Agisymba and the Ngizim?

 Whilst continuing our early forays into the Teda, Daza and Kanuri languages, we randomly wondered if Agisymba was perhaps a reference to the Ngizim. Lange and others have mentioned this, although Murray Last made the point more explicit through reference to the Teda and Daza languages. If the Garamantes referred to the distant land of Agisymba by a name heard from people speaking an early form of the Teda language (possible due to the Tibetsi region likely being in contact with the Garamantes in some form), it is plausible for the Ngizim to have been the ethnonym in question. Basically, the addition of -ba at the end of word would have meant "people of" (as in the -bu in Kanembu, although modern Kanuri seems to use -wu). So Agisymba would have meant A (perhaps the -ng initial sound was difficult to render in the Berber and Greek languages) + ngisym + -ba. 

This theory, of course, assumes the continuity of the "Ngizim" name for 2000 years, so it is highly speculative. To our knowledge, the earliest written source referring to the Ngizim dates to the 16th century. When a French source mentions them at the end of the 17th, they were referred to as "Enghzeum." If oral traditions east of Lake Chad and clan affiliations are anything to go by, the Nguedjem (or Ngizim) were also once living in what is today Chad, in Kanem (and as part of the Bulala sultanate in Fitri, perhaps contributing to Carbou's confusion about their origins). While some ethnonyms in the Lake Chad region have persisted for several centuries (Kuri, for instance, or Bade), Ngizim would have to had survived as a name for over a thousand years before its attestation in the 1500s and 1600s. 

Naturally, one would also have to account for Ptolemy's description of Agisymba seeming to imply it was near mountains. Was this meant to refer to an area closer to Air? Tibetsi? Assuming the Garamantes would have traveled south through Kawar and reached the Lake Chad area, the land of Agisymba with rhinoceroses and mountains (nearby?) had to have been to the east or west without necessarily traveling too far. The Air massif to the west is an intriguing idea, but presumably the ancestors of the Teda and Daza may have known the region in or near Kanem more intimately if they approached it from the Tibetsi and Borku regions. Again, if one assumes the Garamantes were in contact with peoples who spoke a language like that of the Teda and Daza since the days of Herodotus, and the latter groups likely knew of Kanem or Lake Chad, it is possible the presumably Berber-speaking Garamantes adopted words for southern peoples acquired through them. 

Ultimately, the Agisymba as Ngizim remains just a theory. For it to be true requires one to make a number of unproven assumptions that the current data do not fully support. We assume that ancient peoples who spoke an early form of the Teda, Daza, and Kanuri languages in the central Sahara or Tibetsi were referring to a group as something like Ngizim nearly 2000 years ago. Then, through their contacts with the Garamantes, the latter referred to the Ngizim in a way that was later recorded in ancient sources as Agisymba. If the Ngizim and Bade (Bedde) were perhaps both living near Lake Chad in antiquity, then the name Agisymba could be a deformation of something like Ngizimba, referring to a region near Lake Chad, probably Kanem. One must also assume the Ngijem in Kanem are related to the Ngizim west of the lake, even though traditions in the late 1800s and early colonial period connected them to the Bulala.

Thoughts on Manan and Early Kanem

Dierk Lange's reconstruction of a map of the Lake Chad basin based on Ibn Sa'id (from "La région du lac Tchad d’après la géographie d’Ibn Saʿīd. Textes et cartes").

Manan, the earliest known capital of Kanem, remains an enigma. Its exact location remains unknown. Nor is there much information on the nature of this settlement or town from external Arabic sources. Oral traditions in Kanem and Borno shed little light on the matter, too. Revisiting the various sources on Manan, however, suggests its antiquity predated the 10th century. Moreover, by using what we know of later capitals in Kanem and Borno, it may be possible to reconstruct some of the characteristics of Manan. Doing so emphasizes the importance of continuity over time in the annals of Kanem and Borno. Of course, various changes took place over several centuries, but distinct features of the Kanembu, Kanuri, and various peoples living near Lake Chad persisted. In addition, reconsidering Manan as a political capital also requires rethinking what we know of early Kanem and how the consolidation of Kanem under a single dynasty required the unification of many peoples inhabiting the region.

Manan and Early Kanem in Medieval Sources

Naturally, one must begin with the medieval Arabic sources. Most of them are available in the Corpus of Early Arabic Sources for West African History, edited by Levtzion and Hopkins. The first to unambiguously allude to Kanem was al-Ya’qubi in the 800s. According to him, there was a Zaghawa kingdom in Kanem. He wrote, “Their dwellings are huts made of reeds and they have no towns.”[1] It is possible that a settlement later known as Manan already existed in this century, but al-Ya’qubi did not consider it a city due to all its structures being reed huts. Intriguingly, he also described another Zaghawa group, al-HWDN, with a king from the Zaghawa. Then he mentioned an enemy kingdom called Malal, “who hate the king of Kanim.”[2] Malal, ruled by a king called MYWSY, could have been a smaller polity later absorbed into Kanem. The place name, Malal, does recall Manan and, according to one spelling, M.lan.

Besides al-Ya’qubi, the next detailed accounts of Kanem are by al-Muhallabi from the late 900s (quoted by Yaqut). This author reported that the Zaghawa had 2 towns: Manan and Tarazki, both in the first clime at latitude 21 degrees. Yet their houses were still reed huts, including the palace of the king: “Their houses are all reed huts as is also the palace of their king, whom they exalt and worship instead of Allah.”[3] This source is important as the earliest to unambiguously name a city or town of Manan in Kanem. Although the source also expresses the idea of a multiplicity of Zaghawa, it is quite clear that it is Kanem being described. Moreover, a glimpse of how the king’s authority may have been seen and what counted as wealth can be gleaned from the evidence: the king’s wealth was counted in livestock like sheep, cattle, camels and horses. The latter is especially significant due to the importance of horses in Kanem’s military power. The subjects were also said to go naked or wear skins, while the king wore silk and woolen clothes.[4] Unfortunately, there is no clear indication of where Manan was located, but the site must have had access to trade routes through the Sahara and perhaps to the east.

After al-Muhallabi, al-Idrisi wrote about Manan. To this 12^th century geographer, Manan was 12 stages from Tamalma. According to al-Idrisi, “Manan is a small town without industry of any sort and little commerce. Its people breed camels and goats.”[5] This description suggests that Manan was quite small and about 8 days travel from Anjimi (Njimi, the capital of the Islamic Sayfawa rulers). Because he utilized sources from different time periods without reconciling their inconsistencies, one must interpret al-Idrisi cautiously. For instance, he also wrote that Manan was 13 stages away from the Tajuwa “town” which may have been an example of al-Idrisi inventing a town. He also claimed that Manan was where the “governor” of the country lived, who led an army mostly consisting of naked archers. This is a fascinating piece of information, but possibly evidence that al-Idrisi uncritically repeated outdated information. The reference to naked archers also suggests the Haddad, an artisanal caste group in Kanem who were known in much later times as the only people to use the bow and arrow.[6]

Last, but certainly not least, Ibn Sa’id wrote a detailed account of Kanem that drew heavily on the lost work of Ibn Fatima. The description of Kanem is that of the period of Dunama Dibalemi (c. 1210-1248). It is also thanks to Ibn Sa’id that we know Njimi was southeast of Manan. Apparently, the earlier Kanem capital was level with the angle of the Lake (Lake Chad) at longitude 51 degrees, latitude 13 degrees. Manan was specifically said to have been the capital of the pagan ancestors of Kanem’s king (in other words, the Duguwa branch of the Sayfawa). Ibn Sa’id also specifies that to the east of Manan wandered the Zaghawa and to their north, the Akawwar (presumably Teda-Daza groups in Kawar?).[7] Basically, Manan was to the southeast of Kawar (and south of the Tibesti Mountains and Borku) while to its southeast, Njimi was closer to Bahr al-Ghazal (40 miles away from this river). When one considers the higher water levels of Lake Chad in the first millennium of our era and the fact that the Bahr al-Ghazal was consistently flooded in the period from 900-1150, agricultural settlements could have thrived in Kanem.[8] Manan, located closer to Lake Chad than Njimi, would have made sense for a capital since it was closer to the areas from which the ancestors of the Sayfawa migrated: Tibesti, Borku, Kawar. At the time, it would have been able to support farmers, herders, and enjoyed closer proximity to the trans-Saharan routes. Naturally, shifting the capital to Njimi with Islamization may have been partly motivated by a desire for better agricultural land as the population moved toward greater sedentarism.

Considering Manan and Early Kanem in Today’s Scholarship

Moving into the modern era, where did scholars believe was Manan? Borno historian Muhammad Nur Alkali postulated a possible location in the Shitati region of Kanem. Some ambiguity can be seen in his attempt to locate it along the northeastern shores of Lake Chad yet also indicating a general location in the Shitati area.[9] When this region was visited by Nachtigal in the 1870s, it was in a part of Kanem that had become largely the terrain of nomadic groups. By this era, it did not neighbor Lake Chad but included numerous valleys. In total, more than 50 valleys could be found in Shitati, which also featured a natron lake. In Nachtigal’s time, most of the people residing in Shitati were Yuroa, Orabba, and Qadawa, the latter a Kanembu group of the Dibbiri with Daza ancestry. The Dibbiri, of course, appear early on in the Diwan since the mother of the first “black” mai, Salmama, was the son of a Dibbiri woman named Hawa.[10] Besides these aforementioned groups, some Kanembu and Danoa (Haddad) farmers also resided in the area. Most importantly, Nachtigal named a place called Maten el-Milah that was no longer part of Shitati. Instead, it consisted of valleys on the path to Borku.[11] It is likely a coincidence, but Manan was sometimes rendered as Matan in written Arabic sources. In the case of this place, Nachtigal reports that it was an Arabic name (Fountain of Salt) and not an indigenous one of deeper antiquity. In other words, Shitati may have once harbored the early capital of Kanem, but there is no smoking gun to irrefutably demonstrate it. In its favor is its location northwest of Njimi and proximity to Lake Chad, which enjoyed higher water levels over 1000 years ago.

Also worthy of consideration is the theory of H.R. Palmer. Palmer, a towering figure in colonial-era scholarship on Borno, was guilty of contemporary racial theories, shoddy or questionable linguistic connections, and sometimes lacking transparency for his sources. Nonetheless, Palmer did work with local elites to gather traditions or translate various manuscripts, meaning that his work is unavoidable for any serious interest in the history of Borno. In terms of Manan and early Kanem, he even gathered traditions (which appear to contain anachronisms) of Dugu’s alleged southerly campaign.[12] As for Manan, Palmer apparently connects it to the Kulu or Kuluwan region.[13] Since Madan or Malan appears to have been the place where the early mai Fune died, this is consistent with Manan as a royal capital. Against Palmer’s theory, however, is the area of Kuluwan between Kanem and Bagirmi. This is not consistent with medieval Arabic sources placing Manan to the northwest of Njimi. It was also the area where Katur, a successor of Fune, died, according to the Diwan. Ultimately, Palmer’s attempt to link Manan or Matan with the Kuluwan region is not persuasive and contradicted by the Diwan which places Manan (or M.lan) in Kanem.[14]

Manan, Malal, and Early Kanem

Besides Palmer, John Lavers also proposed an intriguing theory for early Kanem with relevance to Manan. Based on the brief description of Kanem by 9^th century author al-Ya’qubi, Lavers has suggested that in c. 872, Kanem had “Zaghawa” rulers but also competed with neighboring “Zaghawa” polities. One of these groups was called Hawdin, and another was Malal. Since the ruler of Malal was called Mayusi or Mai Wasi, and the Zaghawa king Kakarah (according to one reconstruction), is it possible that the rulers of Malal superseded the early rulers of Kanem and became the reigning dynasts?[15] This theory is, of course, based on the assumption that the title of the ruler of Malal was mai and since that is the title used by the kings of Kanem and Borno, Malal’s rulers may have replaced another polity and became the dominant power in what became known as Kanem. Of course, the absence of sufficient evidence limits its probability though it would possibly correlate with the M.lan (or Manan) mentioned in the Diwan as the place where Funa died. Assuming, of course, that M.lan is equivalent to Manan and possibly related to Malal.

This theory is likewise interesting if one accepts Terio’s notion that the Zaghawa king of Kanem named by al-Ya’qubi was actually the title kireh, used by the Zaghawa for kings.[16] Alternatively, the rulers of Malal may have intermarried with the “Zaghawa” or so-called Duguwa in Kanem, since the Diwan presents Dugu as the father of Funa, the mai who allegedly died in M.lan. Furthermore, Zaghawa traditions remember a Zaghawa king of Kanem named Douk Bourmè, presumably the same Dugu recalled in Kanuri girgams and the Diwan.[17] Since dating these figures is a hazardous exercise, one can only tentatively assign dates. If the excessive reign lengths in the Diwan are meant to refer to generations as well as to stretch the dynasty back to Sayf b. Dhi Yazan, we cannot be sure which kings are semi-legendary or when their reigns may have taken place.

Obviously, this makes any endeavor to tie the polities mentioned by al-Ya’qubi with the tentative chronology of Lange problematic. Yet Lange has written, “Un souverain du nom de Funa semble avoir régné au milieu du VIIIe siècle, Arsu à la fin du VIIIe et Katur au milieu du IXe siècle.”[18] If one accepts this mid-700s date for Funa, who died in a place called M.lan that may be Manan, then it is difficult to reconcile with the theory that Malal in the 870s was in conflict with Kanem and ruled by their own independent king. Unless one proposes that Funa died in a war with Malal sometime in the 700s or later traditionists merely fused the two dynastic lines together after their intermarriage, it is difficult to square with Lange’s suggested chronology for these “Duguwa” kings. Nonetheless, the possible Malal kingdom or polity as a rival of Kanem under the “Zaghawa” could be a reference to a fusion of Zaghawa, Teda-Daza, and Kanembu elements that occurred over several centuries, consolidating as a single dynasty with regional supremacy in the 900s or 1000s. Malal may, if the theory has any validity, have been a smaller polity of Kanembu-affiliated people whose capital was changed into Manan.

Concluding Thoughts

Clearly, the location of Manan is a subject of debate. The early medieval sources provide only glimpses of pre-Islamic Kanem and must be used cautiously. Indeed, due to some of these authors never actually seeing Kanem themselves, their reports are not based on direct experience. Nevertheless, they provide a few clues about how early Kanem developed and a rough idea of where Manan could have been.

Subsequent scholars from the colonial and postcolonial eras offered new theories, but without any definitive evidence to pinpoint exactly where Manan was. With the recent confirmation of Njimi’s likely location at Tié, and the general idea that Manan was to the northwest of Njimi, we can more confidently assert that it was closer to Lake Chad. This makes the theory of Muhammad Nur Alkali plausible. The more speculative theory proposed by Lavers could facilitate identifying the placement of Manan, too. Of course, our interpretation of it relies on the questionable assumption that Manan, Malal, and M.lan were designating the same place in the medieval Arabic sources and the Diwan. Such a theory could elucidate why Manan has been forgotten in Kanuri tradition, too.

In spite of its obscurity, Manan’s position as the earliest known capital of Kanem makes it significant in the growing sedentarization and consolidation of a powerful kingdom to the east of Lake Chad by the 9^th and 10^th centuries. While definitive proof remains elusive, the cumulative evidence suggests that Manan was an early political center northwest of Njimi, and possibly linked to the polity of Malal that appeared in the writings of al-Ya’qubi in the 800s. With future archaeological surveys and excavations, a more confident location for Manan can be found which could meaningfully change our perception of the origins of urbanism in Kanem.

---

[1] al-Ya’qubi in Corpus of Early Arabic Sources for West African History, 21. Some possible references to the area of Kanem may predate c. 872, but the earlier Arabic authors only use the name Zaghawa. This term may have been used very broadly for many different ethnolinguistic groups living between Nubia and the central lands of Black Africa. It is possible, nonetheless, for some Zaghawa groups to have lived as far west as Kanem in the 9^th century (or earlier) and interacted with groups more closely related to the modern Teda, Daza, Bideyat and Kanembu.

[2] Ibid. John Lavers has also proposed an interesting idea about this polity, although it remains purely conjectural without additional sources.

[3] Ibid., 171. The reference to the house of the king is important, even if it was built with reed and not the monumental type of architecture Kanem and Borno developed after Islamization. The town of Tarazki is also intriguing as it bears a resemble to the later Kanem town of Daniski.

[4] Ibid.

[5] Ibid., 114. Earlier, al-Muhallabi reported that livestock and horses were the wealth of Kanem’s ruler.

[6] For more information on the Haddad, see Henri Carbou, La Région du Tchad et du Ouadaï. This group has been the subject of more than a few unlikely or highly problematic theories. Lange, for instance, has proposed identifying the Haddad or Danoa with the so-called Duguwa dynasty. There is perhaps some basis in this theory due to Haddad oral traditions remembering an early ancestor called Dana and the fact that the Haddad, an Arabic name, are referred to as Duu or Dugu by the Kanembu (See Edouard Conte, Marriage Patterns, Political Change). Contrary to the theory of a Banu Duku or Duguwa origin, the Haddad are remembered in oral traditions as sharing a common descent with the Bulala or perhaps with slaves or servile populations in Kanem during the period of Bulala rule. This theory, of course, requires deeper analysis but Carbou’s traditions of origins for the Haddad seem to only go as far back in time as the Bulala period. Interestingly, Nachtigal himself focused more on the N’Galma Dukko as a group descended from an early prince of the Sayfawa dynasty, perhaps Duku or the so-called “Duguwa” branch. See Sahara and Sudan Vol. 3.

[7] Ibn Sa’id in Corpus of Early Arabic Sources for West African History, 188-189.

[8] For information on the fluctuations in the levels of Lake Chad, see “Floods, Droughts, and Migrations: The Effects of Late Holocene Lake Level Oscillations and Climate Fluctuations on the Settlement and Political History in the Chad Basin” by Karsten Brunk and Detlef Gronenborn in Living with the Lake: Perspectives on History, Culture and Economy of Lake Chad. These authors have argued that the Bahr al-Ghazal was flooded throughout this period and the Sahel zone was semiarid and subhumid. But the Sudanic savannah lands would have been humid. Their theory that the center of early Kanem in c. 900 was the Bodele region is fascinating, but this seems too far north (and east) to help one determine Manan’s probable location.

[9] Muhammad Nur Alkali, Kanem-Borno Under the Sayfawa: A Study of the Origin, Growth and Collapse of a Dynasty, 24, 57. For yet another 20^th century scholar’s theory, see Zeltner’s Pages d’histoire du Kanem, which has suggested Manan was in the Egey region of Kanem.

[10] Dierk Lange, Le dīwān des sultans du (Kānem-)Bornū: Chronologie et histoire d'un royaume africain (de la fin du Xe siècle jusqu'à 1808), 70.

[11] Gustav Nachtigal, Sahara and Sudan Vol. III, 65-68.

[12] See H.R. Palmer, Bornu Sahara and Sudan for several examples of similar types of stories, legends or traditions on the kings of Kanem and Borno. Included is one 1751 manuscript which traces the origin of the first Saif to Aghani, a land Palmer claims was the Zaghawa, called Aghna (Arna) by the Kanuri.

[13] H.R. Palmer, Sudanese Memoirs Vol. I, 7, 74-75.

[14] Dierk Lange, Le dīwān des sultans du (Kānem-)Bornū: Chronologie et histoire d'un royaume africain (de la fin du Xe siècle jusqu'à 1808), 66.

[15] John Lavers, “Kanem and Borno to 1808” in  Groundwork of Nigerian History, 189.  

[16] Abdelkerim Souleyman Terio, Origine et évolution des Zaghawa: Du royaume du Kanem aux Etats modernes (VIIIe-XXIe siècle), 94.

[17] Ibid., 89.

[18] Dierk Lange, Le dīwān des sultans du (Kānem-)Bornū: Chronologie et histoire d'un royaume africain (de la fin du Xe siècle jusqu'à 1808), 143. 

Les coutoumes familiares au Kanem

 

Alifa Zezerti

Les coutumes familiales au Kanem by Robert Bouillé is a dated colonial-era study of kinship and social organization in Kanem. Sometimes repetitive, the author focuses on the Kanembu, Buduma,  Fulani, Arab, and Tubu groups in Kanem. Each are distinct but similar, especially in terms of a patriarchal family structure in which wives and daughters were treated unequally. Nonetheless, the author believed that the ethnic diversity in Kanem and inter-ethnic antagonisms made French colonial rule more stable. However, as a strong proponent of rational and stable colonial administration, customary law and practices among Kanem's population required a gradual approach for reforming its population. This meant that slavery, or descendants of slaves, were still very much a part of communities while the Haddad remained an outsider group. Sometimes blood vengeance was still carried out by Tubu or Arab populations. However, with the changes introduced by colonial rule, youths were beginning to show more independence from patriarchal authority while some married women became somewhat more autonomous. Nonetheless, the portrait of the various peoples of Kanem is probably at least partially accurate for how the peoples of the region organized their communities in the 1800s and even earlier, in some instances.

In terms of the alifa based on Mao, Les coutumes familiales au Kanem is not the most detailed. Besides recounting how the French first signed a treaty with the alifa and the pacification of the province, hints of its political and economic structure are largely subdued. The current leader at Mao, Zezerti, was said to be from the Zegbada family. They were supposedly part of the Samarous tribe of its founder's name. His family group was closely aligned with the Mustafa family of Zegbadas who had apparently intermarried with them for around 2 centuries. Zezerti's mother was supposedly from this Mustafa family. Nonetheless, when comparing the genealogical information of the alifas from Landeroin's time to Bouillé, the former listed a "jagata Moustapha" who was the father of an alifa who must have resigned in the early 1800s, Ahmadou Kalli. However, there was an earlier alifa, Moutta Kinguimi, whose "proper" name may have been Moustapha. Is it possible that Zezerti and the Mustapha family represented one lineage group which only effectively occupied the throne in the 19th century, when conflicts between Waday and Borno over control of Kanem (and the Alwad Sulayman) rapidly transformed the political landscape of Kanem? Sadly, Landeroin's Notice historique has little to say about the alifas who reigned before the 19th century. A similar question might be raised about the possible religious status and authority of the alifas before the 20th century. Since Zezerti's grandfather had performed the pilgrimage to Mecca and was renowned for his baraka, Bouillé may have retroactively sought to elucidate this alleged religious/spiritual authority of the dynasty before Zezerti's grandfather in its origin. Intriguinly, Landeroin recorded that the father of Dala Afouno, the first alifa of Kanem, was a Mallam of Magoumi descent. Perhaps descent from an illustrious Islamic scholar or religious leader was part of the dynasty's legitimacy since its inception. 

In terms of Kanem before the fall of the Sayfawa, there are a number of possible hints to the influence of Kanuri groups who came to Kanem after the fall of the Bulala. For instance, the alifas appear to have set up a court with officials often using titles of Borno origin (some undoubtedly began when the Sayfawa ruled in Kanem). Their court also used a system of "fiefs" for some dignitaries and officials and relied on tribute of various types. During the period of Bornoan suzerainty, Kanem also had an institution called the Bit el mal, a sort of public fund that produced food for the poor, gifts to mallams, aided foreigners, and paid masons for building or maintaining mosques. Furthermore, a type of mutual aid practiced by farmers in Kanem, the debdou custom, allowed for Kanembus experiencing misfortune to receive presents of animals, clothes, grain, or money to help. The custom also featured dancing and drums. Whether or not this practice of mutual aid predated the period of the alifas in Mao is unclear, but Bouillé emphasized a Borno origin. 

It would be worthwhile to compare the description of Kanembu social and kinship patterns with those of Borno during the same time (or earlier). It may help establish certain patterns or trends which could have begun in Kanem during the medieval period and were maintained by the Kanuri in Borno. Similarly, it may help elucidate patterns of social change in Kanem under the Bulala and alifa rulers. Naturally, Kanem's landscape differed quite significantly from that of Borno and the region's growing aridity led to some parts being occupied mostly by nomadic groups. Nonetheless, a close examination of family structure, marital patterns, customary law, and land tenure systems could provide possible clues for earlier centuries in the annals of Kanem and Borno. For similar reasons, determining the origin of the Haddad and the interactions between Kanembu and Buduma populations may also shed light on how some aspects of the history of both regions diverged. 

Tarikh el Kawar

Michel Le Sourd's "Tarikh el Kawar" was published in Bulletin de l'I.F.A.N. in 1946. However, since the author was able to collect oral traditions from various parts of Kawar from people still old enough to recall precolonial days, there is some value in this rather short article. Lamenting the lack of interest by literate Kanuri and others in Kawar, Le Sourd attempts a broad overview of Kawar history based solely on oral traditions. Without even referring to the external medieval sources or references to the region from Kanem and Borno, Le Sourd's study is probably most useful for a glimpse of Kawar in the last 2 or 3 centuries. 

According to Le Sourd, the populations of Kawar in chronological order are the following: Soo (Sao), Konnas, Terras, Kanuris and Tubus. Unfortunately, none of his informants apparently gave him information on the Konnas and Terras, who were perhaps the people of Kawar when the region was described in medieval Arabic sources from the 9th century onward. As for the Sao, that appears to be a legend accompanying every Kanuri population. Indeed, some of the same traditions of the Sao familiar from Borno are repeated here, with a few minor differences. In fact, Fachi appears to have retained a stronger connection to legends of the Sao, perhaps a testament to the region's importance during the reign of Ali b. Dunama in his campaigns against Air (Le Sourd 4). However, Fachi traditions remember the sultan of Borno also killing the Sao of Fachi. Moreover, the story of the Kanuri using henna and a trick to trap the Sao is reported, even though this tradition is really more pertinent to the area of Birni Gazargamo or Borno proper (5). Alternatively, the memory of the Soo in Kawar could also be linked to the Tubu name for Seguedin, called Soo by those in the Tibetsi (15). Perhaps the name Soo for Segedin once covered a broader area of Kawar, but then mixed with Kanuri legends of the Sao? Intriguingly, the Sao are remembered in tradition for the use of the ksur or fortified sites across the area (2). Even the town name Agram supposedly means fortified village in Kanuri, although a possible Berber etymology has been proposed elsewhere (35). The Kanuri association of fortified sites with the Sao probably means that the towns or villages already had fortifications before their arrival, though the Kanuri may have expanded or maintained them. 

As for the Konnas and Terras, perhaps the latter is an allusion to the Tura? As in the Tura who also migrated to Borno and of whom some examples of early mahrams have been located. The early Tura mahrams suggest close ties between Kanem and Dirkou which began by the 1100s. Indeed, close relations between the rulers of Kanem and Dirkou (as well as other parts of Kawar) likely began before the 11th century. As for the Konnas, the appellation brings to mind a word used in Borno for pagans to the South, said to be associated with Kwararafa. Since it is highly unlikely for people from Kwararafa to have preceded the Kanuri in Kawar, if there is any connection with the Kwona or Kwoana in Borno, it may simply be a reference to a pre-Islamic population living in Kanem before its Islamization during the time of the Terras (Turas?).

With the arrival of the Kanuri, a better sense of the history of the region begins. According to his informants, Le Sourd found that nearly every site in Kawar was previously inhabited by the Kanuri when the Tubu arrived. Thus, the Kanuri were at Guezebi-Guassar first, before the Tubu (6). The Tomaghera Tubu who migrated into the area intermarried with them, becoming the Guezebidas (7). Similarly, Dirku was said to have had a Kanuri population already when the Tubu arrived (10). Furthermore, Kanuri were the first at Sara, Djado, and Djaba. When recounting the legend of Dalla Kerkaou of Sara, Le Sourd also alludes to the Tuareg of Air ruling Djado, but the chronology for this probably legendary story is ambiguous (16). Is this a reference to the period of Air control in Kawar during the 18th and 19th centuries, or referring to a more ancient conflict between the peoples of Air and Kawar? Last but not least, the Koyam were also present in villages stretching from Agram to Termit (28). Their presence in the region, if they are indeed the Kay referred to in the Diwan, is likely of far deeper antiquity and may be linked to the early salt trade between Kawar and Kanem-Borno.

Also of interest is Le Sourd's observations on differences in spoken Kanuri in Kawar. According to his observations, the Kanuri spoken in Fachi, Djado and Tedjerhe in the Fazzan are similar and distinct from other Kawar villages (23). It is also asserted that these areas were once closer to Zeila, too  (22). Although travel between the Kanuriphone populations was negatively impacted by insecurity, their similarities despite the distance between the Fezzan and Fachi illustrate the latter's economic importance. The similarities to Djado lead one to wonder if that area was once also more economically important for a short route to the Fezzan. Fachi certainly was of great economic significance for trade. Moreover, traditions suggest Fachi, or Agram, was also an important site as early as the 15th century for the Sayfawa ruler, Ali b. Dunama. Its links to the Koyam salt traders are areas worthy of further investigation. 

In summation, Le Sourd's exposition on Kawar history, based solely on oral traditions, raises a number of interesting questions. That the current peoples of the area, of Kanuri and/or Tubu origin, attribute a Sao presence predating other groups could likely be a result of widely shared Kanuri legends of the Sao. If so, then this suggests the Kanuri population may be associated with movement into Kawar during the period of Sayfawa rule in Borno. Since some of the legends current in Kawar even refer to traditions on the origins of Birni Gazargamo to the South, one wonders if the 1400s might be a plausible period of origin for some of the Kanuri in Kawar. Of course, the close relations between Kawar and Kanem before this period could also have led to Kanembu and Kanuri-speaking communities in the area long before the 15th century. But the traditions of the Kawar Kanuri of peoples called Konna and Terra preceding them may be an allusion to the pre-Islamic and medieval population of the oases. Their disappearance may have been due to a merging into the Kanuri population, as the Tura of Borno suggest.

Rambling on Pre-Islamic Kanem Religion

Although the extant source materials are far from ideal, it is possible to reconstruct the pre-Islamic foundations of early Kanem. As a major Central Sudanic state that was already in existence by the 9th century, when its religion was described as worship of kings, attempting to recover the "pagan" foundations of the state can shed light on its early origins. Indeed, as a state with strong links with clan names that appear among Teda, Daza, Kanembu and Kanuri peoples, the origins of Kanem seem to be significantly correlated with the non-Islamic, ancient roots of these related peoples. The Zaghawa and Bideyat, also part of this larger world of Central Saharan and Sahelian "black" nomads and semi-nomads, provide even further evidence of early Kanem's foundations, particularly in the relevance of the king in ceremonis tied to rain and agriculture.

Unfortunately, the medieval Arabic sources are light on details of pre-Islamic Kanem. Written off as pagans by al-Bakri and described as worshippers of their monarchs by al-Muhallabi, early Kanem's "paganism" did not interest Muslim visitors. However, when reading later ethnographies of the Zaghawa and their pre-Islamic rituals, one may begin to see what exactly al-Muhallabi and others were describing. To be specific, it is unlikely that the people of Kanem actually worshipped the mai or king (kireh, or kakireh). Instead, royal authority was tied to rituals reflecting a larger cosmovision or worldview. The ancestors and spirits needed to receive offerings (or, sadaga, Chappelle 87). Kingship among the Zaghawa studied by Tubiana or Kanem as described by al-Muhallabi was a position closely linked to indigenous religious practices. In early Kanem, the mai and the royal lineage were necessary in rituals that led to rain and bountifulness. Without their position as sacral rulers who could lead the necessary rituals and offerings, the people may face droughts, poor harvests, famine, and other turmoil. It is in this way that the subjects of the early mais of Kanem venerated their king, who was believed to be the source of life, death, illness or good health (Hopkins and Levtzion 171). 

Therefore, it is unlikely that the kings of Kanem itself were actually the object of worship. The sacral nature of their position and their appearance in rituals designed to communicate with spirits perhaps led to the view of the king as semidivine or associated with mystical and spiritual authority. By virtue of their special position and the political and ceremonial demands, the king must be separate from the commoners or rest of society. This led to more developed forms of ritual seclusion of the king and veneration of the royal lineage. Indeed, in Borno sacrifices were offered to ancestors and alms given while the king was ritually secluded (Bjorkelo 33). One can even detect echoes of this among the Islamic Sayfawa dynasty for centuries after al-Yaqubi and al-Muhallabi wrote of Kanem. Indeed, the interest of the Sayfawa mais in performing the pilgrimage and Islamic scholarship likely resonated with pre-Islamic beliefs of the mai's position being rooted in sacral patterns. For this reason, the kings of Kanem were often ritually secluded, endeavored to avoid be seen eating, and made appearances as specific festivals or public celebrations. Indeed, Borno natives enslaved in Saint Domingue reported to Descourtilz that the mai of Borno never left the palace and those who dared to look directly at him were executed (Descourtilz 146). Remarkably, al-Umari in the 14th century also reported on rare public appearances of the mais of Kanem, who only appeared during 2 festivals (Hopkins and Levtzion 260). Consequently, Islam may have not represented so much of a rupture to this process as it could have been incorporated into "traditional" patterns of kingship established in ancient Kanem.

In addition to the ritual seclusion of kings in Kanem (and Borno under the Sayfawa), the specific reverence attached to ancestors and remembering clan affiliations through maternal lines may have been an additional remnant of pre-Islamic Kanem. According to Abraham Rosman's analysis of Kanuri social structure, clans were once very significant in the history of the Kanuri people. Kanuri were divided into jili with distinctive facial scars. Each jili had a common ancestor and a head appointed by the mai to collect taxes (Rosman 93). This system of clans and clan sub-groups, organized patrilineally but also emphasizing the maternal clan, may have been important to the ancestors of the Kanuri if they did not marry within the same clan If early clans among the ancestors of the Kanuri were similar to those observed among the Tubu and Zaghawa, then clan founders would have been of great importance for ancestral veneration. Specific clan taboos, areas for offerings, and even animals associated with the clan founder could have been major concerns in early Kanem (Chapelle 362). The mai, or head of the senior lineage and clan group, would have been at the zenith of clan leaders whose marriage alliances with other clans brought some degree of centralization and common investment from each one. The spread of these clans into Kanem, Borno and even other areas like the Tibetsi and Kawar, must have, even if fabricated in some cases, brought a degree of unity and shared belief in a shared origin defined by similar ceremonial offerings to ancestors. Women and the clan of the queen mothers of Kanem must have been important for this reason, or at least important enough to be recalled by the later editors of the Diwan for the early centuries of Kanem.

An instance of the significance of ancestral veneration and sacrifice can be found in Idris b. Ali's campaigns in Kanem. While campaigning against the Bulala,  Idris b. Ali stopped to pray at the tombs of previous Sayfawa rulers near Njimi. Although the prayers Idris b. Ali recited for his ancestors were Islamic in nature, ethnographic work on the Teda and Daza also include special reverence tied to the graves of ancestors. In addition, Idris b. Ali then gave alms and cattle after the Quran was recited at his ancestors' graves in Njimi. Through commemorating and making offerings to ancestors, one also asks for their intercession. The past rulers of the Sayfawa dynasty and the specific clans of their maternal ancestors points to the possible role of clan intermarriage with forging alliances among the various Teda, Daza, Zaghawa, Kanembu, and Kanuri groups who comprised the population of Kanem. Each clan would have been linked to ruling dynasty through intermarriage, rituals, taboos and ceremonies that brought them to the royal court and created bonds through offerings. Manan, the "pagan" capital of Kanem, possibly featured a stream, grotto, or spiritually significant place for offerings. It is likely that a form of divination ceremony took place at caves, the kurkuri of the Daju and Zaghawa mentioned by Palmer (in Bornu Sahara and Sudan 212). It is also possible that specific clans, such as the Tomagheras and others mentioned in the Diwan, may have had names derived from a wadi, or an animal believed to have assisted the clan founder. The clans could have identified their ancestor or founder with a mountain, tree, or even serpent (Tubiana 16). If so, there may have been specific clans associated with specific roles in rituals involving the king (Tubiana 45). 

The example of the rain ceremony among the Zaghawa, for instance, brought together members of various Zaghawa sub-groups and clans to a spiritually significant place for a ceremony that included the offering to a spirit and ritual feasting. A similar practice was also observed among the Keira sultans of Darfur and the rulers of Wadai, states which were also influenced by Kanem-Borno. Indeed, something similar was likely practiced in Borno, too. Muhammad Bello's letters with al-Kanemi accuses the Muslims of Borno of shirk, as they performed sacrifices at rivers and possibly even sacrificed a virgin in a river (Bello, 72). While these practices might reflect some of the pre-Islamic beliefs of Buduma, Marghi, Kotoko and "Sao peoples" who initially resisted the Sayfawa, they also resemble those practiced among the Teda, Daza, and Zaghawa. Thus, the alleged worship or sacrifices to idols in trees and rivers could have also been a remnant of Teda, Daza, and pre-Islamic Kanuri beliefs. Indeed, even if the improper practices written of by Muhammad Bello were exaggerated or perhaps a product of indigenous populations living near Borno who gradually adopted Islam, the continuity of ritual seclusion among the mais suggests a strong link to the pre-Islamic past. Yet it would be intriguing to study the words for God among the neighbors of the Kanuri to establish the origin of ramabe, a term which may have been the original Kanuri word for God (Koelle 275).  In light of these considerations, perhaps for reasons of ritual seclusion and sacred places more so than other factors, Idris b. Ali's mother had the brick mosque and palace at Gambaru constructed for his use. Njimi in Kanem, the capital after Manan, also used fired-brick for constructions that clearly designated the area of the mai as distinct.

Addressing the issue of the mune may further elucidate the matter at hand. The mune, often presented as a pagan relic that, once opened or destroyed by Dunama Dibalemi, led to dissolution and civil war among the Sayfawa (which certainly was common enough in 13th and 14th century Kanem as civil wars, assassinations, deposed mais, usurpers, and conflict with the Bulala and Sao indicate), presents a number of historical questions. Palmer, whose Sudanese Memoirs is an important yet problematic source on Borno, mentions the Zaghawa of Wadai referring to the Quran covered in skins as Mani. Moreover, the "pagan" Beli and Zaghawa allegedly believed mani was a ram stored in a cave. Even though one must use Palmer very carefully, the mune of Kanem was likely something derived from pre-Islamic beliefs yet redefined or adapted to fit Islam. Thus, Ahmad b. Furtu described the mune as something wrapped up and vital to the success of the Sayfawa. Indeed, our Bornoan chronicler compared it to the Ark of the Covenant. Thus, opening or destroying the mune was perceived by a 16th century Borno intellectual as un-Islamic. This is a testament to the degree to which pre-Islamic Kanem practices were capable of adaption to new religious and political paradigms. The mune, from pre-Islamic Kanem to the Muslims of the Sayfawa, was a talisman or highly charged spiritual object capable of reinterpretation through Islamic lens. According to Muhammad Yanbu, the mune was also carried into army and therefore associated with the military successes of the mais (Bobboyi 87).

Although there is inherent danger in relying too heavily on 20th century ethnographic studies of the Zaghawa and Tubu peoples, their ethnographic present likely illuminates aspects of early Kanem. After all, Islamization among the Tubu and Zaghawa is a late phenomenon, and they appear to have been closely linked to the major clans that contributed to the formation of Kanem. Thus, their practices of sadaga or offerings, veneration of ancestors, practice of clan totems and taboos, and belief in the centrality of the king in specific rituals tied to rain and agriculture are probable indicators of what early Kanem looked like. Thus, the early kings of Kanem were probably not objects of worship of their subjects. However, their position was closely linked to spiritual power or contact with ancestors and spirits that required offerings. Through ritual seclusion and the maintenance of specific ceremonies of rain-making or, in the case of Darfur, a festival in which the king physically engaged in a sowing festival, the kings of Kanem were the ritual center of the state. Marriages with other clans must have further cemented the unification around an early polity as other clans would have contributed participants in important state ceremonies. Early Kanem's conversion to Islam was not necessarily seen as disruptive to this "traditional" order as Islam and its message of a universal God could have been used to buttress the belief in the king's sacral powers. Thus, the mais could have continued to venerate ancestors, maintain some of the pre-Islamic customs, and incorporate Islamic scholarship and pilgrimage into their toolbelt of spiritual (and supernatural) power. Even after conversion and despite the resistance of some Tubu and Zaghawa to Islam, the maintenance of clan identities and traditions, remembered through oral history, written genealogies and chronicles, must have contributed to some degree of common identity and investment in the Sayfawa state from its core.

Works Consulted

Barth, Heinrich. Travels and Discoveries in North and Central Africa: Being a Journal of an Expedition Undertaken Under the Auspices of H.B.M.'s Government in the Years 1849-1855. New York: D. Appleton, 1857.

Bjørkelo AJ. State and Society in Three Central Sudanic Kingdoms Kanem-Bornu Bagirmi and Wadai. Bergen: Universitetet i Bergen; 1976.

Bobboyi, Hamidu. The'Ulama of Borno: A study of the relations between scholars and state under the Sayfawa, 1470-1808. Northwestern University, 1992.

Chapelle, Jean. Nomades noirs du Sahara. Paris: Plon, 1958.

Cohen, Ronald. The Kanuri of Bornu. New York: Holt, Rinehart and Winston, 1967.

Descourtilz, M. E. (Michel Etienne) 1809. Voyages d’un naturaliste. Paris: Dufart, père.

Dewière, Rémi. Du Lac Tchad À La Mecque: Le Sultanat Du Borno Et Son Monde (XVIe - XVIIe Siècle). Paris: Publications de la Sorbonne, 2017.

Galaty, John G., and Pierre Bonte. Herders, Warriors, and Traders: Pastoralism in Africa. Boulder: Westview Press, 1991.

Hopkins, J. F. P., and Nehemia Levtzion. Corpus of Early Arabic Sources for West African History. 1st Princeton: Markus Wiener Publishers, 2000.

Ibn Furṭū, Aḥmad, and Dierk Lange. A Sudanic Chronicle: The Borno Expeditions of Idrīs Alauma (1564-1576) According to the Account of Aḥmad B. Furṭū : Arabic Text, English Translation, Commentary and Geographical Gazetteer. Stuttgart: F. Steiner, 1987.

Koelle, S. W. (Sigismund Wilhelm). African Native Literature, or Proverbs, Tales, Fables, & Historical Fragments in the Kanuri or Bornu Language. To Which Are Added a Translation of the Above and a Kanuri-English Vocabulary. London: Church Missionary House, 1854.

Lange, Dierk. Le Dīwān Des Sultans Du (Kānem-)Bornū: Chronologie Et Histoire D'un Royaume Africain (de La Fin Du Xe Siècle Jusqu'à 1808). Wiesbaden: F. Steiner, 1977.

Magnavita, Carlos, Zakinet Dangbet and Tchago Bouimon, « The Lake Chad region as a crossroads: an archaeological and oral historical research project on early Kanem-Borno and its intra-African connections », Afrique : Archéologie & Arts, 15 | 2019, 97-110.

Palmer, H. R. (Herbert Richmond). The Bornu Sahara and Sudan. Nigeria. J. Murray, 1936.

------. Sudanese Memoirs: Being Mainly Translations of a Number of Arabic Manuscripts Relating to the Central and Western Sudan. 1st ed., new impression. London: Cass, 1967.

Redhouse, J. W. “Translation from the Original Arabic of a History or Journal of the Events Which Occurred during Seven Expeditions in the Land of Kānim, against the Tribes of Bulala, &c., by the Sultan of Burnu, Idris the Pilgrim, Son of ’Ali; Preceded by Some Details of the Sultan’s Ancestors.” Journal of the Royal Asiatic Society of Great Britain and Ireland 19 (1862): 43–123. http://www.jstor.org/stable/25228715.

Rosman, Abraham. Social Structure and Acculturation Among the Kanuri of Northern Nigeria  / by Abraham Rosman. Yale University, 1966.

Tério, Abdelkérim Souleyman, Origine Et Evolution Des Zaghawa: Du Royaume Du Kanem Aux Etats Modernes (VIIIe-XXIe Siècle). N'Djaména, Tchad :Éditions SAO, 2014.

Trimingham, J. Spencer. A History of Islam in West Africa. London: Published for the University of Glasgow by the Oxford U.P

Tubiana, Marie José. Survivances Préislamiques En Pays Zaghawa. Paris: Institut d'ethnologie, 1964.

Unesco. International Scientific Committee for the Drafting of a General History of Africa. General History of Africa. London : Berkeley: Heinemann Educational Books; University of California Press, 1981.

Yusuf, Salahudeen., and Muḥammad Bello. A History of Islam, Scholarship and Revivalism in Western Sudan, Being an Annotated Translation with Introduction of Infaqul-Maisur Fi Tarikh Bilad Al-Tukur of Sultan Muhammad Bello Bin Fodio  Zaria: Tamaza, 2013.

The Zaghawa and Early Kanem

Terio Abdelkerim's Origine et évolution des Zaghawa: Du royaume du Kanem aux Etats modernes (VIIIe-XXIe siècle) is a worthy effort to unveil the deeper history of the Zaghawa peoples and the kingdom of Kanem. Although we have reached rather different conclusions on the early history of Kanem and the relationship between the so-called Duguwa dynasty and the Sayfawa, Abdelkerim uses oral traditions and a familiarity with the history of Zaghawa or Beri clans across Chad and Sudan to illustrate the clear "Zaghawa" origin of the ruling clan of Kanem. However, as indicated by the Diwan and Kanuri, Kanembu, and Tubu traditions and clan names, the modern Zaghawa (Koubara, Wegui, Touba Koube as the three main groups with smaller clans and divisions) have ancient ties to the Teda, Daza, Kanembu and Kanuri groups. 

Indeed, if the Diwan is reliable on the mothers of various early mais, the ruling Zaghawa of Kanem intermarried with Tubu and Kanembu clans. Furthermore, as the Teda, Daza, and Zaghawa were mobile pastoralists, they also migrated back and forth through the Tibetsi, Ennedi, Borku, Kawar, Kanem, Borno, and other regions. The Zaghawa today, located on the frontier between Chad and Sudan, possibly migrated there in larger numbers after the so-called replacement of the Duguwa (Banu Duku) dynasty by the Sayfawa in the late 11th century. Or, alternatively, given the wide expanse of territory between Lake Chad and the Nile occupied by related Nilo-Saharan pastoralists, the "Zaghawa" of the medieval Arabic sources was actually a reference to the entire ensemble of black pastoralists in the central Sahara. This would probably explain why the "Zaghawa" were already known to Arabic sources as early as the beginning of the 8th century, since black Saharan populations west of Nubia and south of the Fazzan may have already formed the early state of Kanem to the south.

In addition, linguistics possibly supports a "Zaghawa" or Beri affinity for the early rulers of Kanem. According to al-Yaqubi, writing in 872, the kingdom of Kanem was "Zaghawa" and called Kakira. The Zaghawa language, however, retains the word kireh for emperor or kings, as indicated by Abdelkerim. We are thus inclined to view al-Yaqubi's brief account of 9th century Kanem as evidence of a Zaghawa ruling clan that must have been related to various Teda, Daza and proto-Kanembu groups through marriage. Indeed, genetics likely supports this scenario as population admixture studies of Chadian groups indicates mixing between a group of Eastern African origin with one of West-Central African origin around 1000 years ago. Of course, the Tubu of Chad also have significant Eurasian ancestry that reflects deeper histories of Eurasian backflow into this part of Africa. Nevertheless, this must have been the fusion of the Zaghawa, Teda, and Daza with sedentary populations already in Kanem and the Chad Basin. Their fusion would ultimately lead to the Kanembu and Kanuri populations. Perhaps the use of the term Beriberi by the Hausa to refer to the Kanuri is actually an allusion to the Zaghawa, or Beri, peoples?

Unfortunately, Abdelkerim believes the Zaghawa language is a Berber one and its population of Berber origins because of Ibn Khaldun and other medieval writers. Since some of them have classified the "Zaghawa" as Berbers or mentioned their use of the litham, many scholars have uncritically repeated this. However, the alleged Berber influence was probably restricted and more likely than not simply a case of Berbers in the Sahara and Sahel who were assimilated into Tubu and Zaghawa populations. After all, even al-Idrisi cited a similar case of this in the Sahara during the 12th century. The Nilo-Saharan language family of the Zaghawa and all the available evidence from the Diwan and Kanuri oral traditions supports an early ruling dynasty of Teda, Daza, and Zaghawa origin. Even the name Dugu or Duku, remembered as Douk Bourme in Zaghawa oral tradition, is a further testament to Zaghawa origins. Even if the name as remembered by the Zaghawa meant courageous young man with light skin, this does not require a Berber origin. "Red" and "black" clans among the Tubu, for instance, could be what the Diwan was referring to when they identified the first "black" mai as only appearing in the late 12th century. 

In consideration of the available evidence, one can justifiably question whether or not there really was a dynastic change from the Duguwa to the Sayfawa in the 11th century. Abdelkerim agrees with scholars such as Lange who see dynastic change because of the language of one copy of the Diwan or a few other sources. However, since the Diwan, Ahmad b. Furtu, and Kanuri oral traditions and Magumi sub-clan divisions support overall continuity from the Duguwa to the Sayfawa, one can plausibly reject the theory. Even if it is true that the Sayfawa embrace of Islam was resisted by some of the Zaghawa ruling elite in Kanem, who had to be replaced by Humme, one is still left in the dark about why the Magumi royal clan of the Sayfawa retained divisions claiming descent from pagan Duguwa mais? Moreover, why was the last Duguwa mai remembered as Abd Abd al-Galil, a very Arabic name? We are inclined to believe that some of the 11th century mais of Kanem may have already embraced Islam, perhaps the Ibadi sect which was active in the nearby Fazzan? 

After all, centuries of trade and contact with Ibadi traders and even Berber merchants in Kawar would have facilitated the spread of Islam by the 1000s. Perhaps the Sayfawa were really just a more devout, Sunni Muslim branch of the ruling elite who saw greater benefit through the full embrace of Sunni Islam. Whether or not this actually precipitated a mass exodus or dispersal of Zaghawa to the east is unknown, as Kanem may have exerted influence on "Zaghawa" and "Daju" in the Chadian-Sudan borderlands. As for Islam's impact on the ruling dynasty and possible anti-Muslim resistance, perhaps it is important to note the late and slight Islamization of various Tubu and other populations in Kanem and Borno. If the masses were, for the most part, left alone, and some pre-Islamic rituals like coronation rites or certain beliefs in the king's power to influence fertility of the land persisted, then the conversion to Islam could have been accomplished in a manner that did not necessarily trigger too much resistance. Perhaps something comparable to various coronation rituals and harvest festivals later observed for the Keira sultans of Darfur was practiced by the early Muslim mais of Kanem.

If the rulers of Kanem had already, by this time, been accustomed to wearing fine textiles imported from trans-Saharan trade and interacting with Muslims for centuries, it is possible that a Muslim faction was already present in the court at Manan. Like Ghana in the Western Sudan, perhaps there were local and foreign Muslims already incorporated into the administration by this time. If our admittedly speculative theory is correct, then Kanem before the reign of the first Sayfawa mai may have had Islamic rulers or at least prominent Muslim administrators, traders, and teachers. Perhaps, although our only evidence is al-Bakri, a branch of the Ummayads may have fled to Kanem. Like the later recorded history of the Sayfawa, maybe members of the royal family already converted and began their study of Islam through the help of pious teachers like Muhammad b. Mani. Muslim traders established in Kawar and the Fazzan would have been an additional vector for Islamic propagation that might have appealed to the rulers of Kanem. Embracing the religion officially would have increased the stature of the kingdom to many of its Muslim trading partners and perhaps offered way for a new branch of the ruling family to build alliances with powerful allies. 

In conclusion, the Duguwa and Sayfawa were really one single dynasty. The Zaghawa or Beri origins can be deduced from the external Arabic sources. The "Zaghawa" in the Arabic sources likely included various related populations like the Teda and Daza. Over time, the Zaghawa kingdom of Kanem, which may have arose as early as the 500s in Lange's view, became known through trans-Saharan trade. A growing Muslim presence in Kawar and Kanem itself would have created conditions propitious for an eventual royal conversion. Since many of the Tubu and Zaghawa groups were nomadic or semi-nomadic and likely not forced to convert, Kanem's Islamization did not necessarily create a Zaghawa exodus to the east. The evidence for dynastic continuity throughout the history of Kanem-Borno may be further deduced from Magumi clan divisions claiming descent from various pre-Sayfawa kings. The incorporation of various sedentary populations already established in Kanem when the Saharan forebears of the Duguwa/Sayfawa entered the region were gradually incorporated into the state to the point where even non-Kanembu groups claimed descent from pre-Islamic kings like Bulu. Amazingly, references to the Zaghawa role in the origins of Kanem may be recalled in the Hausa appellation of Beriberi for the Kanuri. 

The Oasis of Salt

Knut S. Vikør's The Oasis of Salt: The History of Kawar, a Saharan Centre of Salt Production is required reading for anyone interested in the history of Kawar and Kanem-Borno. While suffering from a number of typos, this book actually surpasses the incomplete history of Kawar written by Marguerite Le Coeur. Unlike her work, Vikør reaches more level-headed conclusions and theories while also challenging Lovejoy on the economic and political dimensions of Kawar and the salt trade (slavery was marginal to salt production). Of course, as its focus is on Kawar, Kanem and Borno only appear in references to the Kanuri population of the region or in periods when direct authority of Kanem-Borno in the region was attested by various sources. Thus, Kawar is the center of this history, showing how a periperhal central Saharan oasis (and nearby Fachi and Jado) could be significant in the Sudanic salt trade as well as trans-Saharan contact. Kawar as a center-periphery is a key theme here, shifting in its reception of varying cultural influences based on the power dynamics of the Central Sudan and Sahara.

Vikør begins the study with an overview of Kawar, the geography, the relevant populations and the method of salt production. This precedes the second half of the book's chronological analysis, but is very helpful for contextualizing Kawar, Fachi and Jado. Then, the rest of the history analyzes Kawar from the time of Herodotus and the ancient Garamantes to the French colonial conquest. Le Coeur's earlier published history covers the same era, but does not rely on questionable assumptions (Le Coeur believed the Tomagra and Kaye were "Berbers"). In fact, Oasis of Salt is a very careful reading of Kawar's history based on our limited sources. Thus, moving from Herodotus and references to the Garamantes in the south from the Roman era, Vikør endeavors to cautiously interpret the various medieval Arabic sources from the 9th century to the late 14th. Moreover, for those able to read Arabic, an appendix includes the original texts with Vikør's translations. Needless to say, his interpretation of some of these Arabic sources is nuanced enough to give a more accurate impression of Kawar from the 7th century onwards. 

While we may never the full historicity of Uqba b. Nafi's "conquest" of Kawar or what exactly the famous alun of Kawar actually was, it becomes clear that Kawar was probably known or at least heard of by the Garamantes and probably an entrepot for trans-Saharan trade by the 9th century. The references to Kanem and Zaghawa or Zaghawi captives certainly suggests the Ibadi trade network had penetrated Kawar and likely introduced Islam into the region. This early Ibadi influence and links to the Fezzan are the reasons Vikør sees early Kawar as falling more into a Maghrebi or North African sphere of influence (exemplified in the ancient Arabic place names and the gassars, which seem to owe more to Kanuri or Sudanic influences but use an Arabic term), later shifting to predominantly "Sudanic" with the assertion of Kanem's control or interests by the 12th century. With the rise of Kanem (even under mai Arku, hints of Kanem's influence in Kawar and ties to the Tomagra and Dirku suggest early links), Sudanic civilization exerted its influence to a greater extent in Kawar and even the Fezzan. Then, despite the decline of Kanem by the late 1300s, the Bornoan state under the Sayfawas reasserted their claims to Kawar. To what extent Arku, Dunama, Ali Gaji, Idris Alooma, or other mais of Kanem-Borno were ever able to exert strong authority over Kawar is unclear, but surviving Borno mahrams and the Diwan suggest familial, tax exemptions, marital, and military strategies.

However, unlike Lovejoy, Vikør does not seem to see Borno's interest in Kawar as being connected to the trade in salt in the Sudan. Instead, Borno was more interested in securing Kawar for trans-Saharan trade to Tripoli, via the Fezzan. Of course, it is hard to imagine Borno not trying to take advantage of Kawar's salt production, as the Koyam were said to have been involved in it and perhaps the Bornoan state added Kawar's salt exports to local natron production in Manga and near Lake Chad to supplement their trade with Hausaland and beyond. In other words, even if trans-Saharan trade and the Kawar salt caravans usually operated independently, it is difficult to imagine Borno not actively promoting Kawar's salt trade. This does not exclude the possibility of Tuareg caravans carrying salt from Kawar to Hausaland, which the Kano Chronicle may suggest was occurring by the mid-15th century. Indeed, the ensuing conflict between the Tuareg of Ayar and Borno in the late 17th century and 18th century, even if it was often small-scale raids in Kawar, does suggest Borno's interest in Kawar were important for security of trans-Saharan trade but perhaps the profitable salt trade, too. Indeed, Tuareg incursions on the borders of northern Borno plus Tuareg settlements of enslaved farmers in Damergu must have been seen as as combined threat to Kawar and Borno's direct borders. The Koyam must have been threatened too, by the rising importance of Tuareg, specifically Kel Owey salt caravans to Fachi and Kawar. 

Overall, this is mandatory reading for anyone interested in Kawar, Kanem-Borno, or the Central Sudan. While there are some arguments of the author we disagree with or find implausible or lacking adequate backing, Oasis of Salt is so far the best read on Kawar's history. Lovejoy's work is on the larger Central Sudan and should be read in conjunction, and Le Coeur's work, based on field experience and exposure to the Teda of Kawar and Tibetsi, should likewise be added to one's list. Taken together, the three provide a great overview and list of sources on Kawar's historic importance and role as a bridge between Kanem-Borno and the Mediterranean. We disagree with the author about the Kanuri of Kawar being the remnants of an original proto-Kanuri population that once lived further north, but he's likely right about the population always being, mostly, Sudanic. We would like a bit more details on the clan systems of the Teda and Kanuri before accepting his theory, though it does seem that early Kawar's population was closely related to the Kanembu and Kanuri. In addition, he's almost certainly correct about the Zaghawa/Banu Duku to Sayfawa "dynastic" change, too. It is far more likely that the Sayfawa dynasty and "Zaghawa" represented two local lineages rather than Hume being a Berber outsider. But based on our limited sources, Vikør does not agree with Lange that there definitely was a shift in dynasties from pre-Islamic Kanem to the Sayfawa. In light of his generally careful scholarship and useful references and translations, this is an enlightening history of a Saharan region whose entire existence is predicated on trade. 

Black Nomads of the Sahara

Although only about a third of Jean Chapelle's ethnographic study of the Tubu peoples, Nomades noirs du Sahara: les Toubous, was particularly useful for our purposes, we highly recommend it as a synthesis of various colonial-era sources on the Teda-Daza peoples of the Sahara and Sahel regions. Despite Chapelle's personal experience on the ground in Chad, many chapters draw on earlier studies by colonial officers who collected local traditions which formed the basis for this books conclusions and historical outlook. It is undoubtedly a dated work and reflects, perhaps, some of the colonial obsession with the "mixed" racial origins of the Tubu. Nevertheless, it was a refreshing read for its acknowledgement of the Tubu origins of the Sayfawa dynasty and the role of constant migrations and shifting lifestyles (from nomads to sedentary agriculturalists) across the history of the region. We only wish more was said about clans of Tubu origins in Kanem and Borno and the ways in which Kanembu and Kanuri communities have continued or inherited Tubu practices. Nonetheless, if the ethnographic present can be used to fairly deduce how the Saharan ancestors of the rulers of Kanem lived several centuries ago, this book might give one an idea of how early Kanem could have looked with a "Zaghawa" ruling dynasty and sedentary farmers in the Kanem region paying tribute and engaging in reciprocal exchange. Like the Tubu chiefs and Derdes of more recent times, perhaps the early mais were associated with religious rituals on behalf of the clan, including rites attached to rain, harvests, and sadagas.