ARTICLES 95. THE ORIGINS OF THE MQSSI-DAGOMBA STATES by A.A. Illasu* In a recent article,Fage ascribed some of the misinter- pretations of the early history of the Moss I-Dagomba states to what he termed the "MossI - centred approach" by which con- clusions for the whole group of states are drawn exclusively on the basis of evidence from the Mossi states alone.' Thus did Marc (1909), Delafosse (1912), Tauxier (1917, 1924), and Dim Delobson (1932),2 -' attempt to date the foundations of the Whole group of states although they were unable to take account of evidence from Mamprusi and Dagomba, traditionally regarded as "senior" to the Mossi states of Ouagadougou, Yatonga, and Fada N'Grumah. Fage, as did Westermann (1952) and Prost (1953), took account of the Mamprusi and Dagomba evidence published by Tamakloe (1931), Blair and Johnson (1932) and Rattray (1932).3 He also consulted an unpublished work by D.V. Mackay,** an official of the Colonial Administration of the Gold Coast, which deals exclusively with Mamprusi. Unfortunately the Mamprusi traditions collected by Rattray and Mackay on which he relied, are, as he himself recognised, both fragmentary and demonstrably unreltable. Consequently, he makes a number of misleading conclusions In his extremely useful outline interpretation of, the early history of the Mossi-Dagomba states. The aim of this paper is to re- appraise this interpretation in the light of additions I have made to the fragmentary accounts of Mamprusi oral traditions collected earlier by Rattray and Mackay. The published works mentioned earlier as well as some archival material have also been consulted. The ethnically - related Mossi-Dagomba states were founded by "small bands of strangers'^ who migrated from the east or northeast of Lake Chad to the lands south of the Niger bend. The strangers were acquainted with the idea of chief- tainship as opposed to the politico-ritual organisation of the acephalous peoples they encountered. Because of their political and military superiority they consequently overran those scattered and independent peoples and revolutionized their political or tribal patterns by welding them into kingdoms. The most salient Dr. A.A Legon. Iliasu Is a Lecturer in the History Department, * . #»- , *> -*» ARTICLES 96. feature of this spectacular change was the office of a terri- torial and secular ruler, "an unheard - of conception", as Rattray put It, which replaced "the Immemorial institution of e ruler who was the high-priest of a totemic clan and dealt only in spiritual sanctions".6 The descendants of those state-forming strangers who still rule over the Mossi-Oagomba states have almost the same traditional accounts about their origins. Thus the lesser-known Mamprusi oral tradition tells much the same story as the more accessible and better known Dagomba oral tradition. Both point to a migration from somewhere east of Lake Chad; both recount the adventures of Tohazie (the "red hunter"), whose son, Kpo- gonumbo, married Sohiyini, the daughter of Abdul RahamanI, a king of Grumah; and both recognize Gbewa, the most famous issue of this marriage, as the great ancestor of the Mossl-Dagomba peoples. There is however a significant difference between the starting points of Dagomba and Mamprusi oral traditions. The former version nearly always opens with the migration from the east as well as the Saga of Tohazie and his relations with the earlier established Mande and Gur-speaking peoples south of the Niger bend. Mamprusi oral tradition on the other hand Invariably omits this prefatory chapter end starts with Gbewa's migration from Grumah territory to Pusiga from where he subdued the neighbouring peoples and, as would be argued later, founded Mamprusi, the oldest of the Mossi-Dagomba states. This differ- ence probably accounts for the misleading conclusion sometimes made that the great Gbewa is the fons et origo In Mamprusi oral tradition. The explanation for what can be termed the abridged and unabridged versions of their common oral tradition lies primarily In the scale and quality of their respective machinery for preser- ving and recounting their past. The Dagomba machinery is elaborate, accessible, and efficient. The state drummers or "lunsl", who form the core of this machinery, are a highly specialised segment of Dagomba society. They have a hierarchy of their own, at the head of which Is the Namoo Na, who, like any other important chief of the kingdom, has his own titled officers.7 They enjoy a fairly high social status partly because they are handsomely rewafded for their services and partly because they claim direct descent from a "nabia" or prince of royal blood, Bizung, son of Na Nyagse, the first Ya Na or Paramount chief of the Dagomba kingdom. As . court historians and drummers, they chant the genealogy and exploits ARTICLES 97. of the privileged; as custodians of traditional beliefs and constitutional practices, they make pronouncements of facts during grave political disputes. Theirs Is an arduous and painstaking task which requires long hours of patient appli- cation and practice and for this reason, they have their own time-honoured and efficient system of recruiting and training. Given the nature of the Oagomba machinery, particularly Its size and distribution - f or there Is hardly a Oagomba chief's village without Its drummer - It Is probably hot surprising that there has been available for a long time now a fairly detailed account of Dagomba oral traditions. The Mamprusl kingdom on the other-hand has nothing comparable either In grandeur or efficiency to the Dagomba machinery. There are, to be sure, some able drummers who nre as eagerly sought after as the Oagomba ones by connoisseurs during the festivals of the Mamprusl and Dagomba peoples. But these are the exception rather than the rule and In any case they are either natives of Dagomba or have acquired their skills at the feet of Oagomba drummers. Consequently the collector of Mamprusi oral tradition has to rely largely upon the "Naylrt Kpambaya" or state elders and counsellors. And the danger from this source Is perhaps not so much as the likelihood of Its being unreliable and Inaccurate, the elders being generally less well- equipped than the drummers to recollect past events. It Is rather that because of their estate within the kingdom, the elders seem to think that the chief function or ralson d'etre of the oral tradition Is to explain the existing political structure with its differentiated status between rulers and the ruled. They therefore attach more Importance to the great political advance which Gbewa made when after his migration to Puslga he changed his career as a leader of mercenaries to that of a state-builder. Only by this means which, In a sense, Is their charter, could they account for the origins of chieftaincy as well as the secular authority vested In their officers. Thus It was not uncommon for the Mamprusl state elders, when asked of the origins of their kingdom, to.start with Gbewa's exploits and the success of his arms over the peoples he met around Puslga and absorbed into his kingdom.9 It Is possible that both Rattray and Mackay, the pioneer collectors of Mamprusl oral tradition, relied exclusively upon the elders, hence their Incomplete and not entirely satisfactory accounts of Mamprusl oral traditions. , * •• I*-** ARTICLES 98. The much-shorter Mamprusi version, however, has at least one advantage over the unabridged Dagomba one, namely, it dwells at length upon Gbewa's conquests and even attempts to delimit the geographical extent of these conquests. According to this version, Gbewa left Grumah land after a disputed succession following the death of his maternal grandfather, Abdul Rahamani. Although his faction carried the day, he migrated with a large following to Sana or Sanga, south of Fada N'Grumah. From there he began the gradual conquest of the territories of the Busansi and Kusasi further south, and In due course he overran part of their territories. He then moved his base of operations from Sanga to Pusiga from where he came to exert power over, and levy tribute on, the less powerful peoples around him. As was the case with most other kingdoms similarly founded, his authority was more or less constantly effective only over a relatively small area around Pusiga. However Mamprusi oral tradition claims that this authority extended as far north as Sanga, as far south as present-day eastern Dagomba, then largely populated by the Konkomba, as far east as Mamprugu where the mother of Tohugu, one of Gbewa's illustrious sons, came from, and as far west as Zangu and Nabare, both east of the Walewale - Bolgatanga road. It is possible that even the Oaboya district, a later accretion to the state of Gonja, formed part of Gbewa's kingdom. It was thus an extensive domain, stretching to the Black as well as the White Volta. Gbewa ruled over this domain until his death. Here again, both Mamprusi and Dagomba traditions give the same accounts about the circumstances of his death. According to these accounts Gbewa had nine children, the eldest of whom was Kachiogu. But Kachiogu being a woman could not succeed her father, so Zirili, the eldest son, became the heir apparent. But Zirlll's impa- tience and love of adventure made him an unsuitable successor to a newly-founded kingdom which required a period of peace to consolidate. Gbewa therefore preferred a younger son, Kufogu, as his successor and made his preference known. When this was revealed to Zirili, he plotted arid killed Kufogu and so over- whelmed was Gbewa oh receiving news of this tragic event that he kept on moving restlessly upon the skins on which he sat until the ground around him opened and swallowed him, whereupon Zirili succeeded him. There is a material difference between the Dagomba and Mamprusi accounts regarding Zirili's reign. According to the ARTICLES 99. former account, Zirill reigned peacefully until his death; the Mamprusl version, which seems the more plausible one, claims that ZlrtlI's reign ended abruptly and In tragic circumstances. According to this version, soon after he had succeeded Gbewa, a chief of a Grumah village died and ZfrII I, through Sohtyfnf, his grandmother, claimed the right of succession. Partly because of the accepted practice at that time and true to his own character, Zirlll took up arms to enforce his claims. The expeditionary force he led into Grumah territory was however decisively routed and a number of prisoners, Including himself, were taken. This disaster was a signal for the Irreconcilable BusansI and Kusasi to rebel against their rufers. Faced at once with external and internal disaster, and without a leader, the elders could afford neither the luxury of a funeral for Zirlll nor the time-consuming but customary process of electing a successor. They therefore drafted Tohugu, the eldest surviving son of Gbewa to succeed Zirlll.'O But Tohugu's brothers, Inspired and led by SItobu, challenged the const I tut tonality of these preceding* and a fratricidal struggle therefore ensued. With the Busansi and Kusasi up In arms and uncertain of support from his subjects In the capital, Tohugu, upon the advice of the elders, fled to Mamprugu to solicit the support of his uncles. In this he set a precedent, for In later years, princes who either found themselves in situations similar to Tohugu's or were parties to disputed successions invariably appealed for assistance from their maternal uncles. So common was this practice that it became a convention amongst the princes of royal blood and the divisional chiefs, much to the benefit of the kingdom as a whole. But for this convention, civil wars would not have been what they actually were in the history of the Mamprusl kingdom, namely, fratricidal struggles involving a handful of mercenaries'' and the villages of the princes' uncles. And since ft was not uncommon that most of these villages were of no consequence In size, these struggles were generally concluded in a matter of days, so that the kingdom was spared all the usual horrors of civil wars - ruined villages, plundered homes, downtrodden crops and attendant famine. Tohugu's flight took him first to Gambaga In the south- eastern part of the kingdom and then eastwards to Mamprugu. Sitobu and his followers pursued him only to Gambaga and after a brief stay there, they proceeded westwards to Nabare and then southwards to Vend I Dabarl. It was from this former capital of the Dagomba kingdom that Nyagse, the eldest son of Sltobu, sub- dued the neighbouring "Black" Dagomba and Konkomba peoples to 3,. ARTICLES 100. extend and consolidate the Dagomba kingdom founded by his father J2 Meanwhile Tohugu ruled over the kingdom founded by Gbewa, with Mamprugu as his capital. The name of this little village was later given to the whole of that kingdom. Tohugu's successor, Na 2ob.z.laj however, removed the capital to Gambaga from where, according to Mamprusi oral tradition, Yampoga a princess of royal blood, one day rode a stallion northwards, where she met and married a hunter and their son became the founder of the Mossi dynasty. Moss I oral traditions give credence to this portion of Mamprusi tradition. According to the most widespread version of their origin, their pregenetrlx was not Yampoga but Nyennenga, the daughter of Naba Nedega or Nedega a "Dagomba" chief who lived at Gambaga. He valued his daughter's warring skill so highly that he would not grant her permission to marry. She therefore fled from Gambaga and rode northwards where she met a man called Rtatle, who according to some traditions was the son of the king of Mall, and according to others was a Busanga hunter. They had a son whom they named Ouedraogo ("stallion") in honour of the horse that had carried Nyennenga In her flight to the north. A few years later, Nyennenga sent Ouedraogo to Gambaga to visit his uncles. On his departure from Gambaga to his parents, Ouedraogo was accompanied by several "Dagomba" horsemen with whose help he drove the Busamsi from Tenkodogo (TInkurugu In Mamprusi) andmade it his capital, With the assistance of more "Dagomba" horsemen from Gambaga, Ouedraogo subsequently subdued the Grunsl, NIntsi, Foulse and Klrplsi populations around him. His dynasty event- ually proliferated in all directions: one of his sons, Rawa, established the kingdom of Zandona to the north of Tenkodogo; it later became the kingdom of Yatenga under Rawa's classI- ficatory brother, Yadega. In the east, Rawa's brother Diaba Lompo founded Fada N'Grumah while a nephew Oubrl founded Ouagadougou in the west.13 Dagomba tradition does not recall these events and it has been suggested that this was partly because Mamprusi being "the de jure father-state /*of the Moss!-Dagomba states_7 needed to retain a ritual relationship with the Moss! states, whereas Dagomba did not" and partly because "geographically Mamprust Intervenes between Dagomba and the Mossi lands".'1* I find this explanation unconvincing. The failure of Dagomba tradition to -0 [** I* ARTICLES 101. recall the Moss! breakaway can be explained qufte simply on the grounds (i) that the Mossi states derived not from Dagomba but the Mamprusi line and (II) that since the Dagomba split from Mamprusi had preceded the Mossi one, her traditions could not possibly recall events which concerned only the Mamprusi and the Mossi. In other words, Dagomba tradition Is silent over the Mossi breakaway precisely for the same reason that Mamprusi tradition makes no mention of the circumstances, retained in virtually identical terms In Dagomba and Nanumba traditions, that led Nmantambo, one of Sltobo's brothers, to go "away from him in anger" and found the state of Nanumba.15 Secondly, Nedega or Naba Nedega is referred to In Mossi tradition as "a Dagomba chief at Gambaga" while the men who accompanied Ouedraogo after his visit to Gambaga are said to have been "Dagomba horsemen". It has been suggested that this apparent confusion of Mamprusi with Dagomba arises from the fact that Dagomba was the better known kingdom following "the early weakness of Mamprusi vis-a-vis Dagomba".'6 There Is a less interesting but more accurate explanation for this confusion. The Mamprusi call themselves Dagbamba, a name anglicized to Dagomba; "Mamprusi", derived no doubt from Mamprugu the name of their kingdom, is the name given them by outsiders. They call their neighbours to the south, the so-called Dagomba, the "Yoba" from the word "Yooba" which means forest people. The name Dagbamba was once used of Fada N'Grumah also, and it is said that it was originally the Grumah name for the peoples living to the south of their territories which the ancestors of the Mamprusi and Dagomba took after they had conquered these peoples.'7 This probably explains why the two peoples are still known by one name among certain tribes; thus in Gonja; the; term "Nwong" refers both to the Mamprusi and the Dagomba and similarly, in the History of Wa Yendl, the capital of Dagomba and Nalerigu, the Mamprusi capital, are spoken of as the capitals of two tribes of the Dagomba.'", Thirdly and finally, it seems conceivable that Na Gbewa (Dagomba tradition) is the same as Nedega or Naba Nedega (Mossi tradition^ and Bawa or Bwongwa (Mamprusi tradition) The Mossi variants are probably due to euphony and habit since Na Gbewa sounds much like Nedega or Naba Nedega while the Mamprusi ones are obviously the mis-spellings of the name Gbewa by Rattray and Mackay respectively. ARTICLES 102. It Is evident from the above comparison of oral traditions that the emergence of the Mossl-Dagomba states was by no means contemporaneous. The first of them to emerge was the Mamprusl kingdom: It was founded by Gbewa but It was not known by that name until Tohugu made Mamprugu Its capital. Thus, unlike Sltobu and possibly Nya'gse, Tohugu never founded a kingdom; all that he did or, more accurately, was dj-Iven to do, was to remove the capital of an established kingdom from Puslga to Mamprugu. This was not an uncommon practice and, indeed, Tdhugu's immediate successor, Na.Zobzia,- removed the capital from Mamprugu to Gambaga while Na Atabia took It finally to Nalerigu.19 Yet neither Na .Zobzia nor Na Atabia has ever been credited with founding a kingdom. It is thus misleading to conclude that Nyagse's kingdom was founded about the same time as that over which Tohugu ruled.20 It is even doubtful, considering that Tohugu was a generation removed from Nyagse, whether they even ruled contemporaneously. The immediate off-shoots of the Mamprusl kingdom Were Dagomba and Tenkodogo; the latter in turn gave rise to Yatenga, Fada N'Grumah and Ouagadougou. The flssionary sequence is not clear but it seems that Nyagse's wars of expansion occurred more or less about the same time that Ouedraogo drove the Busansl ''''"''""- from, and established his dynasty at, Tenkodogo, from where, as mentioned earlier, it proliferated In all directions. The Moss! bards relate that while Ouedraogo sent out Rewa and 01aba Lompo to carve out kingdoms for themselves, he detained their younger brother, Zoungourana, at Tenkodogo because He did not wish the young man to leave him. Zoungourana subsequently married a Nimisi woman called Poughtoenga and she gave birth to Oubrf, who along with Ouedraogo, became known as one of the two founders of the Ouagadougou dynasty. It seems then that while Yatenga and Fada N'Grumah were founded almost contempora- neously, the pacification of the kingdom of Ouagadougou took at least a generation more to complete. This, quite apart from the fact that OubrI was the nephew of both Rawa and Diaba Lompd, explains why Ouagadougou has never been accorded traditional primacy though It came to eclipse the other Moss I states, the Mamprusl kingdom on the other hand never established Its ascen- dency over the other Mossl-Dagomba states; on the contrary, it appears that from about the middle of the 16th century onwards, It became the least political force amongst these states. Yet it has always been regarded as the "father-state11 by both the Dagomba and Moss I kingdoms and its royal tine as the ten Ior of • ** ARTICLES 103. the Mossi-Dagomba royal lines descending from Gbewa. The explanation for this - and it is an explanation which is supported by "oral tradition - is that the Mamprusi kingdom is the oldest amongst the Mossi-Dagomba group of states, having been founded by the great Gbewa himself. Finally, a word about the tentative dates for Gbewa and the emergence of the Mossi-Dagomba states, these vary widely, as can be judged even from the dates for Oubri alone. According to Marc, Oubri appears to have lived about the middle of the 14th century. Delafosse places him In the 11th century. Tauxler first of all suggests the 12th century, but after comparing the dates of several scholars, he revised this to the beginning of the I4tl> century. And finally Fage dates him to c, 1500-on his reckoning that Dagomba emerged c.1480.21 Nearly all these dates - the possible exception Is Fage's - are based largely on the sum total of theoretical average relgn- lengths and the number of rulers on the regnal 1Ists of the various kingdoms. The limitations of this type of analysis are as weighty as they are obvious and one cannot therefore select among the alternative dates with any degree of confidence. In any case, the eleventh or fourteenth century dates for Oubri are, as will be shown later, unsubstantiated by evidence from Mamprusi and Dagomba. Fage's dates on the other hand are quite satisfactory* but as will be shortly shown, his main evidence in arriving at these dates is open to serious objections. He postulates a close link between the emergence of the Mossi- Dagomba states and the re-establishment of imperial power in the Niger bend under the Songhai kings Sonnl All (c.1465-92) and Ask Ia Muhammad (c.1493-1528). According to this thesis, the invaders, whose descendants later founded this complex of states, entered the lands south of the Niger bend about the 13th century; that for the most part they either took service In the armies of the Mali empire or organized raids against established towns of the Niger valley; and that these raids came to an end when they provoked serious retaliation, such as those of c.1498, 1549-50, 1561-2 and c.1575. "It seems likely", observed Fage, "that this change In the balance of power In the Niger valley may have pushed the early Mossi-Dagomba southwards and have inclined them to seek profit from the levying of tribute on the kingship groups of the Upper Voita basin rather than from raiding northwards"." effective foundation of the Dagomba state In its present area to C.1480. This conclusion appears to be confirmed by Tamakloe's On this reckoning, Fage based the ARTICLES \0k. regnal chronology, as revised by Fage, In which Na Nyagse, the first Ya Na of the Dagomba state probably ruled, not In l*tl6-32 as Tamakloe suggested, but from c.1^76-1^92.23 If one accepts the revised dates of c.1^76-92 for Nyagse as well as Fage's suggestion that the ancestors of the Mossi- Dagomba peoples first entered the lands south of the Niger bend sometime during the 13th century, then one Is left with the strange spectacle of a group of state-forming peoples, who for almost two centuries were unable to establish their hegemony over the kinship groups of the Upper Volta basin, a feat they accomplished within only three years after pressure from the north. Thus unless the almost two centuries of apparent Inactivity, that Is, vis-a-vis the formation of states, Is accounted for, It seems that the more probable consequence of the re-establishment of Imperial power In the Niger valley was not the emergence of any new states as Fage believes, but rather consolidation of existing ones within more defensible boundaries. This Interpretation Is borne but by even the most cursory account of the Mossl confrontations with the Sudanese empires of Mall and Songhay. The Mossl were, to begin with, "propelled by a powerful drive of expanding conquest"2*+ northwards. How far north they got Is an open question but the claim in.the 17th century Timbuktu Tarikhs that they reached Timbuktu and Its neighbouring regions In C.1333A Is probably an exaggeration.25 This first phase of expansion came to an end when the invaders were driven southwards where they consolidated their positions In Grumah and Yange to the east of Ouagadougou and among the non- Mossi populations of Ouahigouya. But after a while, they felt sufficiently strong to relnvade the Niger valley and It was during this second period of invasion that they reached eastern Massina and Lake Debo In c.lUOO, Benka in c.1^33 and Walata In 1^77-83. This second period of success also came to an end with the restoration of Imperial power towards the close of the 15th century. Again the consequences for the Mossl ambitions of territorial expansion were the same: they were compelled to abandon part of their conquests, this time all territory north of Ouahigouya, and withdraw Into their present boundaries.26 In short the evidence seems to suggest that the significance of the re-establIshment of Imperial power In the Niger bend was that It merely determined the northern limits of the Moss! domains, and not their emergence. Fage's Interpretation of the origins of these states also requires some comment. According to him they emerged primarily , Jfe- ARTICLES 105. because the Mossf-Dagomba peoples had been compelled to seek profit further south, after the so-called raids, hitherto their main source of profit, had ended with the restoration of imperial power In the Niger valley.27 This interpretation ignores the most Impressive and significant point about the traditions of migrations recountered earlier in this paper. This is the evolution of states following the movement of populations especially after disputed successions. Thus dTd Gbewa migrate from Grumah territory to Puslga, Tohugu and Sltobu from Puslga to Mamprugu and Yendl Dabari respectively, and Nmantambo from Bagale to Blmblla. Each of these movements resulted in the foundation of new states. So also did the much later migrations of peoples In the Mamprusl and Moss! states*: the Mamprusf principal Ities of Janga and Tongo, for Instance, were founded early in the 18th century by the supporters of unsuccessful candidates to the paramountcy.*" Similarly the Moss I principal itIes of Boulsa, Boussouma, Conqulzftenga, and Yako emerged as a result of migrations following disputed success ions.29 It would $«em then, that what Fortes termed "the dynamics of constant movement"30 was built Into Mossi-Dagomba social structure and that Its raison d'etre at the political level was to drive away from the seat of government certain sections of the population, e.g. supporters of possible competitors, or rejected candidates,.for chJefshlp. The immediate result of these expulsions was the reduction of tension and therefore the maintenance of political stability in the state. Ultimately, however, they led to the formation of new states or principalities. It Is, perhaps, to this built-in mechamfsm of the Mossi-Dagomba social structure, rather than the desire to seek profit that one must look for the migrations of the Mossi-Dagomba peoples and the evolution of their states. There Is no suggestion here that either the mechanism or Its consequent process of state-format Ion Is peculiar to the Mossi-Dagomba states although It might be pointed out that In the case of these states, the splinter groups maintained links with, and voluntarily recognized the primacy of, the parent-state. The suggestion is that In the case of the Mossf-Dagomba peoples, ft probably offers a more fruitful explanation for their migrations than does any other interpretation. It is, however, to an economic Interpretation that one might turn for the evolution of these states in that part of the Volta basin. It is not certain when the kola nut trade ARTICLES 106. between the northern areas of Ghana and Senegal, Mall and the Hausa states began.31 According to Delafosse the Dyula had been trading In kola with northern Ghana long before the 11th century, and it was In the course of that trade that Islam was first Implanted there at the beginning of the U th century. Apart from the view of Delafosse, there Is also that of the Kano chronicler: according to this authority the nuts were first brought to Hausaiand during the reign of Dauda, the 15th century king of Kano, who ruled from 1^21-1^38. Perhaps, as in the case of the Introduction of cowries Into Hausaiand, the chronicler of Kano Is most probably wrong over the date of the entry of kola. It seems then that the trade began before the 15th century, possibly the 13th or 14th centuries, as suggested by one authority.32 Could the desire to control this trade not have influenced the evolution of the Mossl-Dagomba states on the Upper Volta basin? In conclusion, It Is necessary to place the expeditions Into the Niger valley In their truer perspective. Strictly speaking, these expeditions were undertaken not by the entire Moss I but by those of Yatenga, the northern-most of the Moss I states. These northern Mossl had earlier under Yadega claimed complete Independence from Ouagadougou and after absorbing all the lands of the kingdom of Zandona, they began to extend their territory northwards. It was In the course of these wars of expansion that the Yatenga rulers Bonga and Nasare carried the northern Mossl to Masslna in c.HtOO and Walata in 1^77-83 respectively. The evidence, In short, does not support the view that besides the northern Mossi in Yatenga, those of Ouagadougou and Fada N'Grumah in the south, much less the Mamprusi and Dagomba, took part in the Niger expeditions. These southern Mossl as well as the Mamprusi and Dagomba have no traditions of any loss of their territory following the restoration of imperial power in the Niger bend. The Dagomba, for Instance, lost territory to the Gonja along their western frontier, the Ouagadougou Mossl and the Mamprusi hardly at all, the latter, thanks to their geographical position and the Chakosi mercenaries. It was only the lands of the Yatenga Mossi that were reduced as a result of the change In the balance of power In the Niger valley towards the close of the 15th century. It would therefore seem that any attempt to date the emergence of the entire Mossl-Dagomba states from an event which undoubtedly has relevance for the history of only one of them, Is yet another example of the "Mossl-centred approach" which Inspired Fage's ARTICLES 107- paper and which, as mentioned earlier, is partly responsible for the misinterpretation of the early history of those states. It was earlier mentioned that Fage's date of 1^79-92 for Nyagse is more accurate than Tamakloe's 1416-32. It now remains to be seen how relevant 1^79-92 Is to the emergence of the Mossi-Dagomba states. Strictly speaking, Nyagse consol(dated rather than founded the kingdom of Dagomba; its true founder was Sitobu, the father of Nyagse, The Dagomba kingdom therefore emerged soon after the initial diaspora from Pusiga. Thus a foundation date in the first half of the 15th century or at the latest c.1^50 for Dagomba might not be . put of place. But Nyagse, like Ouedraogo, Is traditionally two generations removed from Gbewa, whose reign Is said to have been a particularly long one. On this not entirely satisfactory reckoning, It would seem that the kingdom of Mamprugu emerged early in the 14th century; the date cannot be later than c.1350, but may be a little earlier. On the same reckoning, It is pfobable that by c.1480 at the latest, the Moss I states of Fada N'Grumah and Yatenga had taken shape as determinate political entities while the foundations of Ouagadougou had been laid by Ouedraogo. The foundation dates just mentioned ere substantiated by evidence from the regnal IIsts of these states. A fairly sound chronology has been worked out for Mamprusi and Oagomba for the period c.1700-c.1900.33 In those two hundred years both states had the same number of generations (five) and the same number of rulers; (thirteen)•.}? This works out to forty years for a generation arid ah average of fifteen and half years< per reign. A similar comparison cannot be made for th>.perfod prior to c. 1700 largely because the Mamprusi regnal 1 ^st is demonstrably defective. Not only does ft have eight rulers•< In five generations compared, with: Dagomba's1 nineteen rulers In eight generations; there is. also, unlike the post c.1700 . period, no agreement either on the names of the chiefs or the •• order in which they reigned;35 No'such discrepancies exist In the Dagomba I ist which, has •'twenty-nine chiefs back from c.190,0. If the average of fifteen and a half years per reign Is taken, Nyagse's reign will fall In the middle of the fifteen century, that is about the same; date reached by Fage. Coming now to'the Mossf states, Ouagadougou records eighteen generations from Oubrl to c.1900 and Yatenga twelve ARTICLES 108. The latter figure compares favour- or thirteen generations. ably with Dagomba's eleven since Nyagse who, as mentioned earlier, reigned about the same time that Yatenga was founded.37 By correlating traditions of Yatenga and Bambara, Tauxler dates Naba Kango of Yatenga to 175^-87.38 This ruler was the last of the four generations back from c.1900. The following three generations ruled for about one hundred and fifteen years, or an average of thirty-eight years per generation, only two years less than Mamprusi and Dagomba. If one calculates thirty-eight years for a generation, Yatenga must have been founded about the middle of the 15th century, which compares favourably with c.l*t80 mentioned earlier. On the basis of c.1^50 for Yatenga - which is applicable to Fad a N'Grumah - Oubrt who Is a generation removed from Rawa and Olabo Lompo, the founders of these two Moss I states, can be dated to c.1500. As Oubri is also traditionally five generations removed from Gbewa, the kingdom of Mamprusi probably emerged during the early decades of the fourteenth century. The least hazardous conclusion one can therefore draw Is that this complex of states emerged at various times between the first half of the fourteenth and the close of the fifteenth centuries. 4- ARTICLES 109. FOOTNOTES J.D. Fage, "Reflections on the Early History of the Mossi-Dagomba Group of States". J. Vansina ed.,- The Historian in Tropical Africa (Oxford, 196*0, pp. 177-189. L. Marc, Le Pays Mossf (Paris, 1909; M. Delafosse, Haut - Senegal-Niger Qarls, 1912), 3 Vols.; L. Tauxier, Le Npir de Yate'nga (Paris, 1917) and Nouvelles Notes sur Ie Moss? et le Gourounsi (Paris, 1924) A. Delobsom L'Empire du Mogho - Naba (Paris, 1932) and L. Frobenlus, The Voice of Afr ica, "Uondon, 1913). E.F. Tamakloe, 'A Brief History of the Oagomba People1 (Accra, 1931) and R.S. Rattray, Tribes of the Ashanti Hinterland (Oxford, 1932), 2 VolT ~" D.V. Mackay, lsFhe Mamprusf". (Unpublished Manuscript1:) Ghana National Archives (G.N.A.), Accra. Rattray, 1932, Vol.1, p. Ibid. C. Oppong, "A Preliminary Account of the Role and Recruit- ment of Drummers in Dagbon". Research Review tfnstItute of African Studies, Legdn), Vol.6 No.I, pp. 38-51. Another account claims that the drummers came from a place called Bizung near Dtarf. According to this version, Na Luro (11th Ya Na) built a bridge over a river and wanted the deed to be recorded. He therefore summoned a drummer called Bezung whom he entitled Namoo - Na, the first chief drummer and reputed ancestor of all present day drummers in Dagbon, Ibid. Compare this to the tendency among Muslim writers to regard Muhammad Zangina (16th Ya Na or Paramount chief of the Dagomba State) as the first king and founder of the State because he was the first Ya Na to embrace Islam. 5. 6. 7. ARTICLES 110. See I. Wilks, "A Note on the Early Spread of Islam in Dagomba". Transactions of the Historical Society of Ghana, Voi . V IM (1965), pp. 67-98. It is said that Tohugu at first declined the honour because being a modest person, he felt unequal to the grave problems confronting the kingdom as well as the task of ruling over it. Pressure was, however, brought about him to accept. His reluctance has formed part of Mamprusl tradition: the Nayiri-elect Is customarily expected to show a token of resistance when approached by the king-makers and It is to overcome his resistance that the king-makers have to "arrest11 him. For instance the Chakosi and the Tampolensi who were settled at Sansane Mamgu and Langblnsl respectively. Both peoples made significant contributions to Mamprus! history. The Chakosi, for example, protected the northern and north-eastern flanks of the kingdom. It Is also possible that It was because of their presence within the kingdom that even Samoury had to skirt Mamprusl during his triumphant drive across the Upper Niger basin. The Tampolensi supplied the officers when the wing of musketeers or "Kambonsl" was Introduced Into the Mamprusl State army. See my "The Kambonsl of Mamprusl and Dagomba", Department of History (Legon), Paper, October, 1968. The story of Tohugu's flight and the foundation of the Dagomba State has been greatly abbreviated here. For a fuller account, see Tamakloe, 1931, pp. 11-17. For a fuller account of these accounts see, In addition to the sources listed under footnote 2 above, E.P. Skinner, The Moss I of the Upper Voita (Stanford, 1964), pp. 7-12 and footnote 14, p.205. 10. 11. 12. 13. 14. Fage, 1965, p.185. 15. Tamakloe, 1931, p.14 and D. Talt, "A History of Nanumbe" (Unpublished Manuscript.) Institute of African Studies, Legon. ARTICLES 111. JK* !*>* 16. 17. 18. 19. Fage, 196*f, p.185. Information of Yldana Yelzori of Gambage, 15/3/69. See also Susan Drucker, "Colloque sur les cultures Voltalques" - Recherches Voltaiques, 8, 1967, p.79. J. Goody and K. Arhln, ed., (Legon, 1965), p.23. "~" "Ashantl and the North-West" There is no oral tradition regarding the removal of the capital from Mamprugu to Gambaga; it Is likely that this was politically and strategically motivated- unlike Mamprugu which was on the eastern periphery of the kingdom, Gambega was almost In its heartland. It was therefore • more suitable administrative centre and, considering the Grumah Incursions on the eastern marches of the kingdom, a more defensible position. The transfer from Gambaga to Nalerfgu, the present capital, has been variously explained. One tradition says that the court was transferred after Na Atabia had repelled a Grumah Invasion which reached as far as Nalerfgu (see Mackay, Unpub!fshed Manuscripts). This may explain why it was possible to establish the court at Nalerfgu, freed from Grumah threat, but not very it was necessary to remove from Gambaga. Another version is that Na Atabia was attracted to Nalerfgu district In his search for vast fertile farming lands to raise crops and sustain his Chakosl mercenaries. (Information of Yisifu, Kpanarana or Custodian of State Spears, Nalerfgu, 15/8/68). Yet another Is that with the development of the north-east trade-route to Hausa lands and the growth of the Muslim community at Gambaga, which became more and more open to strangers, Gambaga may have been regarded •* unsuitable for the residence of a Paramount chief. H» Atabia therefore chose Nalerfgu, five miles away to the east, because it was at once far enough from the cosmo- polttan atmosphere of Gambaga and close enough to secure control over this trading centre. (See N. Levtzlon, Muslims and Chiefs In West Africa (Oxford, 1968), p.129. 20. But see Fage, I96*f, p. 185 where he argues, rather uncon- vtnclngly, I think, that Dagomba and Mamprusl emerged contemporaneously In c.t480. ARTICLES 112. 21. Marc. 1909, pp.131 and 136; Delafosse, 1912, I, 306 ff; L. Tauxier, La Noir du Soudan (Paris, 1912), 458 ff; 1917, 16 ff, and 1924, p.672; and Fage, 196**, pp.182 and 187. 22. Fage, 1964, p.178. 23. ib]d., pp. 179-181. 24. Tauxier, 1912, p.459. 25. On this issue and on whether the history of the Moss I of the Tarlkhs is an integral part of that of the founders of the Moss I States of the Upper Volta Basin, see M. Izard, Introduction A L'Histoire Des Royaumes MossI (Paris, 1970), I, 35-70. 26. Ibid., and Skinner, 196**, p.9. 27. Cf. this Interpretation of the MossI expeditions to his "vTew that the Asante wars of conquest between c.1700 and 1824 were little more than raids in which the inland peoples were captured and sold Into slavery while the southern peoples were attacked to clear the paths to the European forts on the coast. J.D. Fage, Ghana: A Historical Inter- pretation (Madison, 1959), p.55 and Introduction to the History of West Africa (London, 1962), p.£7. 28. See for Instance, Rattray, 1932, Vol. II, pp. 340-1. 29. Skinner, 1964, p.10 and Tauxier, 1917, pp. 55-75. 30. 31. 32. M. Fortes, "Colioque sur las cultures Voitaiques". Recherches Voitatques, 8, 1967, p.174. A. Boahen, "The Ghana Kola Trade". Ghana Notes and Queries, No.1, January-April, 1961, pp. 8-14. Ibid. Extensive excavations at Klsare and Mpatia strongly suggest that these places were for a long time directly linked with the flourishing north-east trade route to the Hausa States. See R.D. Mathewson, "Klsare: A Preliminary Report". West African Archaeological Newsletter, No.3, (0 4 . ARTICLES M3. lyoa, pp. 22-25 and R.N. York etc. Archaeology In the Volta Basin, 1963-1966 (Legon, 1967), pp. 37-45. 33. Levtzlon, 1968, pp. 194-203. *• 34. 35. 36. Levt2lon (1968, p.199) gives back from c.1900, twelve and thirteen chiefs for Mamprusi and Dagomba respectively. Mamprusl also has thirteen so that between c.!700-c.1900, the two states had the same number of generalIons and the same number of rulers. See my "Mamprusl Regnal List and Chronology: A PrelImlnary Analysis (forthcoming). See regnal lists by Rattray (1932), Mackay (unpublished manuscripts) and R.A. Irvine, 1898, G.N.A., 1371/53, Accra. Ouagadougou's high number (18) of generations Is probably due to filial succession during the early period of her history. This was,, however, unusual for a colleteral mode of succession (and Indeed Inheritance) Is the rule among the Mossl-Dagomba States. 37. Yatenga, however, had more rulers (44) than Dagomba (29) from Nyagse. The higher figure Is explicable by both Internal and external wars. 38. Tauxler, 1917, pp. 667-8 and Levtzlon, 1968, p.200.