The Song of Muhammad al-Amin b. Muhammad al-Kanemi

The first page of Denham's English translation of Muhammad al-Amin al-Kanemi's Q. daliyya Nasim al-saba arajta rab'an bi-Kukawa wa-shahadta man bi ' l-babi min mutaraddidi

Although we read it a few years ago in the work of Denham and Clapperton, it somehow escaped our attention until recently. We are speaking of, naturally, an Arabic poem penned by Muhammad al-Amin al-Kanemi, the founder of the Shehus who have ruled Borno after the fall of the Sayfawa. In this case, we found the poem by al-Kanemi intriguing for its resemblance to Kanuri praise songs from the period of the Sayfawa dynasty. In this case, one sees a song or verse poem about the martial prowess and victory of al-Kanemi against Bagirmi (and other unnamed foes). This, naturally, brings to mind Kanuri praise songs for the mai, yerima or galadima for similar military victories or skills. Some of the metaphorical language used by al-Kanemi is also quite rich and draws on what must have been some of the standards of Borno Arabic poetry and oral literature. For instance, references to "forests of spears" when referring to his large army is quite evocative or to the hyena bring to mind Kanuri praise songs referring to lions, bush cows, and the rogondimi snake. Moreover, metaphors and similes comparing al-Kanemi's favorite mistress, rescued during this lengthy campaign (over 6 months away from Kuka, or Kukawa) to Indian silks are rather well-done. Indeed, this personal element in which al-Kanemi celebrated this reunion with a lover is a personal dimension to Borno's written literature that is not easily found. This makes it very much a personal statement even as it celebrates the contributions of other military chiefs or officials like Tirab. Last, but certainly not least, the song from 1821 refers to the palace of the Bagirmi sultan facing an attack from al-Kanemi's troops. This is another piece of evidence for Borno's role in the sacking and pillaging of Massenya, the Bagirmi capital.

The Tuareg and Ali b. al-Hajj Dunama

Map from Studies in the History of Pre-Colonial Borno

One key factor of Borno’s 18th century decline during the “Late Sayfawa Period” was the loss of Kawar, a major salt production center and part of the trans-Saharan trade routes. During the long reign of Ali b. al-Hajj Dunama (r. 1747-1792), Borno not only suffered defeat at the hands of the Tuareg at Bilma in c. 1759, but some of Borno’s vassal states and regions suffered Tuareg depredations, particularly areas like Gaskeru under the shaykhs of the Koyam. Lovejoy, who has written extensively about the salt trade in West Africa, has relied heavily on Palmer (and, via Palmer, Jean) for the basic narrative of the Tuareg seizure of the Kawar salt trade to Hausaland. However, this momentous development in the 18th century Central Sudan is poorly documented. Besides passing references to Kawar and a conflict or war in Bilma in c. 1759-1760 in the Chronicles of Agadez, Palmer (and Jean) presumably drew on Tuareg oral traditions to show resistance from Borno to Kel Owey and other Tuareg acts of aggression. Intriguingly, however, oral traditions from Borno do not seem to refer to wars with the Kel Owey or Agadez Sultanate over Bilma. Instead, traditions collected from the Koyam refer to a disastrous battle in which forces from Gaskeru and additional warriors sent from Birni Gazargamo were massacred by the Imakiten Tuareg at Kaiguem, an area north of Kelle in Koutous. We propose a return to the sources and oral traditions to better understand relations between Borno during Ali b. al-Hajj Dunama’s reign and the Tuareg associated with Aïr. It shall be argued that Borno’s conflict with the Tuareg during the reign of Ali b. al-Hajj Dunama cannot be understood as a single, decisive war for Kawar, but rather as a series of fragmented commercial and military struggles involving multiple Tuareg groups, local intermediaries, and uneven Borno intervention. Beginning with the written sources from the time period in question, we will then attempt an analysis of the various traditions on this period recorded since the 19th century. It is hoped that the connection between the Koyam conflict with the southeastern Tuareg was indeed tied with the Kel Owey commercial rise in Aïr and Kawar.

Review of Precolonial Sources

Beginning with the main written sources, a number of 18th century or early 19th century texts refer to the Tuareg and the salt caravan to Hausaland. In addition, the Chronicles of Agades alludes to conflict with Borno or in Kawar in the second half of the 18th century. According to the French translation of Urvoy, the “combat” of Bilma took place in 1760.[1] Another conflict erupted in Bilma or Kawar during the affair of Dirfass in 1777-1778. The same collection of Agadez chronicles also contain earlier references to conflict with Borno in the 17th century, such as a possible war in 1685 and the Kel-Owey Tuareg attacking Borno in 1679.[2] These references to conflict in Kawar or with Borno are, unfortunately, very thin on the details. It is possible that the sultans of Agadez were not directly involved in instigating the conflict with Borno or the Tubu for Kawar, meaning that their royal chronicles shed less light on the conflict than Kel Owey oral traditions. But, according to Hamani, the Agadez sultans Muhammad Humad and Muhammad al-Udala (r. 1768-1792) were allies of the Kel Owey.[3] 

Besides the Agadez sources, written sources from Borno do, if indirectly, refer to conflict during the second half of the 18th century. A mahram dated by Palmer to 1752 referred to a failed attempt by the mai to enlist the Tura for war. When the Tura threatened to leave Borno and return to Dirku in Kawar, the Sayfawa ruler relented and reaffirmed their privileges.[4] While this predates the c. 1759 aggression in Bilma, there must have been some threat or campaign important enough to tempt the mai to enforce military service from a privileged group. A later mahram dated to 1785 also affirms the town of Laluri as belonging to the Beni Mukhtar Tura.[5] More written sources, such as poems and texts on Islamic rulership, the meaning of the Mune, and other texts exist, but not in translation. The sense appears to have been one of insecurity and conflict with various neighbors or vassal states. Indeed, a 19th century Mandara kirgam may be obliquely referring to this troubled political climate when elucidating how the motive of Ali b. al-Hajj Dunama’s aggression was Mandara’s resources.[6] This is certainly plausible as the Mandara campaign of 1781 took place after Borno’s loss of Bilma. Perhaps eager to reassert Borno’s status in the Central Sudan, the Sayfawa mai also wanted to acquire extra resources to make up for the losses to the Tuareg.

Salines of Bilma (Abadie).

In addition to these aforementioned sources, a number of European 18th century descriptions also establish that the Tuareg had achieved a monopoly of the trade in Kawar’s salt to the markets of Hausaland in the second half of the 1700s. For example, the journal of Hornemann referred to the Kel Owey Tuareg’s dominance of Asben. The same source referred to the Koyams (“Kojams”) living to the north of the principal town of Borno (Gazargamo) so his account was presumably written before the final abandonment of Gaskeru. More importantly, Beaufoy is cited for the salt caravan from Agadez to Domboo.[7] Thus, Hornemann’s journal, written in the 1790s, clearly establishes a Tuareg (and specifically Kel Owey) ascendancy, as well as the ties to Agadez for the trade in Kawar salt. Carsten Niebuhr, writing in German and basing his information on the Sudan from talks with Tripolitanian envoy and his African slaves from the Central Sudan, reported that Afnu (Hausaland) gets its salt from Asben. Niebuhr also referred to the Tuareg by the Kanuri term for them, indicating one of his informants was Kanuriphone or came from the Borno region.[8] While some of the salt from Asben’s own salines likely contributed to the supply of desert salts to Hausaland, Niebuhr’s information was collected in the 1770s. Consequently, it is possible that Kawar’s salt production was also being exploited by Tuareg traders in Katsina, Kano, and other Hausa cities. The use of the Kanuri term for the Tuareg also shows the degree to which the Tuareg were a known and identifiable group in 18th century Borno, possibly through their raids and pillaging of Kutus, Muniyo, Koyam centers, and other parts of Greater Borno. Last, but certainly not least, North African and Fezzani informants in the late 18th century reported that Borno still exported salt (perhaps natron is meant), but the people of Agadez controlled the salt caravans from Domboo to Cashna (Katsina). Supposedly, the Tuareg only paid brass and copper to peasants or oasis dwellers of Kawar for their salt. The overall picture suggests the Agadez merchants controlled the trade in salt from Borno (Kawar) to Katsina.[9]

This sense of Tuareg control of the Kawar salt trade to Hausaland is confirmed in a number of 19th century writings by European travelers to the region. Richardson, for example, noted that the Aïr Tuareg were in possession of the Bilma salt caravan. Moreover, a “short time ago,” Borno destroyed or stole the supplies used by this caravan, causing the loss of hundreds of camels.[10] Since Richardson did not give a year for this act of Borno’s sabotage, we can only assume that well into the 19th century, Borno tried to impede the Tuareg traders. Heinrich Barth also observed the history of Tuareg conflict with Borno and the salt caravan. According to Barth, the Kel Owey had an alliance with the people of Azanéres that was linked to the Bilma salt exchange. Furthermore, Barth believed this trade of the Asben-affiliated Tuareg trading Bilma’s salt to Hausaland only began about 100 years ago.[11] This agrees with the Agadez chronicle of the Tuareg of Aïr engaging in a war or battle at Bilma in c. 1759. Elsewhere in his account of his travels, Barth also referred to Ali b. al-Hajj Dunama’s violent struggles with the Tuareg in the middle of the 18th century: “He made great exertions in every direction, but his efforts seem to have resembled the convulsions of death…”[12] Although he confusingly referred to the mai as Ali Omari, recalling Ali b. Umar of the 17th century, it’s quite clear that Barth was referring to Ali b. al-Hajj Dunama. He even named his successor correctly as Ahmad. That this period was a turbulent one in Borno’s history can also be found in Barth’s allusion to Ali b. al-Hajj Dunama’s campaigns against the Bade (Bedde).[13] Nachtigal, for his part, wrote of the responsibilities of the yerima including oversight of the southeastern Tuareg groups while the galadima was responsible for the Bade and western Borno. Thus, during the conflicts with the Tuareg in the 18th century, the yerima was presumably the first official meant to respond to the crisis. This is perhaps contradicted by oral traditions of the Koyam, however, who remembered a kaigama named Ali, allegedly a son of Ali b. al-Hajj Dunama, who came to their aid when their community was attacked by the Imakiten Tuareg at Gaskeru.[14] Clearly, a major dimension of conflict with the Tuareg during this period involved Koyam and other groups living outside of Kawar.

Map of Hausaland, Asben and Borno by a Hausa mallam (in Hornemann's Journal).

Revisiting Colonial-Era Scholarship

Moving into the colonial period, more precise traditions of conflict between the Tuareg and Borno were recorded. Jean, whose Les Touareg du sud-est. L'Aïr. Leur rôle dans la politique saharienne was published in 1909, was the first to explore what actually happened in c. 1759. Unfortunately, his account is garbled and may mix events associated with Borno’s failed siege of the Tuareg forces at Bagzam with events in the 18th century. In Jean’s retelling, the Bornoan forces abandoned their siege of the Tuareg at Bagzam and, when fleeing to the east, were pursued by the Kel Owey to the wells of Ashegur. There the Kel Owey defeated the forces of Borno, took hundreds of prisoners, and established them at Bilma.[15] It is difficult to say to what extent this tradition is accurate, but it does correctly identify the Kel Owey as protagonists in this conflict. Later, the sultan of Agadez and the Kel Gress went to Fachi and Tuareg suzerainty of the region was established. But, in c.1765, Kel Owey traders were massacred in Borno. This prompted a counterattack in which the sultan of Agadez, Mohammad Almoubari invaded Borno and seized cattle. Supposedly, the Tuareg marched almost to Kuka (Gazargamo) and ensured free access to the area for Tuareg traders.[16] A number of problems exist with this narrative, despite Palmer, Lovejoy and later authors repeating it. If the named sultan is correct, then this was probably Muhammad al-Mubarek, a powerful sultan of 17th century Agadez. Various sources refer to specific wars or battles with Borno during his reign. On one of those occasions, forces from Agadez did reach the capital of Borno. But, that attack occurred in the 1660s, not the 1760s. Besides, the sultan of Agadez in c. 1765 was Muhammad Humad. While Rodd identified him as the son of Mubarak, Hamani’s genealogy of the kings of Agadez names his father as Usman.[17] That some type of conflict likely emerged may be seen in the allusion to the Kel Owey as the Tuareg group targeted by Borno’s violent massacre. In this context, their presence in Borno was due to people of Bilma demanding supplies from Borno. Once there near Lake Chad, they were poorly received and most were killed. The specific reference to the Kel Owey, the sultan of Agadez (at the time, an ally of this group) and the bravery of the Kel Férouan in the expedition against Borno are details that could very well refer to the 1760s.  Furthermore, that the sultan of Agadez at the time was an ally of the Kel Owey reinforces the idea that the sultan may have sent forces to aid the Kel Owey against Borno. 

Despite his garbled reconstruction of Tuareg oral tradition, the narrative of a 1765 invasion from Aïr that reached just before Gazargamo has been repeated by subsequent scholars. The unavoidable Palmer, in Bornu Sahara and Sudan, essentially repeated Jean. Thus, the War of Bilma began in 1759 and in that conflict, Kel Owey achieved victory against Bornoan soldiers. Again, the 1765 reference to the massacre of Kel Owey in Borno and the response from the Agadez sultan are presented uncritically.[18] As Palmer did, other colonial-era scholars largely repeat this narrative or merely stress different aspects of Tuareg tribute demands or the workings of the salt trade. Abadie, as one colonial-era example, merely referred to the Tuareg domination of Kawar probably beginning before the Tubu presence but had to admit the Kanuri traditions remembering a Tubu presence first.[19] In his "Tarikh El Khawar," Le Sourd reported that the Kanuris of Guezebi-Guassar sent tribute to the Kel Owey. Yet prior to that relationship, they paid tribute to the “Temaghra” Tubu. Further, the Koyam were said to have once lived in villages between Agram and Fachi in the distant past.[20] This reference to the Koyam in Kawar and their ties to its trade before the Tuareg may be implied by the mountain east of Bilma’s main mountain, called Diotko, after the Diotko of Borno. Abadie was told the Diotko, or Jetko, were said to have been traders of salt and dates in the region.[21] The Jetko, like the Koyam, were a pastoralist group related to the Koyam. Urvoy, on the other hand, provided some key information on the Tuareg groups involved in Kawar and to the lands south of Aïr. For instance, his Histoire de l’Empire du Bornou mentions the Amakitan Tuareg in Koutous, demanding tribute from sedentary farmers, yet the Ikaskazane of the Kel Owey were installed to the west of the Amakitanes, with a center at Garazou in Alakoss.[22] It would be worthwhile to further explore the extent to which the Kel Owey and Amakitan groups coordinated their actions or engaged in joint raids, especially if directed against Koyam and Jetko traders in Kawar with links to Borno’s trans-Saharan and desert salt trade.

The area of southeastern Niger to Lake Chad (Maikorema Zakari).

Unquestionably, this memory of Koyam or Jetko traders active in Kawar was surely another aspect of the Koyam conflict with the Tuareg during the 18th century. Unsurprisingly, this is precisely where the Koyam traditions described by Landeroin are the most detailed on Tuareg conflicts. In fact, conflict with the Imakiten (and likely other Tuareg) began as early as the 17th century, when Kalumbardo was destroyed by Tuareg raiders. The 18th century settlement at Gaskeru, under the effective leadership of shaykhs descended from the founder of the earlier Kalumbardo, was probably connected to traders in Kawar. Hence, it is likely not insignificant that the only conflict with the Tuareg reported by Landeroin was with the Imakiten Tuareg of Koutous who attacked Gaskeru with the aid of other Kel Aïr:

Sous le commandement d’Abdoullaÿ, frère cadet de Beker, les Touaregs Imakiten du Koutous, aidés des Touaregs de Azbin, vinrent attaquer Gaskérou. Abdoullaÿ demanda du secours au Maï Ali, qui lui envoya son  fils, le kaïgama Ali. Ensemble, ils repoussèrent les Touaregs jusque dans leur pays. Mais, ayant voulu pénétrer dans l’intérieur de celui-ci pour reprendre les captifs et les troupeaux razzés, les pillards firent front avec toutes leurs forces et au combat de Kaïguem, au Nord de Kellé, les Koyams furent massacrés en grand nombre.[23]

Perhaps even more significantly, the mai sent his kaigama, the top military official, to aid the Koyam against the Tuareg. When they pursued their enemy to Kaiguem, north of Kelle, the Borno forces were massacred and the shaykh, Abdoullay, was killed. This detailed account of utter defeat is, of course, referring to the Imakiten Tuareg who regularly raided parts of Borno. Nevertheless, it is possible this event took place before 1781 since a kaigama named Ali died in that campaign.[24] If the kaigama remembered in tradition was the same man, Abdoullay may have died in conflict with the Tuareg before 1781. To verify this, of course, would require a gargam or more precise genealogies or successions for the kaigamas of 18th century Borno. Overall, this tradition of conflict with the Tuareg supports the notion of how complex and multivalent the relationship with the Kel Aïr Tuareg was at this time.

Thoughts on Postcolonial Scholarship

With the postcolonial era, more precise studies of Kawar, Borno, and the Tuareg of Aïr appeared. Lovejoy, whose detailed study of the desert salt trade drew heavily on Jean and Palmer for Borno’s conflict with the Tuareg, also provides some dates for a Tuareg attack at Gaskeru (c. 1787-1792) before it was finally abandoned.[25] Maikorema, on the other hand, reported the Jetko of Yari Jetko were attacked by the Tuareg, so they moved to Tal Bul, north of N’Gigmi.[26] This is in concordance with attacks on the Jetko and Koyam during the 18th and 19th century by Tuareg bandits. Scholars such as Muhammad Nur Alkali likewise noted the importance of the loss of Kawar for Borno’s economy. Furthermore, the growing independence of the vassal rulers of Muniyo and the Daagirawa, based on an alliance against Tuareg raiders, may have been another sign of Borno’s failure to provide the necessary security for its client states in the 18th and 19th centuries.[27] 

Vikor's map of the Tuareg Salt trade in The Oasis of Salt.

Vikor, in The Oasis of Salt, provides the best synthesis for the Kawar or Fachi-centered aspect of the conflict with the Tuareg. Thus, the familiar tale of the Tuareg defeating Borno forces at the well of Ashegur north of Fachi is maintained. This victory was achieved with the aid of the sultan of Agadez. Additionally, the Itesan held Fachi while the Kel Owey chased the Bornoan forces to Kawar. But, and here is a key difference, Vikor stressed the role of Teda raiders in attacking the Tuareg. One such attack took place in 1768-9, when Isandalan Tuaregs were attacked at Jado. Another Teda attack on the salt caravan led to a major battle in 1777, where the Tuareg leader was named Dirfass.[28] This notion of the Tubu playing a disproportionate role in these conflicts with the Kel Aïr was also believed by Djibo Hamani. For his part, Hamani viewed this period as one in which Borno’s rulers were less involved in the affairs of their vassal states on their western border, areas often raided by the Tuareg: Muniyo, Kutus, Damargu. In Kawar, the Tuareg were seen by Hamani as clashing with the Tubu in 1759, 1768/9, and 1778.[29] Though Vikor recognized the role of Borno in the conflict of 1759, Hamani seems to prefer contextualizing it as one of Tubu raids on Tuareg traders or pilgrims, thereby causing counterraids or battles. Such a perspective, however, ignores the role of Borno in the earlier conflict and the tradition of a Kel Owey group massacred in Borno in c. 1765. Even if that tradition has been problematically reported by Jean, there seems to have been some clear Borno involvement in the fight with the Tuareg. Such a view can find quick support in Borno’s support for the Koyam shaykh, Abdoullay, during this period.

In the study of Aïr’s political economy during this era, Kathleen O’Mara’s dissertation presents a compelling case for the commercial and political ascent of the Kel Owey. In her study, the Kel Aïr alliance won a battle at Bilma in 1759/60. The rights to tribute and monopoly over Kawar’s salt trade went to the Kel Owey and Kel Geres. The Sarkin Turawa of Agadez also received taxes or tribute from a bulama of Kawar. Lastly, she argued against Grandin’s theory of the Kel Geres controlling Bilma in the 1500s, finding no evidence for such an assertion.[30] More importantly, her materialist analysis of the Kel Aïr in the late 1700s and early 1800s emphasizes how the seizure of Kawar’s salt trade was an impetus for agro-pastoral expansion. With the growth of a Kel Owey commercial class investing in salt, agricultural production, livestock, and trade in textiles, slaves, leather products, and items acquired through trans-Saharan trade via the Fezzan or Ghat, the Aïr Tuareg system became a fully tributary one that maintained the dominance of the "nobles."  In fact, the continuation of the sultanate structure in Agadez as an intermediary of Tuareg groups in Aïr, plus their own source of legitimacy via Islam, provided a balance with Kel Owey elites who rose to dominance during the 18th century. Finally, the growing economic integration with the Hausa states to the south and, eventually, the Sokoto Caliphate, contributed to a southward migration as more Tuareg elites began to rely on the labor or tribute from servile or free dependents laboring in agriculture. This process must have fueled further Tuareg settlement in lands to the south, in addition to the ecological pressure during periods of extended droughts in the Sahel. As one can likely imagine, this process among the Kel Aïr of the second half of the 18th century would have included the Amakitan and Kel Owey groups. 

Conclusions and Evidence 

As the evidence suggests, Borno’s conflict with the Tuareg during the lengthy reign of Ali b. al-Hajj Dunama was complex. The battle for Kawar was merely one (and an important one) aspect of this era, the Late Sayfawa Period. The fissures in the imperial system of the Sayfawa state were widening as vassal states rebelled, Borno failed to provide security for others, and internal contradictions and conflicts fueled dissension. Borno’s response to the disastrous 1759/60 war over Bilma revealed how significant it was to the state’s trans-Saharan and Sudanic economy. By also responding to attacks on the Koyam at Gaskeru, the Sayfawa were likewise acting to defend their frontier inhabited by groups who were also tied to Kawar and the trade in desert salts. This means that Borno’s response was based on the goals of protecting its own economic, political, and military interests. That the Bilma salt trade was of such importance can be seen in the dynamic rise of the Kel Owey Tuareg during this time, whose burgeoning power also encouraged further Tuareg depredations and settlements in lands south of Asben. That conflict extended to Borno’s peripheral tributaries such as Muniyo and Koutous is linked to the growth of Tuareg settlement to the south, itself facilitated by the Kawar salt trade and the Hausa markets of centers like Kano and Katsina. To properly analyze this period, one must interrogate closely all sources, particularly the problematic oral traditions (or the debased or elaborated forms reported in colonial sources).

[1] Yves Urvoy, “Chroniques d’Agades,” 161.

[2] Ibid., 160, 162, 170

[3] Djibo Hamani, Le sultanat touareg de l'Ayar: au carrefour du Soudan et de la Berbérie, 261.

[4] H.R. Palmer, Bornu Sahara and Sudan, 46.

[5] Ibid., 47.

[6] Eldridge Mohammadou, Le royaume de Wandala ou Mandara au XIXè siècle, 44.

[7] Friedrich Hornemann, The Journal of Frederick Hornemann's Travels, from Cairo to Mourzouk: The Capital of the Kingdom of Fezzan, in Africa, 113, 180-182. 

[8] Carsten Niebuhr, “Das Innere von Afrika", 981, 986.

[9] Association for Promoting the Discovery of the Interior Parts of Africa, Proceedings of the Association for Promoting the Discovery of the Interior Parts of Africa, Volume 1, 154, 157, 167.

[10] James Richardson, Travels in the Great Desert of Sahara in the Years of 1845 & 1846 Containing a Narrative of Personal Adventures During a Tour of Nine Months Through the Desert Amongst the Touaricks and Other Tribes of Saharan People; Including a Description of the Cases and Cities of Ghat, Ghadames and Mourzuk, Volume 2, 143.

[11] Heinrich Barth, Travels and discoveries in North and Central Africa Including accounts of Tripoli, the Sahara, the remarkable kingdom of Bornu, and the countries around Lake Chad, (1890), 154, 158.

[12] Ibid., Travels and Discoveries in North and Central Africa Including accounts of Tripoli, the Sahara, the remarkable kingdom of Bornu and the countries around Lake Chad, Volume 2 (1857), 281.

[13] Ibid., 599. 

[14] Landeroin, “Notice historique” in Documents scientifiques de la Mission Tilho, 399. 

[15] C. Jean, Les Touareg du Sud-Est: l'Air ; leur rôle dans la politique saharienne, 119-120.

[16] Ibid., 121.

[17] Rodd, People of the Veil, 415, Hamani, 452. 

[18] Palmer, Bornu Sahara and Sudan, 86.

[19] Maurice Abadie, Afrique centrale: La colonie du Niger, 177.

[20] Michel Le Sourd, "Tarikh El Kawar,: Bulletin de l'I.F.A.N VIII, (1946) 6, 28.

[21] Abadie, 5.

[22] Yves Urvoy, Histoire de l’Empire du Bornou, 91. 

[23] Landeroin, 399.

[24] Mohammadou, 55.

[25] Paul E. Lovejoy, Salt of the Desert Sun, 229.

[26] Maikorema Zakari, Contribution à l'histoire des populations du sud-est Nigérien Le cas: Mangari (XVIe - XIXe s.), 83.

[27] Muhammad Nur Alkali, Kanem-Borno Under the Sayfawa: A Study of the Origin, Growth, and Collapse of a Dynasty (891-1846), 268, 314.

[28] Knut S. Vikør, The Oasis of Salt, 212, 216.

[29] Hamani, 267.

[30] Kathleen O’Mara, A Political Economy of Ahir (Niger): Historical Transformations in a Pastoral Economy, 1760-1860, 114, 124, 343.

Brick Architecture in Massenya (Bagirmi)

Heinrich Barth's plan of Massenya, the Bagirmi capital.

Although the kingdom of Bagirmi also used fired brick, this part of Chad does not appear to have received as much analysis as sites in Kanem and Borno. This is unfortunate since details about the arrangement of the royal center in Massenya and the use of fired-brick for walls as well as the porio may provide clues for understanding earlier sites in Kanem. Moreover, understanding the arrangement of the space and its larger cosmological significance in Bagirmi could very well reveal something about the nature of space, royal palace layouts, and cosmology in Kanem and Borno. In our own very limited and exploratory manner, we will attempt a brief overview of Massenya's history and the use of fired brick in its elite constructions. The tentative analysis here is merely one of raising questions and exploring the deeper political and social meanings connotations of brick constructions in this part of the "Central Sudan."

A view of Massenya based on Barth.

A quick commentary on the sources is in order. Researching the history of Bagirmi, particularly its history before the 19th century, is an arduous task. Heinrich Barth, Muhammad Bello, Escayrac de Lauture,  Nachtigal, al-Tunisi, and Denham wrote about the region in the 19th century. Before them, one finds occasional references to Bagirmi from other sources, both European and African. Unfortunately, most of the pre-1800 sources are very vague or imprecise, often only referring to Bagirmi's geography or mentioning its role in provisioning captives for the trans-Saharan trade. That Bagirmi was at least used as a name for the region by the 15th century can be seen in the map of Fra Mauro. Furthermore, references from the Diwan and the chronicle of Ahmad b. Furtu (as well as Anania's writings) refer to Bagirmi. Fortunately, the travelers, colonial "scholars" and administrators, and anthropologists and historians of the 20th century have contributed greatly to a fuller picture of Bagirmi's history. Nonetheless, the loss of a Barma equivalent of the Kanuri girgam, and the absence of written chronicles has impeded the development of a more complete history. This has forced Lebeuf, Reyna, Pâques, Gaden, Lavers, Lanier, Devallée, and others to compare lists of kings, look to oral tradition recorded at different times, explore Barma cosmology, and, in some cases, write about Fulani Muslim scholars based at Bidderi or other parts of the kingdom.

Pâques, Le roi chasseur

As for the  possible deeper antiquity of a kingdom of Bagirmi before the 15th or 16th century, as once proposed by Lange, the evidence is too ambiguous to say. The general chronology of Bagirmi's kings, with dates based on Nachtigal's interviews with Bagirmi elites, appears to be the most accurate, and that places the dynasty's origins back to the 16th century. We are inclined to favor such a date, with the period before the rise of the dynasty of Bagirmi as one of Bulala domination over a heterogeneous population (which included Fulani and other groups). In fact, the Bagirmi tradition on the foundation of Massenya's royal palace suggests the kingdom had already been in existence for at least a generation and Fulani Muslim clerics were highly esteemed. Therefore, a 16th century origin for Bagirmi as a singular kingdom is likely. Since Nachtigal relied on informants from Bagirmi and his list omitted names remembered by other informants (as reported by Escayrac de Lauture), there is a possibility that the dynasty's origins began earlier than Nachtigal's chronology suggests. For instance, Nachtigal was told that it was Lubatko who ordered a hedge of thorns placed around Massenya (Nachtigal 401). This mbang was said to have reigned from 1536-1548. Only under Abdallah (1568-1608) was the royal dwelling walled, with a plan by 4 Fulani Muslims named Dede, Shille, Aji Amede, and Majangala (Nachtigal 403). 

Besson's plan of old Massenya's palace complex.

Yet Escayrac de Lauture, writing decades earlier on Bagirmi's history, reported that a large number of brothers of Birni Besse ruled Bagirmi before Malo, the son of Birni Besse, ascended to the throne (Escayrac de Lauture). If Malo, who was said to have become mbang 28 years after his father, was preceded by a number of uncles (only one of whom is mentioned by Nachtigal, Lubatko), then Nachtigal's list omitted several kings. It also raises questions about the nature of succession in early Bagirmi. If it was practiced collaterally until the rise of Malo, and 5 kings ruled in between Birni Besse and Malo, the kingdom may have begun a little earlier than 1522, assuming that some brothers may have been significantly younger than Birni Besse. Of course, Escayrac de Lauture had access to fewer sources than Nachtigal, so all of his information on the kings of Bagirmi must be used cautiously. This nonetheless illustrates how difficult it can be to construct a coherent chronology without additional sources to corroborate or date events more firmly.

Vestiges of a wall in Old Massenya (from Lebeuf).

What of the city of Massenya? The sources disagree on some key details, but are largely compatible. Pâques attributes the foundation of Massenya to Dala Birni (Pâques 17). This is interesting since Nachtigal named Dalo Birni as a brother of Birni Besse. Most sources (Lavers, for example) concur on this point with Nachtigal. Mbang Bolo was said to have built Massenya and the royal palace by Lanier, but this may be a reference to Malo or perhaps another name of Abdallah (Lanier 460). In terms of Massenya itself, Nachtigal believed that Malo had a hedge placed around Massenya. Then, his brother and successor, Abdallah, built a wall around the royal palace. But only in the 17th century, during the reign of Burkomanda I (1635-1665), was a brick wall built to surround the gur, or palace (Reyna 62). Nachtigal, Reyna, and Devallée all associate Burkomanda with the first brick walls around Massenya's palace. 

Plan of Old Massenya in Lebeuf.

This early date, assuming the regnal years of Nachtigal are accurate, point to the 17th century for the first fired brick walls in the town. Indeed, when Heinrich Barth visited Bagirmi in the 1850s, he saw the brick wall of the palace was in a state of decay and probably more than 100 years old (Barth 518). This would suggest the wall was built by the 1750s at the last, and an earlier date in the 17th century is not impossible.  Later on, the town walls of Massenya were rebuilt during the reign of Wanya (1722-1736). Another 18th century mbang, Loel (or Loen, 1741-1751), was said to have added an additional room in the palace complex for his 500 horses and began to employ 1000 slaves as a royal bodyguard force in the royal square (Nachtigal 407-408). Lastly, the mbang named Haji (1751-1785) was remembered in some sources for promoting the use of fired brick at Bousso (Pâques 200). Haji was also known for promoting Islam in Bagirmi, which including the use of bricks for mosques. This did not stop him from endeavoring to limit the use of brick for elite structures outside the capital (Seignobos 276).

Ruin of part of the fired brick wall for the palace (Lebeuf). 

Unfortunately, Massenya had been sacked and pillaged on a number of occasions in the 19th century. Consequently, Barth's description of the fired brick wall around the gur or his brief mention of brick buildings outside Massenya are based on a visit to the site after at least 2 periods of pillaging or invasions. The first, when Sabun of Wadai invaded Bagirmi and took Massenya, involving destroying part of the town and palace walls (El-Tounsy 158, 162). An attack from Borno, led by al-Kanemi, was likewise said to have resulted in much destruction. Lanier attributes the destruction of the palace to this Bornoan invasion as Massenya was once again pillaged (Lanier 462). Despite these attacks, the city of Massenya was still described as a very large town by a variety of sources. Aboubakar, a Bagirmi native living in Brazil during the 19th century, described the capital "une tres grande ville" (Comte de Castelnau 31). Seetzen, writing from Cairo and basing his writing on interviewing a native of Afade, reported that Bagirmi "c'est une ville considerable" (Seetzen 166). Even buildings of two or 3 storeys were said to exist in Massenya when Escayrac de Lauture collected information on the kingdom (Escayrac de Lauture 91). Whilst Escayrac de Lauture's information is perhaps an exaggeration, one senses that Massenya was a large town. Even when, by the second half of the 19th century, when Massenya was known for having a large circumference but small number of houses and people for its size, it covered a large space (Nachtigal 423). It is very likely that the royal capital was more densely populated in the past before the kingdom's capital was sacked and Bagirmi became a tributary of Wadai.

Ruins of the old houses in Massenya (Lebeuf)

We are fortunate to know something about the nature of the gur of Massenya. Its walls were 10 feet at the base, 15-20 feet high (Reyna 62). Reyna, who drew this observation from Barth, who wrote that the palace wall was 18 feet high, 10 feet thick and built of decaying baked bricks (Barth 516). The town's circumference, in Barth's reckoning, was about 7 miles but only half of the capital's space was actually inhabited. Barth similarly noticed the city had a dendal  and baked-brick ruins outside the city on the road to Abu-Gher (Barth 518). Pâques, who also visited the site, found evidence for the vestiges of 20 houses, a cemetery and a mosque in the palace complex. Unfortunately, many bricks from the site were later used for the road between N'Djamena to New Massenya, meaning we have less evidence available (Pâques 17-18). Lebeuf has also written about the fired brick mosque in the palace. Evidence for additional fired brick buildings 1 km west-northwest of the palace were also noted (Lebeuf 219). The palace's grand entry gate and porio were made of fired bricks. The latter structure was only accessible to the mbang and certain adims (Lebeuf 227). Unsurprisingly, the palace cemetery had a short brick wall around it (Lebeuf 229). As for the length of the palace wall, Lebeuf measured it as 245 m and 240 m on its north and south sides and 215 and 182 meters on its other sides (Lebeuf 217). In terms of the palace complex's buildings, the ngarman, or "minister of the royal household," was in charge of its oversight. But the galadima was also responsible for the construction and oversight of some palace buildings (Reyna 111, 115). 

Some of the fired bricks from Massenya (Seignobos).

Overall, Massenya's brick structures were either for elites or the walls of mosques. This is similar to the patterns observed in Kanem, Wadai and Borno for this construction material. The palace complex of Massenya not only employed brick for its outer wall, but used for other structures. Outside of Massenya, other elite structures or mosques used brick, too. While Seignobos believes the earlier use of fired bricks at sites along the Chari predated Bagirmi's territorial expansion, other sources contradict this. For instance, Boussou, where bricks were used for the circular homes of the local mbang, was supposedly a site where their use predated Bagirmi's conquest (Seignobos 265). Contra Seignobos, however, Pâques argued that the use of fired bricks at Bousso was associated with the Bagirmi king, Haji, who facilitated the Islamization of the region (Pâques 200). Intriguingly, the local ruler of Bousso was also one of the few dignitaries allowed to sit on a carpet, perhaps another sign of the association of bricks for elite structures (Barth 562). Similarly, Bagirmi's conquest of chiefdoms or small polities along the Chari may have begun earlier and promotion of brick for elite structures or mosques could have started in the 17th or early 18th century. While Seignobos raised a legitimate point about the differences in brick shapes in Massenya and sites along the Chari, the common use of the word yalul for brick may be another indication of a common origin (Seignobos 273). Moreover, despite the diversity of brick types at Massenya, the most archaic ones were similar to those of Bousso (Seignobos 272). This could possibly indicate the deeper shared diffusion of the technology of bricks among societies south of Lake Chad. 

Overall, the use of bricks for elite structures is consistent with symbolizing power. Seignobos also emphasizes the role of submission in the use of bricks as peripheral towns or villages had to supply the bricks or labor through a corvee (Seignobos 275). This close association of elite status and brick constructions is certainly consistent with what is known about similar sites in Kanem, Borno, and Wara in Wadai. Massenya's palace complex, with its brick outer wall and use of brick for other elite structures, clearly exemplified this trend. It was also likely used to buttress the ritual seclusion and spatial separation of the king from the rest of the city. In the case of Massenya and Birni Gazargamo, where the palace complex occupied a central position, the political message must have been clear: the world revolved around the king. In the case of Bagirmi, where the local cosmological significance of the king as a Sun-like figure loomed large, one can imagine how brick served to highlight this elevated position (as well as that of the royal family and upper-echelon of officials or dignitaries). Therefore, Massenya's gur was yet another example of brick and elite ideology (and Islamization) in the Central Sudan. It reinforced, however, more deeply rooted cosmological or spiritual ideas about kingship.

Bibliography

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, Volume 2. New York: Harper and Brothers, 1857.

Comte de Castelnau, Francis. Renseignements sur l'Afrique centrale. Paris: Chez P. Bertrand, 1851. 

Devallée, J. "Le Baguirmi," Bulletin de la society de recherches congolaises 7, 1925.

El-Tounsy, Mohammed ibn-Omar. Voyage au Ouadây. Paris: B. Duprat, 1851. 

d'Escayrac de Lauture, Pierre-Henri Stanislas. Mémoire sur le Soudan, géographie naturelle et politique, histoire et ethnographie, moeurs et institutions de l'Empire des Fellatas, du Bornou, du Baguermi, du Waday, du Dar-Four, rédigé, d'après des renseignements entièrement nouveaux et accompagné d'une esquisse du Soudan oriental, par M. le Cte d'Escayrac de Lauture. Paris: Arthur Bertrand, 1855-56.

Lanier, H. "L'ancien royaume du Baguirmi," Bulletin du Comité de l'Afrique française 35, 1925.

Lavers, John. "An Introduction to the History of Bagirmi," Annals of Borno 1, 1983.

Lebeuf, AMD. "Boum Massénia, capitale de l'ancien royaume du Baguirmi," Journal des Africanistes 37-2, 1967.

Nachtigal, Gustav. Sahara and Sudan, Volume 3. Berkeley and Los Angeles: University of California Press, 1971.

Pâques, Viviana. Le roi pêcheur et le roi chasseur. Strasbourg: Travaux de l'Institut d'anthropologie de Strasbourg, 1977.

Reyna, Stephen. Wars Without End: The Political Economy Of A Precolonial African State. Hanover: University Press of New England, 1990.

Seetzen, Ulrich Jasper, "Nouvelles recherches sur l’intérieur de l’Afrique," Annales des voyages, de la géographie et de l’histoire ou Collection des voyages nouveaux les plus estimés, 19, 1812, p. 164-184.

Seignobos, Christian. "Les briques cuites du Chari," 2000 ans d’histoire africaine. Le sol, la parole et l’écrit. Mélanges en hommage à Raymond Mauny. Tome I. Paris : Société française d'histoire d'outre-mer, 1981. 542 p. (Bibliothèque d'histoire d'outre-mer. Études, 5-6-1).

Borno's Failed Mandara Campaign (1781): Ungwa-a-Mufaka

A map of Mandara in the early 18th century by Barkindo (The Sultanate of Mandara to 1902).

One particularly interesting textual source on the history of Borno in the 18th century was written in 19th century Mandara. A kirgam written in Arabic decades after a war in which Borno forces were defeated by Mandara, the text is both fascinating and problematic. It nonetheless offers one of the few descriptions of the physical appearance of Ali b. al-Hajj Dunama (r. 1747-1792). It's also one of the few written sources on Borno's disastrous defeat outside Doulo that is dated to the period before the colonial era. Indeed, the defeat in many ways encapsulates Borno's declining fortunes in what we like to call the Late Sayfawa Period (1715-1808, or 1715-1846). This period saw Borno's loss of the lucrative salt trade from Bilma, growing independence of Bedde communities, Gobir's refusal to render tribute, a long drought, Wadai's incursions into Kanem from the east, and the truly calamitous fall of Birni Gazargamo to forces aligned with the jihad of Uthman dan Fodio. Consequently, Borno's defeat by the hands of Mandara, despite mustering what the sources suggest was an enormous army (including cavalry and firearms), is worthy of historical inquiry for understanding the Sayfawa's gradual loss of paramount status in the Central Sudan. In order to address this conflict, the Ungwa-a-Mufaka is an essential source.

Written during the reign of May Iliayassa (1845-1858) of Mandara, the  Ungwa-a-Mufaka was translated into French in Eldridge Mohammadou's Le royaume de Wandala ou Mandara au XIXè siècle. Though written many years after the conflict, dated to 1196 AH, the kirgam cites a number of eye-witness testimonies. For instance, the future prince, Boukar Adjama, was one witness  (60). Another, Oumata Nadji, was cited for a description of the physical appearance of Ali b. al-Hajj Dunama on page 62: 

C'était un homme d'une taille impressionante. Son teint était clair et lui donnait l'apparence d'un Arabe Chouwa. Il avait une grosse tête, des yeux aux globes très blancs mais qui devinrent toutes rouges lorsqu'il monta en selle. S'élancent dans les rangs de son armée, il se mit à exhorter ses hommes d'une voix si puissante que les veines de son cou enflèrent et qu'on put les distinguer nettement à distance. 

Besides the eyewitness reports, the kirgam also quotes letters exchanged between Ali b. al-Hajj Dunama and the king of Mandara, May Bladi. Since it is far more likely that the conflict between the two states happened during the reign of Bladi's successor, Bukar-a-Jama, the kirgam must be interpreted carefully. Nonetheless, the access of the chronicle's author to Mandara elites who were present during the war and possibly correspondence between the Sayfawa mai and Mandara's king, the source presents a plausible account for Mandara's victory. 

A map of Mandara at its territorial zenith (Barkindo).

So, what does the chronicle actually tell us of the conflict? According to this source, the war began after a Shuwa Arab went to Gazargamo and reported to the Sayfawa mai the resources of Mandara. This Arab had lived in Mandara before moving to Borno, thereby making him an informed observer of Mandara's political, economic and geographical conditions. The ruler of Borno, hearing abou this, developed an interest in seizing Mandara, eliminating Mandara's king and dignitaries, and reducing its people to slavery (44). Subsequently, Borno sent its emissary, Katouri b. Boukar, to Mandara with a letter from the mai. This first correspondence compared the mai to a lion or cat while belittling Mandara's king as a chicken or hare. The implied threat clearly stated his hostile intentions: "Ceci je te le dis pour que tu sois informé et que tu puisse méditer sur la conduite, et les conseils les plus opportuns à prodiguer à  ton peuple" (45). Thus, the conflict was presented as one of Borno initiating conflict and declaring hostile intent through letters. The use of the written word to exchange messages between kings in the Central Sudan was also evident in the letters sent by the Bulala and Sayfawa during the conflict in Idris b. Ali's reign (1564-1596). To what extent the conflict was fully instigated by Borno is unclear, but it is difficult to imagine Borno launching a major expedition against a kingdom merely out of a desire for booty and conquest, particularly due to the problems on many other fronts that must have occupied the court's attention. In that light, it is better to understand the conflict as caused by Mandara's assertion of dominance over some of Borno's southern tributaries during the 1770s and early 1780s, as suggested by Barkindo. 

In order to prevent the war, Mandara's king allegedly sent many letters, gifts, money and slaves. The first response included 3000 thalers, 1000 slaves, 100 horses and a promise of tribute (48). Borno's Ali b. al-Hajj Dunama, however, refused to accept this. Indeed, he went as far as imprisoning and later ordering the punishment of a member of his court who urged peace (49). Unsurprisingly, Borno's second letter pledged to invade Mandara. A numerous army consisting of cavalry, archers, spearmen, shield carriers, and men armed with firearms began the march toward Mandara (50). Despite this invasion, Mandara's king continued to attempt a peaceful resolution by sending gifts and money with his letters. Indeed, the kirgam gives an exceedingly high number of 32 letters sent by Mandara to stop the Borno invasion (41). Nonetheless, these repeated peaceful overtures fail despite the Mandara king's message, "Nous mettons entièrement à ta disposition et attendons tes ordres pour les exécuter. Tous tes désirs seront les nôtres" (52). The dream of a Mandara marabout is even mentioned, probably to symbolize the imminent attack from Borno against the capital.

Since the chronicle emphasizes the numerous attempts by Mandara to avoid military conflict, it also highlights the immoral or unjust nature of Borno. For example, Ali b. al-Hajj Dunama's imprisonment of a member of his court for speaking against the war serves to highlight his injustice. Another episode, even more alarming, emphasizes the immoral actions of a Sayfawa prince. Named Yarnam, the prince stayed at Karangou with a man named Ali Ayssami. In addition to being fed by his host, Yarnam demanded payment and later raped his host's wife and daughter. When Ayssami sought redress from the mai, he did nothing (57). This episode is yet another instance in which the cause of the Sayfawa was unjust and immoral. That Borno would go to war against another Muslim kingdom who sought peace was also an argument in favor of Borno's immoral war.

Returning to the chronicle suggests Borno's forces took about 18 days to reach Doulo, the Mandara capital. Along the way, thousands of soldiers joined the campaign and various villages of Mandara were sacked or pillaged (56). Once outside Doulo, a kaigama, named Ali Madjounoun, leading Borno's forces, was instructed by Ali b. al-Hajj Dunama to attack, seize the palace and prevent the enemy from fleeing into the mountains (55-56). This suggests that the Bornoan army was very familiar with the mountainous terrain favoring Mandara. While this was happening, Mandara, which apparently had scouts or spies informing their court of the Borno army's movements, had carefully prepared for their defense. The king planned specific areas for his troops to strike (60). Their careful planning paid off despite being outnumbered by the invading forces. Indeed, they were successful enough since the kaigama was taken captive. When he refused to serve Mandara, he was executed (61). As the battle continued, ultimately lasting from noon until sunset, the mai led the Bornoan forces, exhorting them to fight while mounted on his horse. But, the Bornoan forces began to retreat, losing all semblance of order as Mandara soldiers pursued them (63-64). During this chaotic retreat, many Borno troops were seized and their weapons, valuables, horses, and persons were taken. 

Meanwhile, Ali b. al-Hajj Dunama continued to retreat on horseback as Mandara soldiers chased after him. They apparently let him escape as the night progressed, but they managed to seize the person carrying his scepter (64). The chronicle suggests that the Mandara troops, being simple soldiers, were hesitant to directly attack the king of Borno due to the "sacred person" of a king, even an enemy (64). That characterization, however, is contradicted by Borno traditions, which mention Ali b. al-Hajj receiving a wound in this war (Landeroin 399). Perhaps this attempt by the Mandara chronicle was meant to minimize the idea of soldiers attacking kings to maintain status distinctions and power relations? Regardless, the Mandara campaign was an utter disaster for Borno. Many troops were lost, horses, weapons, and soldiers were seized, and Mandara likely continued to intrude upon Borno's southern tributaries. While the kirgam states only 180 Mandara elites were killed and 35 disappeared, the casualties of the rank and file are not known (but probably very large) to the chronicler (65).

Ultimately, the Ungwa-a-Mufaka is a precious source for contextualizing one of the military defeats in the Late Sayfawa Period. Although obviously biased and perhaps too eager to portray the ruler of Mandara as entirely blameless, the kirgam provides some much-needed context on the affair. It provides a physical description of the mai and some insights into the nature of Late Sayfawa Period military and political thinking of this era. Without the Bornoan perspective as can be reconstructed in the works of al-Tahir al-Fellati, one must use the Mandara kirgam and oral traditions. The benefit of the chronicle is in quoting the testimony of those actually present when the war took place in 1781. Hopefully the other textual sources on this episode will be published or studied for further insights into its causes. Lastly, one should consider the 19th century date for the kirgam and the possible impact of the jihad on delegitimizing the Sayfawa in some quarters. This could be a relevant factor for understanding the kirgam's portrayal of the mai as an oppressive king.

Bibliography

Barkindo, Bawura M. The Sultanate of Mandara to 1902: History of the Evolution, Development, and Collapse of a Central Sudanese Kingdom. Stuttgart : F. Steiner, 1989.

Mohammadou, Eldrige. Le Royaume du Wandala ou Mandara au XIXe siècle. Tokyo: Institute for the study of languages and cultures of Asia and Africa, 1982.

Tilho, Jean (editor). Documents scientifiques de la Mission Tilho. Paris: Imprimerie Nationale, 1910.

Faras

We came across a fascinating video showing some of the beautiful works of art recovered from Faras (in Nubia) on Youtube. In the Women of 1000 AD project, the list of sources for Queen Martha included a link to a related video.

The Early History of Gao and the Songhay: Trade, Power, and Islam along the Niger (AI)

What follows below is an attempt to use AI to write a short essay on a topic based on our notes. The notes in question were written based on several academic publications we read a few years ago. Sadly, the free version of ChatGPT was unable to reproduce proper citations before we used up the free access to the best version. Nonetheless, we thought it would be worthwhile to show what it is capable of doing with information we fed to it. After some time, we will ask it to revise this again using whatever additional sources it can access. A link to our notes that are the basis of this can be found here. 

The early history of the Songhay and the rise of Gao (known to medieval Arabic writers as Kawkaw) illustrate the development of one of West Africa’s earliest powerful political centers. Located along the Niger River at the point where Saharan caravan routes met fertile riverine settlements, Gao emerged as a dominant regional hub before and during the early second millennium CE. Arabic geographers, although writing from afar and often relying on merchant testimony, consistently presented Gao as a sovereign kingdom that exercised influence over multiple subject peoples. Archaeological evidence, particularly from Gao-Saney, reinforces this image by revealing craft specialization, foreign imports, and signs of urban stratification. Yet the limited availability of contemporary written sources means historians must critically combine archaeology, oral tradition, and later West African chronicles to reconstruct this period. Through this synthesis, a picture emerges of a culturally dynamic society grounded in indigenous traditions but strengthened by the adoption of Islam and long-distance trade.

Well before the rise of Gao as a trans-Saharan commercial center, the Niger Bend supported communities that relied on the river’s abundant resources. The Sorko, known as expert fishers and boatbuilders, and the Gow, skilled hunters of hippopotamus and other dangerous river animals, formed some of the earliest identifiable groups associated with Songhay. Their craft specialization facilitated trade along the river and created patterns of settlement that anchored early political formation. Matrilineal descent was a defining feature of these river societies. Authority passed not solely from father to son but through the mother’s lineage, which ensured that elite women played essential roles in validating leadership transitions. Religious life centered on spirits connected to water, the riverbed, and animal life. Kingship was sacred, and rulers derived legitimacy from their roles as intermediaries with supernatural forces. Such cultural foundations did not disappear with the arrival of Islam; instead, these Sudanic social structures shaped the ways Islam was received and practiced.

Songhay historical memory, preserved in later chronicles, places the earliest ruling seat at Kukiya, downriver from Gao. The Zuwa (or Za) dynasty, said to descend from a foreign ancestor or hero figure, represents the first political lineage acknowledged by tradition. Oral narratives speak of battles between these early kings and powerful river spirits, sometimes depicted as fish demons, symbolizing struggles over control of ritual authority and sacred space within the community. Archaeological discoveries in the Bentyia–Kukiya region, including early Muslim burials, suggest that the site was not isolated nor purely mythical. Instead, Kukiya appears to have been a substantial multiethnic settlement that was already interacting with Islamic traders by at least the 11th to 12th centuries. Over time, as trade networks intensified and Gao’s geographic advantages became more important, political power shifted northward, indicating a gradual process of state consolidation rather than abrupt replacement.

By the 8th–9th centuries CE, Gao emerged clearly in Arabic writings as a capital city and political force. The geographer al-Yaqubi noted that the king of Kawkaw ruled a number of subject peoples, including desert-dwelling Sanhaja nomads. This claim aligns with archaeological findings showing that Gao was integrated into long-distance economic systems. Gao’s prosperity stemmed strongly from trans-Saharan trade. Caravans from Tadmekka, Kawar, and the Fezzan arrived with salt — a critical commodity that held both economic and political value. In return, Songhay merchants and rulers exported copper, ivory, animal products, and enslaved captives. These exchanges linked Gao to the wider Islamic world stretching from North Africa to the Mediterranean and the Middle East. Excavations at Gao-Saney have revealed copper-working furnaces, imported pottery and glass beads, and architectural features suggesting elite compounds. Such evidence confirms that Gao was not merely a waystation but a thriving production and consumption center. The kingdom’s ability to control trade routes and levy taxes or tribute helped consolidate political authority and population growth.

Islam did not arrive in Gao through conquest but through commerce and diplomacy. By the 10th–11th centuries, rulers known by the title Qanda (or Kanda) had converted to Islam, according to al-Bakri. Their subjects, however, largely maintained traditional beliefs and cult practices. Gao thus existed at this time as a dual-religious polity, with two towns: one Islamic and one that upheld indigenous spirituality. This structure demonstrates a transitional phase when kings embraced Islam’s advantages for diplomacy and legitimacy while still depending on traditional bases of ritual power among their people. Notably, Islamization appears to have been elite-driven. Conversion supported alliances with North African Muslims and facilitated smooth trade relations. But kings continued to rely on matrilineal structures to ensure dynastic stability, which reflects the persistence of indigenous frameworks for rule.

A political transformation became visible in the late 11th and early 12th centuries. The names of rulers that appear in Arabic texts vanish, and new rulers known as the Zaghe/Za emerge prominently in the archaeological record. Royal tombstones found at Gao-Saney are inscribed with Arabic prayers and titles such as malik (king) and malika (queen). Some of these funerary stelae were carved from marble imported from Almeria in Islamic Spain, revealing direct commercial and cultural connections with the wider Muslim world. This new leadership reinforced Islam at the highest levels of authority while retaining Songhay’s matrilineal basis of succession. The prominent burial of queens indicates that elite women were not diminished by Islam but remained central to the legitimacy of rulers. These developments mark Gao’s first clearly documented period of Islamic monarchy, accompanied by a flourishing of intellectual and cultural engagement beyond the Sahara.

Gao’s influence extended beyond commerce. Medieval accounts record that its rulers engaged in conflicts with the Ghana Empire to the west and Kanem to the east. These wars were not merely territorial but economic, driven by competition for control of trade corridors, especially those linking goldfields, salt mines, and northern markets. Warfare and diplomacy were tools to secure and manage tribute from both riverine and desert communities. Gao’s growing military power suggests a centralized administrative apparatus capable of organizing cavalry and riverine forces. That capacity would later form the basis of Songhay’s imperial strength under Sunni Ali.

Islam in Gao did not override existing beliefs. Instead, the two systems merged into a culturally hybrid form of statecraft. Rulers valued Islamic law, Arabic literacy, and international alliances, while also maintaining sacred ceremonial roles rooted in Sudanic cosmology. Spirit possession, public ritual, and matrilineal inheritance endured. This dual legitimacy — Islamic and indigenous — made Gao’s political structure both flexible and resilient. Rather than abolishing entrenched social customs, Islam was used to reinforce monarchy and expand bureaucratic functions, especially in commerce, taxation, and diplomatic communication.

Reconstructing early Songhay history requires attention to the nature of the available sources. Arabic geographers provide some of the earliest written references to Gao, but their works contain gaps and uncertainties because they relied heavily on second-hand information. Archaeology offers concrete material evidence of urban development and Islamic influence but remains limited to sites that have been excavated, especially around Gao-Saney. Oral traditions preserve indigenous perspectives and cultural meanings but contain mythic elements and were shaped by later political contexts. The Tarikh al-Sudan and Tarikh al-Fattash, written centuries later, reflect Songhay’s imperial ideology rather than early history. Despite these limitations, the convergence of evidence strongly indicates that Gao was a powerful and Islamizing state by the 11th century, rooted in earlier indigenous cultural institutions. Continual comparison of sources allows historians to create a clearer, though still evolving, understanding of this crucial period.

By the beginning of the 12th century, Gao had emerged as a thriving urban center with solid economic, military, and religious foundations. Its rulers commanded long-distance trade networks, projected force against rival states, and served as guardians of both Islamic and indigenous authority. These achievements ensured that Gao remained influential even when incorporated into the Mali Empire in the 13th and 14th centuries. When the Sunni dynasty restored independence in the 15th century and expanded to create the Songhay Empire, they built upon centuries of state formation and a tradition of political innovation rooted in the early history of Gao. The foundations of one of West Africa’s greatest empires began not with sudden transformation, but through gradual evolution driven by commerce, culture, and hybrid spirituality.

Gambaru

Migeod 1924

One of the most illustrious sites associated with the Sayfawa dynasty of Borno is Gambaru. Located only a few kilometers from the royal capital, Birni Gazargamo, Gambaru was said to have been a favorite retreat of the Sayfawa rulers. The site is also remembered traditionally as the construction of a magira named Aisa Kili N'girmaramma. According to H.R. Palmer, this queen-mother ordered the construction of Gambaru for Idris Alooma (r. 1564-1596). Others, referring to Idris Alooma's mother as Amsa, attribute the site to Amsa. Despite the contradictory traditions on the mother of this famous mai, Gambaru was apparently in use until c. 1808 or 1809, when the attacking forces of the jihad pillaged Gazargamo and other sites in the region. In an attempt to reconstruct the history of Gambaru, we shall share our own thoughts on the site based on studies by historians, archaeologists, and travelers. Understanding Gambaru's origins and its function as a palace site will elucidate how the Sayfawa maiwa utilized brick for elite architecture. It may also reveal a Bornoan example of how a small royal center could coexist with a much larger birni 5 kilometers away that was the real economic and political core of the state. 

Migeod (1924)

Naturally, one must begin with the origins. Since Gambaru was referenced multiple times by Ahmad b. Furtu, including an allusion to a mosque there, the site was already established by the reign of Idris Alooma (Lange 131). Traditions collected by Landeroin extend the origins of Gambaru as a Sayfawa center further back in time, to the 15th century. Per tradition, the area of Gambaru was where a Sao hunter, Gouma Kandira (also called Dala Gomami) lived. When the Sayfawa arrived in the region, he granted them permission to settle nearby. Over time, the Sayfawa (or their followers from Kanem) used their abundant merchandise to acquire additional lands from the local Sao. Over time, Gouma Kandira was tricked into buying too many goods on credit that he could not repay. This led to Gouma leaving the region and the Sayfawa making Gambaru their capital. However, they found the region was too densely wooded, too close to the river (causing frequent flooding) and the site was not favorable for horses (Landeroin 356). Interesting, this Sao chief, Gouma Kandira, may be identified with the Sao hunter who settled in Machena, Bolo Kandira (also known as Guma Kerbina). Although C.J. Lethem recorded a version of events which attributed Gouma Kerbina's departure from Borno to his father, the traditions of Gouma Kandira/Bolo Kandira (or Dala Gomami) were known as far west as Machina. By their own traditions, the ruling dynasty of Machina were said to be descendants of this man. 

Graham Connah (1981)

A different version of this tradition was reported by Palmer. In Sudanese Memoirs, Palmer wrote of the "So" Dala N'Gumami, clearly the same Sao leader Dala Gomami. However, instead of Gambaru, Palmer wrote that Dala lived at the town of Gaji Dibun. The followers of the Sayfawa were already using land nearby for pastureland, amazing the gigantic Sao leader due to their small size and horses (Palmer 66). This Sao hunter subsequently led the Sayfawa to the area of Gazargamo. The familiar tale of the trick of the henna was then used to defeat the Sao and kill nearly all, except for Dala N'Gumami. This Sao leader was said to have been a friend of the Sayfawa (67-68). Despite the absence of Gambaru in Palmer's version of this event, it should be remembered that Gambaru was located quite near to the site of Gazargamo. The lack of the toponym in Palmer's version may have been due to different informants attributing the Sao hunter's town to the general vicinity of Gazargamo and Gambaru, along the Komadugu Yobe. Undoubtedly, further evidence is necessary, but a Sau-Gafata connection for these communities may be assumed since they were still living near Gazargamo in places like Damasak at the time of Idris Alooma's reign.

Thurstan Shaw (1978)

Since oral traditions associate Gambaru with the Sao leader who led or welcomed the Sayfawa to the area of Gazargamo, the area was inhabited by Sao peoples long before Gambaru became a royal retreat for the rulers of Borno. Traditions are still imprecise for the exact process in which Gambaru became Sayfawa royal center. For example, Lange heard traditions in the region that reported the site was constructed by Aisa for Ali Gaji or Dunama Dibalemi. Furthermore, one of these kings did not like Gambaru because it lacked adequate space for horse racing (Lange 133). This is clearly an instance of modern tradition recalling the names of two of the most prominent Sayfawa monarchs and anachronistically turning Aisa into the mother or magira of them. Migeod, writing in the 1920s, also recorded that the mai rejected Gambaru as a capital, thereby leading to it becoming the magira's town (226). Palmer, in Bornu Sahara and Sudan, wrote of Aisa Kili N'Girmaramma having great difficulty in protecting the life of Idris Alooma from his cousin, Dunama Fannami and Abdallah. As regent for her son, after Abdallah's death, she was said to have built the palace and mosque at Gambaru to so Idris would not be corrupted by the manners of Gazargamo (Palmer 232). 

Seidensticker's Plan

Muhammad Nur Alkali similarly attributed the construction of Gambaru to the queen Amsa, mother of Idris Alooma, based on a praise song for her (Alkali 145). Likewise, Bivar and Shinnie, in "Old Kanuri Capitals," also reported the tradition of Amsa (or Aisa) as the builder of Gambaru (Bivar & Shinnie 3). Even a problematic praise song in J.R. Patterson's Kanuri Songs, for a magira named Aisa, referred to her as owner of Gambaru. Overall, most traditions concur on the identity (Amsa or Aisa) of the magira who built Gazargamo, although some attribute the son for whom it was built to Ali Gaji. The sense of the mai rejecting Gambaru as a capital may be an interpolation of the other traditions in which the Sayfawa rejected Gambaru (or its environs) earlier in the 15th century. Lamentably, the method in which the magira sponsored the brick buildings or walls of Gambaru is ambiguous. Traditions of craftsmen imported from Tripoli to complete the project may be unreliable. After all, the Sayfawa had sponsored brick buildings since the 12th century in Kanem. It is possible that workers were local artisans using skills that the Sayfawa had long sponsored for elite structures. Possible evidence of this can be found in the existence of a mai jalabvube in Gazargamo, the head of brickmakers (Seidensticker 245). In terms of the construction of the brick walls, Migeod did note that the bricks showed evidence that not all the bricklayers were equally skilled (Migeod 227). Nonetheless, a magira like Aisa Kili N'girmaramma possessed ample resources and had the structures built. Gambaru is the only known site with fired-brick ruins believed to have been built by a woman.

Palmer (1936)

Due to the consensus on the builder of Gambaru, what can we infer on the site's history from the 16th century to the early 19th century? Lange noted the reference to horses in oral traditions on the site, which brought to mind J.R. Patterson's district report on Borsari from the colonial period. If the mulima, a slave official in charge of the royal horse stables, was based at Gambaru, oral tradition may be alluding to this official. According to Patterson, the mulima collected the taxes from the villages or towns of Sage, Gilbossu, and Jaba. The mulima was based at Gambaru in the past. If Nachtigal was correct to identify the post of mulima as one held by slaves, this official was also at Gambaru and received the right to collect taxes or tribute from villages in the area. Of course, in the days before the abandonment of Gazargamo, the metropolitan district was densely populated and full of villages and towns. This means that in the past, the mulima may have held tax rights over a large population. Provisioning members of the royal family staying at Gambaru may have also entailed some fiefs or administrative rights over villages in the region. Sadly, it is unclear if the fiefs held by the magira were also included among those tied to Gambaru.

Palmer, Sudanese Memoirs

Indeed, a possible hint of Gambaru's past connections to the larger region can be seen in Denham's brief description of the site. Denham, presumably drawing on oral tradition, was told the nearby river was once very actively used by boats moving from Gambaru to Kabshary. The surrounding area was also once extensively cultivated (Denham, Clapperton, Oudney 156). This account, though brief, was based on traditions only a little over a decade after the site was abandoned, so it has a degree of reliability. The active river traffic, well-cultivated surroundings, and proximity to Gazargamo meant an economically dynamic regional economy. In fact, farms along the Komadugu Yobe were sometimes irrigated, according to Patterson. This facilitated the growth of cotton and food crops. Lastly, given that horses were of major importance for the military capacity of the state, one can be also be sure that the Sayfawa rulers who occasionally stayed at Gambaru kept a close eye on the mulima. In truth, the name of Aisa Kili N'girmaramma may also be a reference to horses or horse racing (Lange 133). This is highly suggestive of a link between horses, the magira, and Gambaru. Being in an economically prosperous regional economy that was densely settled meant Gambaru could draw on these ample resources.

Sassoon's photo of Gambaru ruins 

Unfortunately, little remains of the site of Gambaru besides the brick walls. Luckily, Musa Hambolu, who partly excavated Gambaru in the 1990s, focused on a few areas that shed additional light. One of them, a Yobe-type archaeological mound was a 2x2 excavation north of the palace. Another spot excavated was 4x4 in the southeast quadrant of one of the round hut foundations within the palace (Hambolu 221). In terms of the actual palace, Hambolu described it as oblong-shaped and featuring 10 compartments or sections. Overall, it measured 24 meters long from the southwest to southeast, 12 meters from southeast to northeast, 27 meters from northeast to northwest and 10.5 meters from its northwest to southwest axis, a total of 243 square meters. In addition, 5 possible hut foundations were detected, 4 of which were located within one of the 10 compartments (Hambolu 223). While less impressive when compared to the palace complex at Gazargamo (100,000 square meters), Gambaru's buildings with circular brick foundations provide possible clues to the use of space between the walled areas, or compartments. Hambolu's excavation of a hut foundation in the southeast quadrant revealed a structure with a diameter 4.4 and 5 meters for its inner and outer dimensions. The wall was 60 centimeters thick but Hambolu did not excavate the entire structure (Hambolu 224). A possible purpose of the site where 4 of these structures were found could have been for the use of the harem or women of the mai. Perhaps one of the circular structures was similar to what can be seen at Machina today, apparently used for giving audiences in front of it. 

Another picture by Sassoon

Besides Hambolu, other descriptions of the site are based on the ruins of the walls still standing. The earliest, based on a description by a Borno native to a European who never saw Gambaru, dated to the 1810s. This brief account merely alluded to the remains of "castles" and houses erected by "Christians" in the past. The association of the site with Christians is certainly perplexing, but Gronenborn has suggested this may be due to a past presence of European slaves or captives in the region. In this light, it is intriguing to note that the original article, published in The Quarterly Review, referred to copper coins used by the Christians being dug up at Gambaro. The use of the word "Christian" in Borno to refer to "pagans" or kirdi should not be forgotten. The association with "Christians" may very well refer to a non-Muslim group in the area or perhaps forcibly resettled nearby. Alternatively, the presence of European slaves and renegades in precolonial Borno, allegedly highly esteemed by the Sayfawa, could also have been present at Gambaru in the 16th and 17th centuries. 

Sassoon picture of standing wall

After The Quarterly Review, travelers who visited the site described Gambaru. Denham was the next to write of his visit to Gambaru's ruins. Noting the mosque walls enclosing a space of 20 square yards, he also observed the house of the sultan with gates opening toward the river. The private mosque attached to the sultan's house was also still observable, as well as other buildings of brick. Barth, who came next several years later, was only able to distinguish 1 wall and a building he believed was part of a mosque (Lange 131). Unmentioned by Lange are the allusions to artificial basins or lakes near Gambaru. Denham, for instance, wrote that Gambaru was located northwest of Lake Muggaby (Denham, Clapperton, Oudney 155). Barth observed what he thought were artificial basins or lakes in the area near Gambaru, too (Barth 345). Consequently, the capital district may have used canals or irrigation ditches that fed artificial lakes or basins of water for agricultural or other purposes. Gambaru, located near a major river, may have benefited from this as water was rerouted to the agricultural settlements whose farmers paid taxes or tribute.

Photo of Gambaru in Bivar & Shinnie's "Old Kanuri Capitals"

Moving into the 20th century, more detailed descriptions of the site emerged. The valuable insights and observations by 20th century visitors and archaeologists are weakened by the plundering of the site for bricks. Nevertheless, some walls still stood to an impressive height in the 1920s. When Migeod visited, he described the palace as 250 yards by 150 yards, with walls of burnt brick. The brick walls were supposedly used to divide the space into multiple yards or sections used for walking, horses, or gardens (Migeod 226). The Gambaru walls impressively still stood up to 8 feet in some sections. However, he did not see evidence of a great palatial house structure, writing, "A great palace in the form of a house did not exist." But, he did view the foundation of a round structure or hut of a large diameter in the northwestern corner of the site (Migeod 227). Even by the 1960s, Bivar and Shinnie viewed walls that still stood up to 3 meters (Bivar & Shinnie 4). Some sections may have been partly restored during the colonial era, but probably not in any way that substantially altered the layout. Even by the 1980s, when Wilhelm Seidensticker produced the only plan of Gambaru, some of the brick walls survived. Seidensticker's plan and study indicates the site's palace was really 10 interior courtyard divided by brick walls. These walls, 1 meter wide with the width of 7-8 stretchers, were differently fired since they were much darker in color (Seidensticker 65). 

Another image of Gambaru's remaining walls in Bivar & Shinnie

As evidenced by the contradictory reports of Gambaru, much of the site's history and architecture remain enigmatic. Even the question of a palatial house or brick house is ambiguous. While Denham, who saw the site a little over a decade after its abandonment, believed there was a large sultan's house and private mosque, later visitors were less confident. Migeod, in the 20th century, believed the brick walls were mainly used to separate the 10 or so compartments or sections, which each had their particular uses or purposes. Gambaru's possible "Sao" origins and link to the Dala Gumami or Gouma Kandira of Borno oral tradition is additionally ambiguous. Thankfully, traditions concur on the magira who ordered the construction of Gambaru. Likely the mother of Idris Alooma, this makes Gambaru one of the few (if not the only) Sayfawa fired-brick site which can be linked to a woman ruler. The site's location and proximity to Birni Gazargamo, the Sayfawa capital established by Ali Gaji, made it a part of a populous and prosperous region that engaged in irrigated agriculture, trade on the Komadugu and overland, and possibly the use of canals or ditches to create artificial lakes or ponds. Gambaru's connection with the mulima and occasional use by the Sayfawa rulers likely made it a popular site for the mai to escape the crowds at Gazargamo. Access to the mulima may have also facilitated the organizing of the horses for military expeditions. In short, the site was well-positioned and functioned as another manifestation of Sayfawa elite architecture in the history of Kanem-Borno. As a smaller royal site near Gazargamo, one cannot help but wonder which, if any, fired-brick ruins in Kanem served a similar purpose. If the antiquity of the birni can be traced to medieval Kanem, then it is difficult to imagine that Tié was the royal capital of Njimi. Was it not, perhaps, akin to Gambaru as a royal center visited by the rulers but not the large town or city sometimes implied in medieval Arabic sources? 

Bibliography

Alkali, Mohammad Nur. 2013. Kanem-Borno under the Sayfawa: A Study of the Origin, Growth, and Collapse of a Dynasty (891-1846) . Nigeria: University of Maiduguri.

Barth, Heinrich. Travels and discoveries in North and Central Africa: Including accounts of Tripoli, the Sahara, the remarkable kingdom of Bornu, and the countries around Lake Chad. London: Ward, Lock, and Co., 1890.

Bivar, A. D. H., and P. L. Shinnie. “Old Kanuri Capitals.” The Journal of African History 3, no. 1 (1962): 1–10. http://www.jstor.org/stable/179796.

Connah, Graham. Three Thousand Years in Africa: Man and His Environment in the Lake Chad Region of Nigeria. New York: Cambridge University Press, 1981. 

Denham, Dixon, Hugh Clapperton, Walter Oudney, and Abraham V. Salamé. Narrative of Travels and Discoveries in Northern and Central Africa: In the Years 1822, 1823, and 1824. 2d ed. London: John Murray, 1826.

Gronenborg, Detlef. "Kanem-Borno: a Brief Summary of the History and Archaeology of an Empire in the Central Bilad-el-Sudan" in West Africa during the Atlantic Slave Trade, edited by Christopher R. Decorse. London: Bloomsbury Academic, 2018.

Hambolu, Musa. "Recent Excavations Along the Yobe Valley." Berichte des Sonderforschungsbereichs 268 (1996): 215-229. 

Lange, Dierk.  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.

Migeod, Frederick William Hugh. Through Nigeria to Lake Chad. London: Heath Cranton Limited, 1924.

Nachtigal, Gustav, Allan G. B. Fisher, and Humphrey J (trans.). Fisher. Sahara and Sudan, Vol. 2. Berkeley and ; Los Angeles: University of California Press, 1971.

Palmer, H.R. The Bornu Sahara and Sudan. London: 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.

Patterson, J.R. Kanuri Praise Songs. Lagos: Government Printer, 1926.

Seidensticker, Wilhelm. "A Note on the Site of Gambaru, Borno State," Zaria Archaeology Papers 5 (1983): 65-67.

__________. "Borno and the East: Notes and Hypotheses on the Technology of Burnt Bricks" in  Lionel Bender and Thilo C. Schadeberg (editors), Nilo-Saharan Proceedings: Proceedings of the First Nilo-Saharan Linguistics Conference, Leiden, the Netherlands, September 8-10 1980.

Shaw, Thurstan. Nigeria: It's Archaeology and Early History. London: Thames and Hudson, 1978.

Tilho, Jean (editor). Documents Scientifiques De La Mission Tilho. Paris: Imprimerie Nationale, 1910.