmistakable ways. Placing West Africa's past connections to the Mediterranean world and its Islamic legacy right at the center of our conception of Africa is crucial to overcome this microcosmic vision and broaden our horizon. We need to free ourselves from the petrified "pre" of tradition, understand regional flows that stand intermediate between the local and the global, and establish greater contact with the present conditions of West Africa without escapist romanticism. This freedom, among other things, may also lead to new research questions.
A topic that deserves more sustained theoretical attention by anthropologists is the rise of massive tides of conversions to Islam in places where previously Muslims constituted only small minorities. A related but distinct question is the growth of "inner conversions," the popularity of the observant and pious variety of Islam, and the clamorous calls for "de-ethnicizing" Islam. At one level, this development demonstrates the contingency of social boundaries, that culture and even language are, in the larger scheme of things, fleeting realities. Changes of religion and language become gateways to new worlds, a concomitant of other transformations, which are occurring simultaneously, but what is the connection between the specifics of these processes and the meshwork of signs inherited from past centuries?
Responses to changed circumstances may be rooted in history in intricate ways, which will not be articulated by the actors. Take, for example, the reaction to colonial domination. Triaud (1997:17) proposes that the presence of small, "encysted" communities of Muslims, close to power-holders as counselors and technical specialists but politically subordinate, prepared many people in West Africa to experiencing colonial occupation as one more misery of this finite world, and not as the eschatological disaster that sympathetic outside observers retrospectively suppose it was. Stewart (1990) suggests that when colonial governments took over the judicial and legal functions of Muslim clerics, space was opened for Muslim leaders who focused their message on mystical experience, and for the Sufi brotherhoods they promoted, which became the prime movers in the twentieth-century expansion of Islam. These are broad perspectives, which could inform anthropological research in different kinds of settings, and no doubt we could find other, equally stimulating ideas in the literature that deals with the intersection of Islam and non-Islam in West Africa.
The larger point is that we need to relocate West Africa in our imagination. This plea goes to the heart of the we-they contrast, which underlies much anthropological discourse. We absorbed a hardened version of this opposition that blocks the way to an empathic connection with our interlocutors and a true curiosity in their mundane lives. Reinserting West Africa in the nexus of Eurasian history would make it less alien. We need to reach a new cosmopolitanism, which would enable us to take in Islam as part of our common heritage and West Africa as part of the Islamic heritage. What Mudimbe (1994:62) called "ethnologization and aestheticization" are largely failures rooted in the incommensurability of the flat conception of African cultures with the stratigraphic verticality of "our" history. Mudimbe is an eminent scholar and author, but he had little consideration for Islam in African cultural history. We need to take his advice and go further. Yes, anthropology should not be the science of the obscure and the arcane, of salvage from the last pagan ritualist, underground possession rites, secrets that exist mostly in the regret of their being lost, or wishful analyses of resistance through magic--ignoring that presidents or diamond tycoons may be its most successful manipulators. We need to be wary of the magnifying glass that distorts difference into strangeness, and go back to the public square, the market, the Friday mosque, and the university classroom. A more resolute reorientation requires attention to the elements of the West African cultural substance, not only their social and communicative uses, and to the centuries of interface with neighboring regions that deposited them. It is worth giving thought to whether we want to shift what we read with priority and what we assign in our classes, how in these decisions we practically reform ourselves as we form others, either to perpetuate or to break the anthropological habit of disregard for West Africa's historical ties to the Mediterranean through Islam.
This article started its life as a distinguished lecture invited by the Association for Africanist Anthropology to the 103rd annual meeting of the American Anthropological Association in Atlanta, Georgia, in November 2004. I extend my thanks to the president of the association, Gracia Clark, its awards chair, Gwendolyn Mikell, its president-elect, Bennetta Jules-Rosette, and its other officers, for the honor and the opportunity. Parts of the lecture were presented in Indiana University, and in a panel of the 48th annual meeting of the African Studies Association in Washington, D.C. (in November 2005), where Michael Lambek served as commentator. On all these occasions, a lively discussion followed.This version incorporates the responses stimulated by these reactions and the critical commentaries of three anonymous reviewers forAfrica Today. I express my gratitude for all of them.
Al-Nagar, 'Umar. 1972. The Pilgrimage Tradition in West Africa: An Historical Study with Special Reference to the Nineteenth Century. Khartoum: Khartoum University Press.
Amselle, Jean-Loup. 1996. Vers un multiculturalisme frangais: A empire de la coutume. Paris: Flammarion.
Appadurai, Arjun. 1996. The Production of Locality. In Modernity at Large, pp. 179-99. Minneapolis: University of Minnesota Press.
Apter, Andrew. 1992. Black Critics and Kings: The Hermeneutics of Power in Yoruba Society. Chicago: University of Chicago Press.
Austin, Allan D.1997. African Muslims in Antebellum America: Transatlantic Stories and Spiritual Struggles. 2nd ed., rev. and condensed. New York: Routledge.
Barth, Henry [Heinrich]. 1857-1858. Travels and Discoveries in North and Central Africa. 5 vols. London: Longman, Brown, Green, Longmans, & Roberts.
Bayart, Jean-Fancois. 1993. The State in Africa: The Politics of the Belly. New York: Long man.
Beek, Walter E. A. van. 2004. "Haunting Griaule: Experiences from the Study of the Dogon." History in Africa 31:43-68.
Berliner, David. 2005. An "Impossible" Transmission: Youth Religious Memories in Guinea-Conakry. American Ethnologist 32(4):576-592.
Bianchi, Robert R. 2006. Guests of God: Pilgrimage and Politics in the Islamic World. Oxford: Oxford University Press.
Binger, Gustave.  1980. Du Nigerau Golfe de Guinee. Paris: Societe des Africanistes.
Birks, J. S. 1978. Across the Savannas to Mecca: The Overland Pilgrimage Route from West Africa. London: C. Hurst & Company.
Bledsoe, Caroline. 1986. Arabic Literacy and Secrecy among the Mende of Sierra Leone. Man n.s. 21(2):202-226.
Boutillier, Jean-Louis. 1993. Bouna: royaume de la savanna ivoirienne: princes, marchands et paysans. Paris: Kathala-ORSTOM.
Bravmann, Rene A. 1974. Islam and TribalArtin WestAfrica. Cambridge: Cambridge University Press.
--. 1983. African Islam. Washington, D.C.: Smithsonian Institution Press. Brenner, Louis. 1984. West African Sufi: The Religious Heritage and Spiritual Quest of Cemo Bokar Saalif Taal. Berkeley: University of California Press.
--. 2000. Histories of Religion in Africa. Journal of Religion in Africa 30(2):143-167.
Caillie, Rene.  1979. Voyage & Tombouktou. 2 vols. Paris: Editions La Decouverte.
Capron, Jean. 1957. Quelques notes sur la Societe du Do chez les populations Bwa du cercle de San. Journal de la Societe des Africanistes 27(1):81-129.
Chazan, Naomi, and Michel Abitbol. 1988. Myths and Politics in Precolonial Africa. In The Early State in African Perspective. Edited by S. N. Eisenstadt, M. Abitbol, and N. Chazan, pp. 28-59. Leiden: E. J. Brill.
Clapperton, Hugh.  1966. Journal of a Second Expedition into the Interior of Africa. London: Publisher.
Conklin, Alice. 1997. A Mission to Civilize: The Republican Idea of Empire in France and West Africa. Stanford: Stanford University Press.
Conrad, David C. 1985. Islam in the Oral Traditions of Mali: Bilali and Surakata. Journal of African History 26(1):33-49.
Copans, Jean. 1980. Les marabouts de l'arachide: la confrerie mouride et les paysans du Senegal. Paris: Le Sycomore.
Cruise O'Brien, Donald B. 1971. The Mourides of Senegal. The Political and Economic Organization of an Islamic Brotherhood. Oxford: Clarendon.
Curtin, Philip D. 1967. Ayuba Suleiman Diallo of Bondu. In Africa Remembered: Narratives by West Africans from the Era of theSlave Trade. Edited by P. D. Curtin, pp. 17-59. Madison: University of Wisconsin Press.
Dieterlen, Germaine. 1955. Mythe et organization sociale au Soudan francais. Journal de la Societe des Africanistes 25:38-76.
--. 1959. Mythe et organization sociale au Soudan francais. Journal de la Societe des Africanistes 29:119-138.
Dunn, Ross E. 1986. The Adventures of Ibn Battuta: A Muslim Traveler of the 14th Century. Berkeley: University of California Press.
Eaton, Richard M. 1993. Islamic History as Global History. In Islamic and European Expansion: The Forging of a Global Order. Edited by M. Addas, pp.1-37. Philadelphia: Temple University Press.
Eickelman, Dale F., and James Piscatori, eds. 1990. Muslim Travelers: Pilgrimage, Migration, and the Religious Imagination. Berkeley: University of California Press.
Ganay, Solange de. 1995. Le sanctuaire Kama blon de Kangaba: histoire, mythes, peintures parietals et ceremoniesseptennales. Paris: Nouvelles du Sud.
Gibbal, Jean-Marie. 1994. Genii of the RiverNiger. Chicago: University of Chicago Press.
Gomez, Michael A. 1994. Muslims in Early America. The Journal of Southern History 60(4):671-710.
Goody, Jack. 1962. Death, Property and the Ancestors: A Study of the Mortuary Customs of the Lodagaa of West Africa. London: Tavistock.
--. 1968. Restricted Literacy in Northern Ghana. In Literacy in Traditional Societies. Edited by J. Goody. Cambridge: Cambridge University Press.
--. 1971. The Impact of Islamic Writing on the Oral Cultures of West Africa. Cahiers d'Etudes africaines 11(43):455-466.
--, and Ian Watt. 1968. The Consequences of Literacy. In Literacy in Traditional Societies. Edited by J. Goody. Cambridge: Cambridge University Press.
Green, Kathryn. 1987. Shared Masking Traditions in Northeastern Ivory Coast. African Arts 20:62-69. Greenberg, Joseph. 1946. The Influence of Islam on a Sudanese Religion. Seattle: University of Washington Press.
Grosz-Ngate, Maria. 2006. Du terrain au texte: reflexions anthropologiques sur Voyages et decouvertes dans l'Afrique septentrionale et centrale, 1849-1855. In Heinrich Barth et l'Afrique. Edited by M. Diawara, P F. de Moraes Farias, and G. Spittler. Cologne: Rudiger Koppe Verlag.
Hammoudi, Abdellah. 1993. The Victim and Its Masks: An Essay on Sacrifice and Masquerade in the Maghrib. Chicago: University of Chicago Press.
Harrison, Christopher. 1988. France and Islam in West Africa, 1860-1960. Cambridge: Cambridge University Press.
Henry, Jos. 1910. L'Ame dun people africain: les Bambara. Munster: Aschendorffschen Buchhandlung.
Hunwick, John O. 1999. Timbuktu and the Songhay Empire: Al-Sadi's Tarikh al-Sudan Down to 1613 and Other Contemporary Documents. Leiden: E. J. Brill.
Jansen, Jan. 2002. Les Secrets du Manding: les recits du sanctuaire Kamabolon de Kangaba (Mali). Leiden: Research School of Asian, African and Amerindian Studies.
Kramer, Fritz. 1993. The Red Fez: Art and Spirit Possession in Africa. Translated by Malcolm Green. London: Verso.
Launay, Robert. 1992. Beyond the Stream: Islam and Society in a West African Town. Berkeley: University of California Press.
Le Moal, Guy. 1980. Les Bobo: nature et fonction des masques. Paris: ORSTOM.
Levtzion, Nehemia. 1968. Muslims and Chiefs in West Africa: AStudy of Islam in the Middle Volta Basin in the Pre-colonial Period. Oxford: Clarendon Press.
--. 2000. Islam in the Bilad al-Sudan to 1800. In The History of Islam in Africa. Edited by Nehemia Levtzion and R. L. Pouwels, pp. 63-91. Athens, Ohio: Ohio University Press.
Levtzion, Nehemia and J. F. P Hopkins. 1981. Corpus of Early Arabic Sources for West African History. Cambridge: Cambridge University Press.
Levtzion, Nehemia and R. L. Pouwels, eds. 2000. The History of Islam in Africa. Athens, Ohio: Ohio University Press.
Lloyd, P C. 1965. The Yoruba of Nigeria. In Peoples of Africa. Edited by J. L. Gibbs Jr., pp. 547-582. New York: Holt, Rinehartand Winston.
Manning, Patrick. 1998. Francophone Sub-Saharan Africa, 1880-1995. 2nd ed. Cambridge: Cambridge University Press.
Mark, Peter. 1974. Africans in European Eyes: The Portrayal of Black Africans in Fourteenth and Fifteenth Century Europe. Syracuse, N.Y: Publishing Desk, Maxwell School, Syracuse University.
--. 1992. The Wild Bull and the Sacred Forest: Form, Meaning and Change in Senegambian Initiation Masks. Cambridge: Cambridge University Press.
Martin, B. G. 1994. Sapelo Island's Arabic Document ... in Context. Georgia Historical Quartely 78: 589-601.
Marty, Paul. 1922. Etudes surl'Islam en Coted'Ivoire. Paris: Leroux.
Masquelier, Adeline. 2001. PrayerHas Spoiled Everything. Durham: Duke University Press.
McCloud, Aminah Beverly. 1995. African American Islam. New York: Routledge.
McNeill, William H. 1963. The Rise of the West. A History of the Human Community. Chicago: University of Chicago Press.
Meillassoux, Claude. 1968. Les ceremonies septennales du Kamablo de Kaaba (Mali) (5-12 avril 1968). Journal de la Society des Africanistes 38(2):173-183.
Moraes Farias, P F. de. 1989. Pilgrimages to "Pagan" Mecca in Mandenka Stories of Origin Reported from Mali and Guinea-Conakry. In Discourse and Its Disguises. Edited by K. Barber and P. F. de Moraes Farias, pp. 152-70. Birmingham: Center of West African Studies, University of Birmingham.
ArabicMedievallnscriptionsfromtheRepublicofMali:Epigraphy,Chronicles,and Songhay-TuaregHistory.Oxford:Oxford University Press.
Morton-Williams, P 1966. Two Studies of If a Divination. Africa 36(4):406-408.
Mudimbe, V. Y 1994. The Idea of Africa. Bloomington: Indiana University Press.
Nadel, S. F. 1946. A Black Byzantium: The Kingdom of Nupe in Nigeria. London: Oxford University Press.
Niane, Djibril TJ 975 . Recherches sur l'Empire de Mali au Moyen Age suivi de Mise en place des populations de la Haute-Guinee. Paris: Presence Africaine.
Ong, Walter J.. 1982. Orality and Literacy. The Technologizing of the Word. London: Methuen & Co. Peel, J. D. Y 2000. Religious Encounter and the Making of the Yoruba. Bloomington: Indiana University Press.
Person, Yves. 1968-1975. Samori: une revolution dyula. 3 vols. Dakar: IFAN.
Pilaszewicz, Stanislaw. 1992. The Zaberma Conquest of Northwest Ghana and Upper Volta: A Hausa Narrative: "Histories of Samory and Babatu and Others "by Mallam Abu. Warsaw: Polish Scientific Publishers.
Prussin, Labelle. 1986. Hatumere: Islamic Design in West Africa. Berkeley: University of California Press.
Reichmuth, Stefan. 2000. Islamic Education and Scholarship in Sub-Saharan Africa. In The History of Islam in Africa. Edited by Nehemia Levtzion and R. L. Pouwels. Athens, Ohio: Ohio University Press.
Renouard, G. C. 1836. Routes in North Africa, by Abu Bekr es Siddik. Journal of the Royal Geographical Society 6:100-113.
Robinson, David. 2000. Revolutions in the Western Sudan. In The History of Islam in Africa. Edited by Nehemia Levtzion and R. L. Pouwels. Athens, Ohio: Ohio University Press.
Sakai, Shinzo. 1990. Traditions orales a Ja: histoire et ideology dans une ancienne cite islamique. In Boucle du Niger: approches multidisciplinaires, vol. 2. Edited by J. Kawada. Tokyo: Institut de Recherches sur les Langues et Cultures d Asie et d Afrique.
Sanneh, Lamin O. 1979. The Jakhanke. London: International African Institute.
Saul, Mahir. 1984. The Quranic School Farm and Child Labour in Upper Volta. Africa 54(2):71-87.
--. 1997. Islam et appropriation mimetique comme ressource historique de la religion bobo. Journal des Africanistes 67(2):7-24.
--. 1998. The War Houses of the Watara in West Africa. International Journal of African Historical Studies 31(3):537-570.
Saul, Mahir and P Royer. 2001. West African Challenge to Empire: Culture and History in the Volta-Banff Anticolonial War. Athens, Ohio: Ohio University Press.
Shaw, Rosalind. 2002. Memories of the Slave Trade: Ritual and the Historical Imagination in Sierra Leone. Chicago: University of Chicago Press.
Shaw, Rosalind and Charles Stewart. 1994. Introduction: Problematizing Syncretism. In Syncretism/ Anti-Syncretism: The Politics of Religious Synthesis. Edited by Charles Stewartand Rosalind Shaw. London: Routledge.
Skinner, Elliott P. 1966. Islam in Mossi Society. In Islam in Tropical Africa. Edited by I. M. Lewis, pp. 173-193. Bloomington: Indiana University Press.
Soares, Benjamin F. 2005. Islam and the Prayer Economy. Economy, History, and Authority in a Malian Town. Ann Arbor: University of Michigan Press.
Steiner, Christopher B. 1994. African Art in Transit. Cambridge: Cambridge University Press. Stewart, C. C. 1976. Southern Saharan Scholarship and the Bilad al-Sudan. Journal of African History 17(1):73-93.
--. 1990. Islam. In The Colonial Moment in Africa. The Cambridge History of Africa, 7. Edited by A. Roberts. Cambridge: Cambridge University Press.
Tamari, Tal. 1996. Lexegese coranique (tafsir) en milieu mandingue: rapport preliminaire sur une recherche en cours. Islam et Societe au Sud du Sahara 10:43-79.
--. 2001. Notes sur la representation cosmogonique dogon, bambara et malinke et leurs paralleles avec la pensee antique et islamique. Journal des Africanistes 71(1):93-111.
--. 2002. Islamic Higher Education in West Africa: Some Examples from Mali. In Islam in Africa. Edited by T. Bierschenkand G. Stauth. Munster: Lit Verlag.
Traore, Dominique. 1947. Makanta Djigi, fondateur de la magie soudanaise. NotesAfricaines 35(July): 23-25.
Triaud, Jean-Louis. 1997. Introduction. In Le temps des marabouts: itineraires etstrategies islamiques en Afrique occidentale frangaise v. 1880-1960. Edited by David Robinson and Jean-Louis Triaud. Paris: Karthala.
--. 2000. Islam in Africa under French Colonial Rule. In The History of Islam in Africa. Edited by Nehemia Levtzion and R. L. Pouwels. Athens, Ohio: Ohio University Press.
Wilks, Ivor. 1966. The Position of Muslims in Metropolitan Ashanti in the Early Nineteenth Century. In Islam in Tropical Africa. Edited by I. M. Lewis. Bloomington: Indiana University Press.
--.1967a.AbuBakraI-Siddig of Timbuktu. In Africa Remembered: Narratives by West Africans from the Era of the Slave Trade. Edited by P D. Curtin. Madison: University of Wisconsin Press.
--. 1967b. Wargee of Astrakhan. In Africa Remembered: Narratives by West Africans from the Era of the Slave Trade. Edited by P. D. Curtin. Madison: University of Wisconsin Press.
--. 1968. The Transmission of Islamic Learning in the Western Sudan. In Literacy in Traditional Societies. Cambridge: Cambridge University Press.
--. 2000. The Juula and the Expansion of Islam into the Forest. In The History of Islam in Africa. Edited by N. Levtzion and R.L. Pouwels. Athens, OH: Ohio University Press.
--, Nehemia Levtzion, and Bruce M. Haight. 1986. Chronicles from Gonja. A Tradition of West African Muslim Historiography. Cambridge: Cambridge University Press.
Young, Tom. 2003. Introduction. In Readings in African Politics. Edited by T. Young. Bloomington: Indiana University Press.
(1.) I see no need here to review the vast literature by historians on the Islamic past of West Africa, but for the record, in addition to the authors mentioned in the text, who are mostly relevant to the Volta region, let me mention Jean-Louis Triaud and the journal Islam et Societes au Sud du Sahara, David Robinson, J. O. Hunwick, H. J. Fisher, Murray Last, Lamine Sanneh, and Louis Brenner.
(2.) In Arabic, the names are al-ahad, al-itnain, al-talata, al-arbia, al-khamis, al juma, as-sabt. Except for juma, these are simply the ordinal numbers, which qualify the word yawm 'day'. The naming of the days of the week is similar in Hebrew.