- 1 As an exception, cf. M. Perinbam (1980: 457), who however a bit vaguely calls Wangara simply a Jula (...)
1More than a decade ago an article of mine on the Mande diaspora on the Malagueta coast and Sierra Leone appeared in this journal (Massing 1985). Inspired since then by the Mande colonies further to the east, in Burkina Faso, Ghana and the Ivory Coast, I have investigated some of the older elements, neglected then, which have played an important role in the history and diaspora of the Western Sudan and were important in the exchange between the Guinea coast and the Sahel, and attempted to clarify their identity and context. While Dupuis (1966), Wilks (1982, 1995), Lovejoy (1978), Fuglestead (1978), Law (1995) and others have dealt with the Wangara in specific contexts, namely the trade in Hausaland in the 16th century or in the Mande-Akan borderland in the 19th century1, I will attempt to outline here the overall context and cover the entire historical time span from ancient to modern Ghana and the geographical range from Mauritania to Northern Nigeria.
2The Wangara, while revealing slightly different identities at different times and in different regions, and being tied at times to specific localities, while at other times covering vast geographic ranges, nevertheless show a common denominator: involvement in the long-distance trade between desert- side and forest fringe and in the propagation of Islamic faith.
3I attempt to document Wangara identity through a) careful analysis of primary historical sources, b) cross-checking of secondary sources with contemporary oral testimony, and c) my own investigations into oral history of Malian, Ivorian, and Ghanaian settlements and clans in order to derive the clues necessary for identifying their origins and activities.
- 2 Different authors have referred to Ghana by various names: al-Bakrî distinguishes Awkar with Awdagh (...)
- 3 The earliest sources on Ghana date from the late 9th and beginning 10th century, but only provide v (...)
4The main primary sources on ancient Ghana are al-Bakrî and al-Idrîsî. Earlier Arab authors, who report on Ghana since the ninth century, do rely on traveller’s accounts unknown to us without first-hand experience in sub- Saharan Africa3. By the 11th century, however, the accounts begin to contain verifiable details on itineraries, production and trade. Wangara is then for the first time defined as a territory or realm on the eastern border of “Ghana” (Cuoq 1975: 134; Monteil 1968: 74).
5Al-Bakrî travelling in 1068 from the Atlantic coast toward the interior along the borderline between Berber and black ethnic groups is somewhat contradictory: on one hand he insists that Ghana was the title of the king (known under the alternative titles Maghan or Baghana) of a country called Awkar with Awdaghast as its capital, and which recognized the suzerainty of “Ghana”. This seems to imply that the suzerain of Awkar, the BaGhana, was not direct ruler of Awdaghast but resided elsewhere. But then he says Awdaghast was not the capital, because it was 15 days from the town of Ghana (Monteil 1968: 75). So, on one hand “Ghana” was defined as title of a ruler, on the other hand as the name of a “town” or state in another location.
6Al-Bakrî mentions the Gangara as a group of Blacks in the neighborhood of the Senhaja town of Banklabîn, and further east, a group of “barbarous” (i.e. non Moslem) blacks Beni Nghmarat: in the neighborhood of Ghiâru and Ytregga, two towns near the Niger (i.e. “the Nile of the Blacks”):
“Ils [Lemtouna] ont eu pour chef Muhw ammad dit Târeshnâ. C’était un homme de mérite et de foi, qui accomplit le pèlerinage et mena la guerre sainte. Il mourut au lieu dit Gangâra, au pays des Noirs. C’est (aussi) une tribu de Noirs, à l’Ouest de la ville de Bânklâbîn. Celle-ci est habitée par les Musulmans Banû Wârith, de la tribu des Sw anhâja” (Monteil 1968: 59). “Le meilleur or local vient de la ville de Ghiyârû (ou Ghayâruwâ), située à dix-huit jours de marche de la ville du roi [du Ghana]. Entre les deux villes, s’étend une zone ininterrompue, peuplée de tribus de Noirs. Si l’on découvre, dans n’importe quelle mine du royaume, de l’or natif, le roi met la main dessus: il ne laisse à ses sujets que la poudre d’or. Sans cela, il y aurait trop d’or sur le marché, et il risquerait de se déprécier. Les pépites d’or pèsent d’une once à une livre. On dit que le roi possède un lingot gros comme un rocher. Il y a douze milles entre Ghiyârû et le Nil” (ibid.: 73). “À l’ouest de Ghiyârû sur le Nil, se trouve la ville de Yaresnâ. Ses habitants sont musulmans, mais le pays est païen...” (ibid.: 74). “Les Noirs barbares connus sous le nom de Banû Neghmarât importent de la poudre d’or dont ils commercent dans le pays...” (ibid.: 74).
- 4 The usual Arabic spelling of Wangara, u n q r a, does not correspond to the phonemes in al-Bakrî’s (...)
7I. Wilks has identified the Banû Neghmarât as Wangara because of their specific characteristics: blacks, Moslems, and traders in gold dust, but his identification is not totally convincing, due to spelling differences and noncorresponding phonemes4.
- 5 Cuoq (1975), probably a misreading of Arabic letters sad and noun, for double ghain. I have not had (...)
8Gangara, or Guangara, on the other hand, corresponds phonetically better to our group, whom al-Bakrî characterizes as black non-Muslims, but he characterizes the Banû Neghmarât in addition as importers and traders of gold dust from the south. In my opinion, Ghiyârû, where they were met, could be Youvarou, situated on the western arm of the Niger–the Issa Ber–at the outlet of Lake Debo. Al-Bakrî’s Yatreggâ or Tîreggâ seems identical with the Tirkw a of al-Idrîsî5. Al-Bakrî’s itinerary reflects well the relation between Berber and black groups and between animism and the progressive introduction of Islam into the western Sudanic States.
9Since the 9th century, the Berber were progressively converting to Islam. In the west, two Senhadja Berber groups, the Beni Djoddala by the Atlantic and the Beni Lemtouna, to the east of the former and four days north of Ghana, converted to Islam since the 10th century. The black populations of TeKrow and Sîlla in the Sénégal valley were converted by one Wara Djabe (Djabi?) who died in 1040, while Sîlla was still fighting with its polytheist neighbors of Qualanbû, one day away (Monteil 1968: 68).
- 6 The two rivers may be the Diaka and Bani further south, or alternatively, further north the Issa Be (...)
10Al-Idrîsî apparently never visited the region himself, but used a source unknown to us which describes Wangara as a continuous gold-producing island between two major rivers. This seems to correspond to the interior floodplain of the Niger, called the Interior Delta 6(Gallais 1967). However, his report only describes gold panning–still practiced along the Niger in the Kangaba area–but not mining thereby suggesting that the Arab traders had not discovered the African mining areas by the 12th century (Cuoq 1975: 134-135, § 210). Both al-Bakrî and al-Idrîsî concur that mined gold was strictly controlled by royal monopoly for the purpose of maintaining high gold prices by quantity restrictions, while the trade in gold dust, on the other hand, was free, and notably in the hands of private traders referred to as Wangara (Monteil 1968: 76).
- 7 “Tamisgidda... wahrscheinlich identisch mit dem Tirka oder vielmehr Tirekka Arabischer Geographen” (...)
- 8 Interestingly, Tireka seems to combine the two names of Dire and Tonka (Direka) –which lie however (...)
11Al-Idrîsî describes one of the Wangara towns which H. Barth identifies wrongly with Tamisgidda7: “Parmi les villes des Wankw â ra, il y a Tîrkw a. C’est une grande ville, très peuplée mais sans murailles et sans enceinte” (Cuoq 1975: 135). C. Monteil (1971: 30) in his monography on Djenné refers to Delafosse’s claim to have identified its location east of Timbuktu. I am rather of the opinion, on the basis of the precise description by al- Bakrî, that its location was further west, near Dire and Tonka, where the two arms of the Niger from the south join, on the borders of the Tedmekkat Tuareg territory8.
- 9 In their discussion on the Almoravid conquest they reject the date of 1076, mentioned in numerous s (...)
12Al-Bakrî and Ibn Khaldoun (1968-1969) report that Awdaghast was sacked in 1055 by the Almoravids, who imposed tribute and Islam on its population causing many to flee east and south and giving rise to Ghana’s domination by the Susu. The question of Ghana’s destruction by the Almoravids has recently been debated and rejected by Conrad and Fisher (1982, 1983)9.
- 10 Valentim Fernandes, factory manager at Arguim, obtained important information on the Malian gold tr (...)
13Ibn Battouta and Ibn Khaldoun are the primary sources for the Malian period; the first Portuguese source seems to date from 145510. But the Wangara are mentioned in Mali from the time of the victory of Sunjata over Sumanguru (c. 1230) until the 15th century (Niane 1982: 3). The Tarikh es-Sudan (Houdas 1981: 18-20) and Tarikh el-Fettach (Houdas & Delafosse 1913: 40-41, 74-78), written during the Songhay period, relate earlier material from the period of Mali’s dominance in the middle Niger.
14Visiting Mali from 1352 to 1354 at the time of Mansa Suleiman, Ibn Battuta identified southern black merchants in the desert-side town of Zaghari as Wanjarata:
“After a distance of ten days’ travel from Îwâlâtan, we arrived at the village of Zâgharî, which is a big place with black merchants living in it. They are called Wanjarâta, and there live with them a group of white men who follow the sect of the Ibâdî from amongst the Khârijites. They are called Saghanaghû. The Sunni mâliki among the white men in that country are called Tûrî. From this town anlî is brought to Îwâlâtan. Then we went on from Zâgharî and arrived at the great river, the Nile. On it is the town of Kârsakhû. The Nile descends from it to Kâbara, then to Zâgha. Kâbara and Zâgha have two sultans who give obedience to the king of Malli. And the people of Zâgha are old in Islam, they are religious and seekers after knowledge. Then the Nile comes down from Zâgha to Tunbuktû [Timbuktu], then to Kawkaw [Gao], the two places we shall mention below. Then it comes to the town of Mûlî, which is the land of the Lîmiyyu¯n and is the last county of Malli” (Hamdun & King 1975: 32).
- 11 Cf. supra. The root Zagha- represents the Fulani word Sare-, Sara- as in Saredina, and Sarafere, me (...)
15Thus Zaghari, ten days journey distant from Walata, was not directly situated on the river. On the other hand Kabara, on the river, is the name of the port of Timbuktu, but also an old name for Lake Debo. Since its situation is given as downstream of Karsakhu, but upriver from Zagha, which itself was situated upriver of Timbuktu, Kabara seems to designate the Debo, and both Zaghari and Zagha–frequently mentioned in the Tarikhs–must have been situated to the south-west and upstream of Timbuktu. Zagha was the point of departure of the Wangara into Hausaland in the late 14th or early 15th century11.
- 12 More research into the Touré clan, one of the leading clans, on the middle Niger would be most inte (...)
16Ibn Battouta’s remarks are of considerable interest for the relationship between different communities and their religions, between foreigners and indigenous people, and animists and Moslems: in Zaghari, the kings (maliks) were orthodox Sunni muslims, and of the Toure clan, while another part of the population belonged to the unorthodox sect of the Ibadites, and the Saganogo clan. This is to my knowledge the first reference to Malian clan names in Arab sources and establishes the Toure–who are often referred as Arma (Arama), descendants of Roum or pre-Arab North African soldiers and local women–well before the Moroccan invasion 12while the Sanogo from the “Ibadi sect of the Kharijites” were probably the imams. In contrast, the Wanjarata are clearly described as blacks merchants.
- 13 Cf. Ibn Khaldoun (1969b: 272; 278), information on the Kharedjite rite among the Zindjas of the Mza (...)
17The distinction between Sunni and Ibadi points to a major religious rift in the area, one which continued among the Songhay at least until Mohammed Askia’s time. While the kings followed the Sunni rite, the imams followed the Kharijia rite, a heterodox doctrine adopted by the Senhaja and Zenata Berber following the first wave of Islamic conversion. According to Ibn Khaldoun the dispute between Kharijite and Maliki rites originated over the mode of selection of the imam, with the older, “heterodox” Moslims, chosing the imams from local lineages, while the Maliki doctrine became prevalent with the growing influence of a new generation of Arab muallim trained in North African universities (at Kaïro, Kairouan or Tunes) who required that the imam should have a connection to the prophet, in particular be chosen from the prophet’s mother’s lineage, the Khoreichi13.
- 14 Information on the khoreich by the Sekou Koreichi, son Oumar Koreichi, old “chef de canton de Dia” (...)
- 15 See Houdas (1981: 9-12) about Sunni origines. Sonni Ali is called the Great Tyrant, the Kharejite. (...)
18Mansa Musa is said to have brought one or several members of the Koreichi lineage on his return from the Hajj in 1324 to hold the imamship at his court14. But, more than hundred years later, Sonni Ali still adhered to the Kharijite doctrine, and for this reason is said to have organized the persecution and killing of the Timbuktu ulema in 1469 who were Sunni Malikites15.
- 16 “Malinke and Wangara are the sons of an early Malian king, the one chose royal power and reigned wh (...)
19A Malian source, cited in the Tarikh es-Sudan, and apparently used by Niane, attempts to assimilate the Wangara to Malinke. The Tarikh makes a fundamental socio-professional distinction between Malinke and Wangara, claiming both to be of the same origin, the former being princes and warriors and the latter “traders who carry gold dust from country to country as the courtiers of princes” (Niane 1982: 3)16.
- 17 Cf. Baiao & Bensaude (1940: 52-54). See also Monod et al. (1951: 85-87).
20In this context the information in V. Fernandes on Djenne is of prime importance and his description of the Ungaros is characteristic of the Wangara trade of the time17.
- 18 Oua or Wa, in Songhay etymology means “milk”, and koré “very white”, meaning “white as milk”; on th (...)
21Three Tarikhs (es-Sudan, el-Fettach, and Tedzkiret en-Nisian) are the primary sources dealing with this period. Several variant spellings of Wangara are used in these Tarikhs: Wakoré, Wankori, Ouankori, and Wangarbe, all variants of the same term in the languages spoken in Mali18–Soninke, Songhay, Fulani.
22In the 15th century, economic power shifted from the West to the East: Walata and its market-town Birou apparently declined in favor of Timbuktu and Gao which supplied the eastern trade routes to the Tuat, Ghadames, Tunis and the Ottoman empire (following the fall of Byzanz in 1453). The Western routes to Sijilmassa and Fes still existed and were actually revived under the Saadian sultans, as is documented by Leo Africanus’ voyages. But Mali’s decline, which had set on around 1375, continued, provoked by the frequent Mossi raids on the prosperous northern trade centers–Walata, pillaged thrice, in 1410, 1446 and 1479, and Timbuktu–so that finally the governor and defensor of these provinces, Sonni Ali Ber, decided to usurp power in 1456.
23In 1479 Sonni Ali besieged Djenne, but died around 1491 in one of his raids into Mossi. In 1492 the Askia Mohammed ended the hegemony of the Sonni dynasty and took Timbuktu (Konare 1980).
- 19 Tarikh el-Fettach, chapter I. See also Barth’s claim that Sonni Ali’s dynasty which was to have mig (...)
24The author of the Tarikh el-Fettach claims that Askia Mohammed was descendant of the Sîlla, a Soninke clan from the Torodo (the Fouta Toro), who had relatives in Gao and among the Toro from Yemen19. Then he attempts an etymology of the various names, by linking the Askia to the Soninke and Ghana, and even further to Yemen and thus legitimizing his origin. He claims to have learnt from a certain Demir el-Yaqub that Ouakoré, Ouangara and Songhay came from the same ancestor, a certain Tarâs from Yemen, who had three sons of the names of Ouakoré, Ouangara and Songhay. The elder son Ouakoré was given the kingship under the title “Kaya maghan” (king of gold).
25Twenty-two Kayamaghans are supposed to have ruled the “empire of Kayamaghan” in the capital of Koumbi preceding the advent of the “Godsent” (presumably the Hejra), and 22 thereafter.
26The Tarikh’s author is uncertain whether these Maghans, were Ouakoré, Ouangara or Songhay, but leans toward a Senhaja origin, because he was certain that the first 22 were not black! In a later passage he claims that the Zendj, black slaves taken from the Ouagadou Farin in an expedition by the askia El Hadj, spoke the Ouakoré language.
- 20 There are 12 citations of Wakoré/Wangara in the Tarikh es-Sudan, and 15 in the Tarikh el-Fettach. H (...)
27Other members of Songhay royal lineages, like the mother of the askia Ismail, Mariam Dabo, and the Mossi Kois are also said to be of Ouankoré or Ouangara origin. At least one pretender to the Songhay throne against the Askia El Hadj is known, Mohammed al-Bankori20.
28H. Barth claims that the Sissilbe in the Niger bend are nothing but a part of the numerous group of Wakoré or Wangaraua belonging to the same tribe as the Sousou and Malinké, who had abandoned their orginal language and adopted Fulfulde and even Hausa, whereas their western associates, Zaberma, still retained their original language. He also identified Wakore with Assuanek or Soninke, the main population of Ghanata, who were later overpowered by the related tribe of Malinke (Barth 1849-1857 (IV): 145; (V): 511).
29Binger (1892: 382-383; 393) claims that the Wakoré formed a principal element in Ghana, namely a class of court functionaries and administrators, and that their family names and titles (Sylla, Doukoure, Niakhate, Nimaga, Konaté) can be found without change among the Soninke; that later this class gave rise to the Mande-Dioula and Maraka who migrated from the Wagadou during the decline of Mali. Malian historians tell me that “Maraka” means “administrator, bureaucrat” in Songhay. It seems then that the Wakoré/Wangara were Soninke clans specialized in trade, Islamic scholarship and law who migrated in the 14-15th centuries from the Awkar, now on Mauritanian territory, into the Mali provinces of Mema, Beledugu, Zaga, Bendugu, Massina, and further East and South, perhaps founding such towns as Odienne, Koro, Boron and Kong.
30C. Monteil (1971: 31; 33) in his monography of Djenné explains that the founders of Djenné were Nono, i.e. “Reds”, who were Sanhaja called Guirga:
“Les Nono étaient de ces Sanhadja dits Guirga [. . .] Les Soninké dits aussi Wakoré, Sarakolé, Marka sont également des Guirga. Le mot Azer ou Adjer n’est qu’une forme berbère désignant les Guirga de Dia qui étaient aussi des Nono. [. . .] Ces Guiriga de race ou d’adoption, voués au commerce, ce sont les Wangara des Mali, les Wakoré des Songhay, les Marka des Bambara, les Malankobé des Macina, ce sont les Nono dont il est ici question.”
31V. Fernandes on Jenné speaks about the race of red or brown people called “Ungaros” who came annually to exchange gold for salt (Baiao & Bensaude 1940: 53). It seems contradictory that the Wangara were described as black in the 12 and 14th century, and then in the 15th as white or red. Can it be assumed that over the centuries some “métissage” of the black Soninke with the Berbers and Moors of Ghana had taken place giving rise to a group of colored (red) individuals?
32The Tarikh es-Sudan does not limit itself to mentioning the Wakoré as subjects of Ghana, but situates them as Islamic scholars, lawyers and cadis (judges), and identifies specific Wakoré individuals, namely:
– Fodiya Mohammed Fodiki Sanou El Wankori, who left his country of Bitou as a result of the internal strife and installed himself in Djenné in 1492 (Houdas 1981: 30);
– el-Abbas Kibi, Oua’kri of origin, and cadi of Djenné (ibid.: 30);
– Mahmoud-ben-Abou-Bekr-Baghyo’o, the father of the lawyers Mohammed and Ahmed Bagayogo, cadi from 1552, and founder of a whole family of “law consultants” (ibid.: 33);
– Mohammed-Benba-Kenâti (ibid.: 34);
– Mohammed-ben-Mahmoud-ben-Abu-Bekr (1524-1593) moved from Djenné to Timbuktu and shared with Ahmed Baba the captivity (ibid.: 71).
- 21 A. Massing, fieldnotes, Timbuktu, February 27, 1997; Jenné, August 5, 1995; Faragouaran, August 17, (...)
33One of the descendants, Mahmoudou Baba Hasseye Baghayogho, formerly director of the center Ahmed Baba, and guardian of the Sidi Yahya mosque in Timbuktu, in the Songhay ward of Bajinde or Maigala, also called Ouangarakunda, told me that his family was the only remaining Ouangara branch in Mali today21. His ancestors operated the caravanserail of the Azlai, the salt caravan, and the forward trade to Djenné. They supplied the caravans with camels, water, food and lodging, received and bought salt in Taoudeni, and hired Sorko or Bozo boatsmen for the forward transport to Djenné where they exchanged it for gold presumably with those of their family members who took care of the southern part of the gold trade. They were also active in Islamic law and religious duties. It is not unusual that the descendants of rich merchant families get a higher education and become scientists and lawyers.
34M. B. Hasseye Baghayogho also gave further details of the episode, mentioned by Mahmoud el Kati, his ancestor, in Tarikh es-Sudan, which triggered the massacre of 1592 of the Timbuktu ulemas by the Moroccans and the captivity of Ahmed Baba where, in which two of his Wangara ancestors were involved apparently resisting the Moroccan soldiers (Houdas & Delafosse 1981: 304).
- 22 A certain number of families in Bamako and in San do this, probably from a pilgrinage which took on (...)
- 23 As the northern branch was involved in the salt caravan, the southern branch could have been involv (...)
35Interestingly, Baba Hasseye Baghayogho also claimed that Abou Bekr Bagayogo came from Kong and that the Bagayogo family has a branch in Seguela. In checking this out, I was told that the Bagayogo claim to have taken their name from Bagdad (Bagadaji)22, and settled first in Djenné and later in Timbuktu, but that their southern branch comes from the Wasoulou (the villages of Keleya, Faragouaran, Danou, Bolou), and that Famusa migrated from Faragouaran to established the Ivory Coast branch in Koro, Odienne, the Mahou and perhaps the Konya in Guinea23.
36In Djenné, the descendants of the Wangara-Bagayogho are known as Garba, live now in the Sankore quarter, also sub-section of Wangarakunda, and are the guardians of the famous well of Nana Wangara and frequently imams of the Mosque. Tarikh es-Sudan mentions Muhammad Saquwa (Saganogo) al-Wankari, first qadi of Jenne under Askia al-Hajj Muhammad, and Muhammad al-Buni (from Bouna) allegedly a pupil of al-Hajj Salim Suwari, and teacher of Umar Fofana, Mohd. Saquwa’s contemporary, and imam at Bighu (Houdas 1981: 30).
37In Kong, however, there seems to be no memory of a Bagayogo family, even though there is a tradition that all the dyamou of Kong are of Wangara origin (going back to Taleb Salim Suwaré of Yelimane, deceased in 1452) (Seriba 1975). However, the imams of Kong have for time immemorial been Sanogo.
- 24 See the passages on the Wangara and Yalsé in H. Barth (1849-1857 (V): 583).
38Paradoxically, the infiltration of Wangara traders into Mossi seems to be a result of the Mossi incursions into the Niger valley and Walata since the early 15th century which contributed more to the decline of Mali than other factors, and which provoked the Songhay usurpation in last consequence. Mali and Songhay were constantly on the defensive against Mossi raids into the Gourma and the Niger valley and Sonni Ali perished in one of his expeditions against them. This might explain the profound penetration of the Marka into Mossi society, where they are known as Yalsé24.
39It seems paradox that the Mossi who were hostile to Islam should have permitted Moslems to enter their country, but their need for trade goods, especially salt, could explain not only their raids, but also the peaceful infiltration by Soninke–generally called Marka on the east of the Niger river. I suggest that the Mossi in earlier times raided the northern markets for salt (and other North African goods), but later permitted traders from these areas to import the desired goods into their own country. In addition, the survival of the Songhay kingdom in the eastern Gourma following the Moroccan conquest of 1592–under Askia Nouhou, the Dendi-dirma–could explain the gradual penetration of the Wangara into these eastern regions, Gourma (with Boulsa, Bilanga), Dendi and Borgou.
40In the Mossi kingdoms, the Wangara established themselves at Mane, Zitenga, Kaya, Bulsa and Bilanga. Members of the Bagayogo family entered Ouagadougou around 1750, the time Naba Dulugu adopted Islam from his Mamprusi relatives in Gambaga:
“The Mamprusi chief is regarded as the ‘father’ of the Mossi chiefs. Sons, one tradition says, imitate the father; so when Moro-Naba Dulugu heard that the Mamprusi chief had an imam besides his other ministers, he wished to have an imam in his court as well. He therefore invited a Muslim from Zitenga, and made him the first imam. This Muslim, Musw tw afâ Baghayogho, was a member of a family of Timbuktu origin, who had first settled at Mane, and then at Zitenga, both in Mossi. [...] Islam had been introduced to Mamprusi during Na Atabia’s reign early in the eighteenth century, over half a century before Naba Dulugu’s reign. [...] Dulugu, it is said, was sent together with his brothers, by their father, Moro-Naba Sagha, to study Koran with Muslims. Two of Naba Dulugu’s brothers, Ngadi and Sigiri, also became Muslims” (Levtzion 1968: 166-167).
- 25 Cf. H. Barth (1849-1857 (V): 292; 573-574; 583) on Dori, Gourma and Togo. See also R. Cornevin (196 (...)
- 26 Cf. N. Levtzion (1968: 240) about Islam in Borgou and Kotokoli, and Kuba (1995: 260)
41In the 19th century Barth and Cornevin in the 20th document the importance of Wangara elements in such places as Dori, Boulsa, Bilanga and many others as far as Sansanné-Mango, Djougou and Nikki in Northern Togo and Northern Benin25. The Mande elements in the commercial sections of Djougou and Nikki are called Wangara. Both settlements lie on the trade route from Gonja to Bornou and Hausa, a route was opened from Borun to Gonja by sultan Burja (1438-1452) and Gonja merchants arrived under Yaqquba (1452-63) (Palmer 1908: 75-77). More research on Wangara clans in Borgou could perhaps contribute further to the clarification of their origins26.
- 27 See Palmer (1908: 70), Al-Hajj (1968: 10-12). On the other hand, there are two more passages which (...)
42The relevant sources, the Kano Chronicle and one used by Ibrahim b. Mhd. n.Idris b.Husai, dated to 1061 (1650/51), mention that the Wangarawa–between 40 and 160 people–emigrated under the leaderhip of Shaikh Abderrahman surnamed Za(gha)iti and came to Kano and introduced Islam, according to the first source in Yaji’s time (1349-1385), according to the second under Mohamad Rumfa (AH 867-904, 1463-99), after having left Mali in 835 AH (1433 AD)27.
- 28 “This is a paper which contains the origins of the Wangarawa who are related to the Shaykh, the jur (...)
“When he and his community wanted to leave Malli, the Sultan of Malli implored them in the name of God to stay. But Shaikh Zagaiti said we must go because our intention is to perform the pilgrimage to the sacred House of God in the year 835. He emigrated together with the descendants of the tribes that were connected with his great-grand-father” (Al-Hajj 1968: 10)28.
- 29 Very likely from Diakha (Diâ).
43The surname is derived from “Zagha” or “Zeghai” and in my view points to one of the above-mentioned towns Zagha (Zare- or Sare-) in the Macina or Lake region south of Timbuktu29. These Wangara left during a time of great insecurity due to Mossi incursions and moved to greater Songhay protection, adopted the Songhay language and perhaps intensified the commercial contacts between Songhay and Hausa.
- 30 Takedda, a land noted for its wealth in copper, most likely the area “Teguidda” west of Agades whic (...)
44The Kano Chronicle associates the Wangarawa with Melle and the introduction of Islam; the Wangara chronicle and H. Barth document Wangara communities in Kano, Katsina and in the Borgou. Hornemann and de Lisle, quoted in Bowdich (1819), found Ouangara on the trade road from Yaoura to Bornou. Fisher (1977: 296) believed that the Wangara split up in two groups in Gobir, one going to Kano, the other going to the Aïr, and Fuglestead concluded that the first Wangara settlers in North Western Hausa came from Takedda, and were linked to the emergence of the Takedda kingdom30 (Fuglestead 1978: 332). Indeed, a major medieval trade and pilgrims’ route to Egypt and Mecca passed via Gao, Takedda, and Ghadames, through the Fezzan and Libya, and not through Hausa and Bornou and Kordofan as later.
45Lovejoy (1978: 176) claims that Songhay expansion brought the Hausa states in contact with the Western Sudan and concluded that since the Wangara in Bornu spoke Dendi, a Songhay dialect, “the foundation of Wangara settlements in Hausa cities was related to the expansion of this Songhayoriented commercial diaspora”. After the collapse of the Songhay kingdom 1590 this was rather a withdrawal and a compensation for the loss of political power. As the Moroccans controlled the middle valley, the Songhay operated from the Gourma-Dendi and the area around Ansongo-Ayorou-Tillabery-Niamey. Lovejoy never states that the Wangara were Songhay, but insists that they were “a Songhay-based commercial group” or a “Songhay merchant class” and that all others connecting them with the Soninke, Malinke or Dyula are off the mark. But Barth (1849-1857 (IV): 145), who was in the area more than hundred years earlier, argues that point documenting the linguistic change from Soninke to Hausa.
46The crucial period of the assimilation of the Wangara into the Songhay kingdom seems to be the century from 1492 to the Moroccan invasion of 1591, the period of the Askias. Yet, Wangara communities in Hausa are mentioned several decades before the ascent of the Songhay and the famous 1523 incursion by Mohammed Askia into Hausaland.
47According to Fuglestead important changes in Hausa socio-political structure took place during the late 15th century, namely the addition of a foreign element of Muslim clerics and merchants to the traditional animist land-priest-chiefship (Fuglestead 1978: 327). These elements sometimes became the occupants of the kingship, and joined the traditional kin groups thereby creating a “dual” structure of local and foreign power-sharing groups. The royal councils consisted from now on of indigenous priestchiefs from among the members of local lineages who elected kings from the incoming lineages. Fuglestead designated the incoming groups who imposed “kingship” on local kin groups as “Muslim clerics of Wangara origin”. Thus, for example, a certain Mohamed Korau, a Wangara, elected in 1492/3, became the first Muslim sarki of Katsina.
48The Volta basin has been important for the Wangara in several respects: it comprised some of the main gold-producing areas (Lobi, Banda) while being linked to others (in the Birim and Pra and Offin river basins, and in Ivory Coast); it marks the southern end of the long-distance trade route from Djenné and Timbuktu where precious goods from the forest zone (gold, kola) were produced; the Black Volta (aswada in Arabic) forms the border and link between the Mande-Diula and Hausa linguistic and economic spheres; it forms the territory of the Mole-Dagbane speakers, Mossi, Mamprusi, Dagomba and Dagari, who as state-forming groups were more receptive to Islam than the autochtonous, segmentary lineage societies.
- 31 Binger (1892: 100) in Kintampo assimilates them with the Ligbi; see also for Salaga (ibid.: 78, 92) (...)
49In contemporary Ghana, “Wangara” refers to Mande speakers and those believed to be of Mande origin and associated with trade. Dioula is even spoken in the border area with Ivory Coast, from Wa down to Wenchi, due to the close association with the important Islamic centers of Kong and Bouna. The Wangara communities further east–many of them named Wangara–along the Volta river, have adopted the local languages and/or speak Hausa, for example in Salaga. Following the familiar complex of “Market-Mosque-Medressa”, the Wangara founded the following colonies on the forest fringe: Bego, Kong, Bouna, Bole, Bondoukou and others31.
- 32 Following the frequent practice in everyday speech to drop inter-vocalic consonants, Bidugu (with a (...)
- 33 I believe the Ligbi-Huéla to be local Nafana converted to Islam since early times and adopting Diul (...)
50Bigou has many variant spellings (Bitugu, Bigu, Bitu, Bi’u, Bew, Beawu, or Beko, Begho, Biégo)32. It apparently goes back to Ligbi, Weila (Huéla) and Noumou elements classed as proto-Dioula by Delafosse and Tauxier, which spoke Southern Mande dialects and lived amongst such autochthonous peoples as Abron, Loro, Gbin, Koulango. If the Ligbi-Weila were indeed Mande-speakers–nowadays they speak standard Dioula–the linguistic proximity to either a Southern Mande or a Northern Mande language needs to be established before deciding on an early Mandeization by “proto-Dioula” of the forest belt33.
51Bigou’s foundation and destruction have been a matter of much speculation: there are radio-carbon dates from 1400 to 1650, of which I. Wilks (1982: 346) states that they were contemporary with Bigou. Y. Person (1964: 325-326, 332) dated the foundation of the Dioula community to the middle of the 16th century based on Mande genealogies (Keita, Kamara, Kourouma) and a west-east migration from the Manden through the upper Niger valley to settle the Dioma in the 15th century, followed by a 16th century emigration of the “Dioman-den” to the Sankaran, the Bate, Northern Konyan and the Wasoulou, and from there to the Mahou, South Konyan and Seguéla-Mankono.
52According to Lt. Benquey, first administrator of Bondoukou, Bigou was founded by a saintly person from the Huélas, around whom several Mande- Dyoula families grouped themselves: Kari-Dioula, Nénéya, Kamaya, Koumala, Dorobo, Donzo-Ouattara, Timité, Nigbi. There was also a quarter of Noumou (blacksmiths). Binger (1892: 161) claimed that “Bondoukou ou Bitougou est plus ancienne que Djenné: sa fondation est antérieure a 1043”, but this was contradicted by Benquey who, while leaving open whether Bitu was older, claimed that at least the settlement of Bondoukou was fairly recent:
“D’après les renseignements que j’ai recueillis, la date de la fondation de Bondoukou serait beaucoup plus récente. Voici du reste les renseignements tels qu’ils m’ont été donnés par l’almamy de Bondoukou, personnage religieux très instruit [. . .] Il existait autrefois à 50 kilomètres environ au nord-est de la ville de Bondoukou actuelle près de Banda (4) [note 4: appelée encore Foughoula (note du cap. Benquey)], (Gold Coast) une importante cité nommée Bégo. Elle avait été fondée par un saint personnage musulman de la tribu des Huéllas (5) [note 5: Ou Huélas (note du cap. Benquey)] qui avait fait le voyage de la Mecque et autour duquel vinrent se grouper plusieurs familles de la tribu des Mandé-Dyoulas. Voici leurs noms: Kari-Dioulas, Nénéyas, Kamayas, Koumalas, Dorobos ou Dérébous, Donzo- Ouattaras, Timités, Nigbis” (Benquey, according to Tauxier 1921: 67-68).
- 34 Usul Begho by Sheikh Kalidu Bamba (1966) quoted in Wilks (1982: 348), and Boutillier (1993: 291).
53The last imam of Begho, Sheikh Kalidu Bamba, kept a manuscript “Usul Bighu” which gives the original quarters in their order of installation: 1. Bamba, 2. Kamaghatay, 3. Timitay, 4. Bane, 5. Jabaghatay, 6. Tarawiri, 7. Kuribari, 8. Watara, 9. Kawtay34.
54The correct identification of the dyamou of Begho, and of their heirs at Bondoukou, will show that the islamized elements claiming Mande origins came from areas along the trade route to Djenné. This identification is made in a footnote in order to spare the reader the somewhat tedious discussion of details.
- 35 Begho’s Mande section, according to Tauxier, was founded by the Ligbi called Kalo-Dioula. I assume (...)
- 36 700 A.H. is inconsistent with 1006 A.H., unless the author, imam Yusuf Ibrahim Kamaghate of Dorma, (...)
55The Ligbi–assimilated to the Huéla founders–bear the dyamou Bamba, which designate the “crocodile”, their “tana” (taboo or interdit). Kamaghate and Timite are clan names found in the neighborhood of Djenné, in the Bwaba (Bobo Oulé) settlements, but also in the Korodougou. The Gbane seem to come from the west, the Djimimi around Kong. Diabaté (Kari- Dioula), Tarawiri (= Traoré), and Kurubari (Coulibaly) are Mande or Bambara dyamou which could come from the Manden or from the area around Segou, but are also among the founders of Kong. Watara is a surname for the founders of Kong and the Traoré of the entire region from Sikasso to Bobo. Kawtay (Kawuté) are Kambara (and imams in Larabanga)35. The break-up of Bigou, attested by local sources to 1594 or 1597, gave rise to the dispersion of its clans. According to Isnad al-Sudan a Watara, the 8th Diula group, according to Tauxier (1921), coming from Mecca installed himself at Bi’u in A.H. 1006 (1597/98) and provoked its destruction: “Bi’u fut détruit et [les gens] émigrèrent vers Maadi en l’an 700” (Niamkey 1985: 79)36.
56Following the destruction of Bigou, the clans spread into all directions, some returning to their homeland, others building a new town, Bondoukou, some 40 km to the north-west of the old one. The Ligbi–founders of Bouna–returned to their old settlements in the Banda hills; the Huéla returned to Sorobango, Genene and Damena in the Gyaman and to the Volta river; the Nafana and Donzo Ouattara also left for Bouna. Another source said that they went back to Segou and from there to Mande and returned via the West. Many Mande-Dioula colonies in the savannah belt claim Bigou as their predecessor: thus Salaga, Kintampo, Wenchi, Banda, Bouna, Bole and Bondoukou.
- 37 According to Konare (1980), by 1599 the valley is occupied by Peul and Bamana till Bamako, the Niam (...)
57The socio-political processes which took place in the 17th and 18th centuries between Mali and the southern savannahs are still not well understood. By the beginning of the 17th century, the Bambara began to leave their homeland in and around Nioro and moved into the middle Niger valley around Segou and San, and in turn displaced Bobo and others who moved towards the Wasoulou and Northern Ivory Coast (c. 1600-1650)37. There it encountered a west-eastern Mande movement, perhaps related to the occupation of the upper Niger valley (the provinces of Hama and Dioma) and the Sakaran, under the leadership of Diomanden lineages in the Mahou and Touba mountains. Those may have assimilated local Senoufo groups such as the Nafana and Nigbi in western, central and eastern Ivory Coast. The internal struggles of the Bamana realm led to formation of alliances which called to help some of these southern Mande elements, especially from Kong. The discussion whether Bigou was identical with the Bitw ou of the Tarikh es-Sudan will be skipped here, as I generally follow Wilks’ argument.
- 38 “Les familles de Ouattara, Daou, Barou, Kérou et Touré seraient venues de la région de Ségou... Les (...)
- 39 “Appelés par les Kamaghaté, les Bambara, liés à Kong et à Ségou, repoussèrent les Sénoufo dans le N (...)
58Kong is said to have been founded by the Nafana, but has probably contributed lineages to Bigou’s foundation who left it again after its destruction. An analysis of Kong’s lineages reveals a dual origin: first, Bambara dyamou which apparently came from the Segou-San area, namely Baro, Daou, Balo, Cissé, Touré38. Second, Mande dyamou like Kanté, Konaté, Kone who seem to have taken the detour via Guinea and the northwestern Ivory Coast through the Korodougou, Boron and Korhogo. If Y. Person (1968) was correct the Sanogo, Kamaghaté, Diabaghaté, Timité also took that direction even though they could also have formed part of the first migration39. According to Blondiaux (1897: 370), the Bamba from western Ivory Coast held the chieftaincies in the Worodougou, Wataradougou and Keso. The French authors from the colonial period maintain that they were Bambara, but I think they were rather Nafana or Ligbi who had adopted a Diula identity and Islam. Binger (1892) also points out numerous Diula elements in the Djimini, Djammala and Ano.
59An early islamization of Southern Mande elements (Nafana and Ligbi) in Kong seems to have taken place through individual muallim of Bobo-Dioula (Marka) origins–namely Saganogo, Kamaghaté, Timité origins– who were reinforced by members of their families from Bigou in the late 16th and 17th centuries.
- 40 Binger (1892: 324) claims that Sékou Ouattara took Kong only by 1790 (ibid.: 394).
- 41 “Au mois de djomada II (5 septembre-4 octobre 1739), on reçut de Dienné la nouvelle que l’armée des (...)
60The rise to power of the Coulibaly, and of Sekou Ouattara in Kong is not dated with uniformity. According to some, Sékou took power in Kong around 1615, and his brother Fa Maghan in Bobo in the same year, Binger however, dates the event to the end of the 18th century40. However, by the middle of the 18th century, Kong was involved in wars with the Bambara to the north and the Ashanti-Abron to the south. The Tedzkiret reports the invasion of Djenné by Famagh the Ouankoré in 173941. Some authors have associated Famagh to Sekou Ouattara’s brother and factional dispute among the Bambara (Konare 1980: 72). If the identification of Famagh with the brother of Sekou Ouattara of Kong is correct, the question arises why Kong attempted to occupy Djenné. Is it pure speculation that the Wangara from Kong attempted to remove trade obstructions imposed by Biton Coulibaly? However, it is said that the Coulibaly established themselves, having come from Biegou, in the neighborhood of Kong, at Ténégéra, before the arrival of Dé-Maghan and his son Tiéba, and gained them to fight against the animist Lasiri, king of Kong (Dé-Maghan died in 1670). Some even claim that Tiéba took refuge with Biton Coulibaly at Segou, and later gained his support (Niamkey 1982: 77, 80). This places the takeover of Kong by Sekou Ouattara around the beginning of the 18th century. According to the Gonja chronicles, Skou (Sheku) died in 1745.
61Kong, involved in wars against the neighboring Abron and Ashanti by the mid-18th century, seems to have recruited warriors from the Ano which already had a Dioula population. Local people were recruited as warriors, and assisted in the defense and expansion of Kong, whose leaders appealed to some of the maraboutic families scattered about after Bigou’s fall to assist them, with talismans among other things, against pagan land chiefs on whose lands in the Black Volta bend they established small colonies.
62Bouna is another one of those establishments which rose from indigenous village towards a Moslem trade towns (with the market-mosque-medressa complex), to be later invested by a centralized Dagomba Muslim chieftaincy. The Muslim section is like that of Bigou said to have been founded by the Ligbi-Huéla. Boutillier (1993: 290-291) established the following chronological order of clans: Ligbi (from Banda = Fougoula), Noumou, Gbane, Grafouté, Kamara and Jabaghte, Kurubari, Wattara and Cisse. Close links with Kong via its Kombala elements are evident.
- 42 According to Boutillier (1993: 282), Ligbi, Numu, Gbane, were the first dyamou to settle at Bouna. (...)
63Marty, another administrator, counted 10 quarters in 1922: Sissera, Ouattara Soura, Tourera, Kambarasoura, Nibisoura Bamba, Soukoulia soura, Tareorera, Granboute soura, Kouroubari soura, Kalediora soura42.
64The similarity to the clans of Kong and to those of Biegou is striking: Ouattara, Touré, Kambara (Kombala and other such lineages like Gbane, Grafouté), Ligbi, Traoré, Coulibaly are represented, while only the Timité are absent:
“Les Cissé occupent le quartier tout entier de Limamso qui tire son nom du fait que, institutionnellement, c’est un Cissé qui est l’imam de la mosquée du vendredi (missiriba), la grande mosquée, c’est-à-dire imam de la communauté musulmane dans son ensemble. En fait, ‘l’imamiat’ de Bouna n’a pas toujours appartenu au groupement cissé: fondé par les Kamara [. . .], l’imamiat fut occupé par eux pendant près de deux siècles, jusqu’à ce qu’ils le donnent aux Cissé peu après l’arrivée de ces derniers [. . .] Le patronyme Cissé qui peut évoquer une origine lointaine wangara ou soninké est très répandu dans les bassins des Volta mais [. . .] le lieu d’origine de leur dernière migration ne serait pas Bégho: ils arriveraient par étapes de l’ouest mais d’une région pas nettement déterminée se situant vers la haute vallée du Niger” (ibid.: 289).
65Apparently Boutillier mixes up the Kamara with Kamaghaté. The present Cissé imam of the Friday mosque told me that the Kamaghaté formerly held the imamate, but were later replaced by the Cissé, who were better trained in Islam.
- 43 Tauxier (1921) links the Karo-Dioula or Ligbi to the Karou in Dapper’s account of Cape Mount and to (...)
66Gbin, Loro and Nafana claim to be the original occupants, who later got joined by refugees from Bigou, in the following order: Hwela, Ligbi, Noumou and Dyoula before the Bono (Abron)43.
67On the basis of rights and customs, Tauxier has concluded that Gouro, Gbin, Loro and Nafana were the original settlers and “owners of the land”, since 5 sectors of Bondoukou township belong to Gouro, Nafana, G’bin, Loro. The Nafana receive sacrifices from the Dyoula, Hausa, Huéla, Noumou and Koulango at annual festivals, in recognition of the cession of land, and in turn hand a chicken to the chief of the Loro-Koulango, and one each to the chiefs of the G’bin and the Gouro (Tauxier 1921: 50-51).
68The Mande of Bondoukou came from Bigou, following the order given by Benquey: Dérébou, Kamayas, Nénéya, Koumala, Donzo-Ouattara, Timite, and some Huéla and Noumou; the Ligbi went to other villages of Banda (Fougoula), and returned with part of the Donzo-Ouattara to the Nasian and Bouna; the Kari-Dioula (Ligbi) went to Dorma, but later returned with the Abron.
- 44 Account by Malam Issa Idrissu Timité, Bole fieldnotes, December 27, 1995. “After the destruction of (...)
69The story of the foundation of Bole not only links it to Mande refugees from Bigou44 but also to the foundation of the Gonja chiefship. It claims that some refugees returned to Mali via Kong. Thus certain families in Mali e.g. the Sanogo, claim to have come from Kong and later they returned there through the northern Ivory Coast, Wangolo, Djerisu, Djimini, Mango (the Ano) to Bondoukou, and then crossed the Volta River and established themselves at Bole. The elder line of Timité thus is in Berekum and the younger one in Bole.
- 45 The sequel to this paper “Mande and the foundation of Gonja”.
70According to the tradition of the imams, the foundation of Gonja then is related to the break-up of Bigou and took place around 1700. Some of the ex-inhabitants of Bigou went to Kong, and from there raided the surrounding northern savannahs and gathered followers–Kamaté, Timite, Gbane and other Moslems–then went to the Ano and from there crossed the Volta river at Sakpe. They obtained the chieftaincy of Bole from the locals and thereafter established mosque, school, and market at Buipe. Kamaghte and Gbane later became the Sakpare imams of Daboya and other Gonja45 chiefdoms.
71By 1750 eight independent Gonja divisions existed among the local tribes, whose leaders claimed to be descendants of Jakpa, the founder hero. A central chieftaincy was created and occupied in rotation by some of the divisional chiefs. At that time the eastern divisions fought against the Dagomba driving them across the White Volta and assisted the Mamprusi with hired warriors from Ano, Kong and locals who established the Mande-Gonja colonies at Sansanne Mango and perhaps even in Borgou.