|
HOME ~
ETUSIVULLE BACK TO MY PUBLICATIONS ~ TAKAISIN JULKAISUIHINI Not Quite Venus From the
Waves: IMPORTANT NOTE:
This version is not identical to the printed version of
1996. Several printing errors are corrected and there are
a few additions, including subtitles. The arguments are,
however, unchanged. Following the good academic manners,
I expect that those who wish to quote the contents of
this text or to create a hyperlink to this page, should
introduce themselves: pekka.masonen@uta.fi Yet the conquest and destruction of Ghana by Almoravid invaders constitute one of the myths which still populate African historiography, like the wonderful voyage of Hanno to the Bight of Biafra, which was carried over from classical Graeco-Roman texts into modern European literature as early as 1533. Since then the story of Hanno has been used for various purposes by western Africanists, for instance in explaining the diffusion of iron technology into sub-Saharan Africa [3]. Just the same, no definite evidence has yet been found for any Carthaginian sailings along the West African coast [4], except the Periplus of Hanno itself, which seems to be a literary composition drawn from earlier classical sources [5]. A reason for the popularity of Hanno, and other such stories in African historiography, has been that many modern writers have been content with using the previous secondary literature only, instead of examining carefully all the available primary sources. Consequently, many subjective and hypothetical assumptions created by previous scholars, working on the basis of even less evidence, have been transferred from one body of research into another. Finally, their origin forgotten, stories like the voyage of Hanno have become established historical facts, through constant repetition in the authorized literature [6]. Similarly, there is no direct evidence for
any conquest, still less a violent and destructive
conquest, of Ghana by the Almoravids [7]. The conquest hypothesis is a European
creation. It was not born of a sudden springing,
fullblown, from the sea, such as was the birth of Venus
in ancient mythology: nevertheless, when the hypothesis
at last assumed its definitive form, in 1912, nearly 850
years after the alleged event, there was a certain modest
drama to that appearance. The purpose of this paper is to
show - by pursuing the textual genealogy, the isnad,
in the European literature on the West African past, as
the story passed from scholar to scholar, altering a
little each time the baton changed hands - how the
hypothesis came to be. In order to understand this
process we need to ask, who invented the hypothesis using
what evidence, and working in which ideological
environment? That is to say, to ask the very basic
questions that every historian should apply to his
sources. Furthermore, this survey also reveals how
important it is to start reconsidering in a wider context
the origin of our conceptions concerning early West
African history. Planting the Seeds: Leo Africanus, Luis del Mármol Carvajal, and Moroccan Propaganda The first seeds, within European learning, of the conquest hypothesis were sown by Leo Africanus [*]. A Spanish Muslim, who lived for a considerable time in Morocco, Leo ended up at the Papal Court in Rome, where he wrote, in rather corrupt Italian, his Descrittione dell' Africa, in 1526 [8]. This was first published in Venice, in 1550 [9]. Several Italian reprints followed, as well as translations into French, Latin, English, and finally Dutch and German [10]. In the western world, Leo came to be regarded as the highest authority on the geography and history of Africa, and he maintained this position until the early 19th century [11]. Even then it was chiefly his geographical ideas which had to be abandoned, overtaken by new European exploration in Africa. In the historiography of Africa, Leo's influence lasted much longer. According to Leo Africanus, the Land of the Blacks, including Guechet and Cano [Ghana], was first opened up in AH 380/1012 AD, after the arrival in North Africa of a certain Muslim [12]. The Land of the Blacks was then "inhabited by people, who lived liked brutes, without kings, lords, republics, government and any customs, and without knowing husbandry" [13]. Continuing with Leo's (very North-African-centric) account, these blacks were dominated by "King Joseph, the founder of Morocco", referring here to Yusuf b. Tashfin (ruled 1087-1106) who established Almoravid power in the Maghrib and Andalusia; and by "the five nations of Libya", meaning the Berber confederacies of the Zenaga, Guenziga, Terga, Lemta and Berdeua [14]. From their Berber rulers the blacks learnt "the Islamic law and divers necessary skills, and many of them became Muslims" [15]. Leo mentions the name Ghana only here, without elaboration, and he does not specify whether it was among the lands which were dominated by King Joseph and the Libyan nations. Neither does the name of the alleged conqueror of Ghana, the Almoravid amir Abu Bakr b. Umar, whom Yusuf b. Tashfin was to supersede in Morocco in early 1070s, appear anywhere in Leo's pages. There is thus nothing to suggest that Leo Africanus was in any sense a direct progenitor of the hypothesis of an Almoravid conquest of Ghana. However, Leo's account, consistently placing the blacks under Almoravid rule [16], was very likely influenced by the far-reaching, and quite unpersuasive, claims of 14th-century Moroccan historians, such as that the Almoravids had once conquered the Land of Blacks as far as the "mountains of gold" [17]. In this respect, Leo himself is more reliable as an indicator of the common negative attitude of Berbers and Arabs towards black Africans, than as a precise historical chronicler [18]. In any case, it is not so much with regard to an explicit conquest hypothesis that the seeds were being sown, but rather in setting a scene of whites over blacks, northerners over southerners, a scene within which the southward penetration of Islam does take place, and where such events as conquest might well be imagined. Following close on the heels of Leo Africanus - less than 25 years separate the two first editions - the Spanish scholar Luis del Mármol Carvajal published his Descripcion general de Affrica (also written very much from a North African point of view) in three parts, two in 1573 and one in 1599 [19]. Mármol relied mostly on Leo, but he knew also various Arabic sources, and he had spent many years in captivity in North Africa [20]. In the opening passage for the chapter dealing with the Land of Blacks in the third volume, Mármol reports: «In the description given by el-Mucaudi, Bubquer, Aben Gezar and other African geographers, only Guequin and Cano [Ghana] are mentioned in the Land of Blacks, because they did not have such detailed information about this area as one has today [21]. All regions which border on the Sahara, or which are close to it, are Mohammedan nowadays on account of the fact that the Almoravids, and the Lumtuna people, rule in Africa. About the year 380 AH, which is equal to 982 of Christ Our Saviour, among them came many morabitos and alcoranist preachers belonging to the cursed sect of Mahoma which they taught this barbarous people, and brought them over to their opinion. And after that, making his way through Ethiopia, amidst those black peoples, Hagin son of Abdulmalic, in the year 469 of the Hijra, began to teach them his rites and ceremonies, and another sectarian, called Yahaya son of Ali Benbucar, started converting all those who live on the bank of the River Niger, and close to it, who... [22]» The text continues with repeating Leo Africanus almost verbatim, including his remarks about the uncivilized state of blacks, and their domination by King Joseph of Morocco and the five Libyan nations. However, it is Mármol's own information which is more significant in relation to our topic than what he took from Leo. In the quoted passage, Mármol evidently connected the conversion of the blacks with the emergence of the Almoravids [23]. He said nothing further about either of the two apparently leading figures whom he mentions here, Hagin son of Abdulmalic and Yahaya son of Ali Benbucar, and it is difficult to identify them now. Furthermore, there is no reference to any military action against the blacks in the context of their conversion: if we take Mármol literally, King Joseph dominated the blacks only after they had already become Muslims. Some of this information, including the date for the conversion (469/1076-7) and the presence of Yahaya, comes most likely from the anonymous chronicle al-Hulal al-Mawshiyya, "The Embroidered Cloaks", completed in Granada in 1381, and perhaps available to Mármol either in North Africa or in Spain [24]. The same information we have quoted above is found also in the first volume, in the chapter where Mármol describes the religions of blacks. Both the date for their conversion to Islam and the context are the same. But in the first volume the agent is called "Yahaya son of Abubequer", his action is directed especially towards the peoples of "Geneúa" (Guinea) and "Neuba" (Nubia), and there is again no mention that Yahaya had used any force against these blacks, though his coreligionists are said to have conquered Spain and attacked Egypt - referring again to the historical Almoravids and their campaigns in the north [25]. Neither is any conquest in the Land of Blacks mentioned in yet a third possible passage, in the section where Mármol relates the history of Almoravids in particular [26]. Nor is the name of amir Abu Bakr b. Umar mentioned anywhere: according to Mármol, the predecessor of Yusuf b. Tashfin and the founder of the Almoravid movement was a certain "Abu Texifien" who ruled ca.1052-87 [27]. Nevertheless, there is in Mármol one reference to southern military campaigning by the Almoravids, and where he speaks of "Lumtunas", or the Lamtuna, having conquered "Tumbuto", or Timbuktu, and other parts of western Sudan in more a remote time. This is Mármol's own addition to Leo's description of Timbuktu, which he otherwise repeats as such [28]. The exact source for this rather general campaigning information is difficult to trace; similar commentary was available - as we have already noticed in the context of Leo Africanus - in the 14th-century Marinid historiography, which was certainly known to Moroccans (and probably to Mármol, too), and of course in the tale of King Joseph and the five Libyan nations. But it is the context of Mármol's remark, rather than its source [29], which is of special interest to us: for Mármol is describing here the ambition of "el Xerife Mahamet", or Muhammad al-Mahdi (reigned 1544-57), founder of the sharifian Sadid dynasty of Morocco, to extend his rule further into the Sahara and beyond, "as the Lamtuna had done in the past" [30]. Greater control over the Sudanese gold trade would have been important to the new rulers, giving them freer access to trade with Europeans other than the Portuguese, and thus enabling them to circumvent the understandable Portuguese reluctance to see gunpowder and firearms passing into the hands of their potential, and often actual, enemies [31]. Sultan Muhammad al-Mahdi eventually abandoned these plans, but the vital point for us is that, once again, the history of the Almoravids was being used to provide precedents for contemporary politics and philosophy, and being shaped to meet contemporary needs [32]. Just as the internal requirements of the Marinid dynasty in Morocco promoted the 14th-century embellishment of the sub-Saharan Almoravid records [33], or just as the external needs of an expanding Europe contributed to the 19th- and 20th-century evolution of the full-blown conquest hypothesis, so it seems possible, even likely, that the Marinid elaboration was being refurbished in the later 16th century, to meet the expansionist needs of Sadid Morocco, which eventually culminated in the Moroccan invasion in Songhay in 1591. Yet evidence quite contradictory to the Moroccan claims comes from the south of Sahara, from the Muslim scholars of Timbuktu. Sultan Muhammad al-Mahdi's attempt to extend his rule into the Sahara and beyond is mentioned also by Abd al-Rahman al-Sadi, the author of Tarikh al-Sudan, which is an invaluable source for Songhay history, and almost contemporary with Mármol [34]. But al-Sadi nowhere mentions that the Moroccan claims over the western Sudan were affirmed by any Almoravid precedent [35]. Perhaps such a justification was not used in this particular instance, or perhaps al-Sadi edited it out - we can never know - but his silence rather suggests that such a precedent did not exist at all [36]. However, there is another scholar, the celebrated Ahmad Baba, who denies without hesitation that any conquest of the Sudan by Muslims in the more remote past ever took place, in a document written in 1024/1615 and entitled Al-Kashf wa'l-bayan li-asnaf majlub al-sudan, "Enslaveable categories amongst the blacks revealed and explained" [37]. The context of this document is that the people in Tuwat, an important commercial terminus in the northern Sahara, evidently anxious about the morality of trading slaves who might have been unlawfully enslaved, wrote to Ahmad Baba for advice. One of the subsidiary questions put to him was this: had there been an early conquest of the blacks, whilst they were still in a state of unbelief? The implication of this question seems to have been that if conquest had preceded conversion, it would in effect impose a permanent enslaveable status, which could not by overriden by later conversion. But where had the Tuwatis picked up this story? Is it the same claim that Mármol had recorded barely more than a decade earlier - the dates at least do fit very enticingly? Or did the question arise from the actual Moroccan conquest which had just occurred and provided plenty of black slaves who might have been enslaved unlawfully, unless they had been conquered in the past, whilst pagans, thus legalizing their present enslavement? Again, we do not know exactly. In his answer to the Tuwatis, Ahmad Baba admitted that there were in his day black peoples who might legitimately be enslaved, because they had never adopted Islam [38]. But not because they had formerly been conquered: rather, he insists that those who became Muslims from among the Sudanese did so voluntarily "without being conquered by anybody" [39]. His emphatic rejection of any form of the "conquest hypothesis" carries particular weight, since he was writing relatively early, more or less on the spot geographically, and - a little curiously, we confess - he cites precisely the Almoravid/Ghana confrontation passage from Ibn Khaldun to prove his own anti-conquest opinion [40]. More than two centuries later, as we shall soon see, this notorious passage would lead European scholars to draw a quite opposite conclusion concerning the same matter. The slavery component in the argument cuts both ways. Perhaps Ahmad Baba, nearer the receiving end of the threat of enslavement, and living in close association with black Muslims, was for this reason likely to play down any idea of original conquest, and we should accordingly take his denial, however robust, with a grain of salt. Contrariwise, since many Saharans and North Africans were closely and profitably involved with the slave trade, and since an Almoravid conquest of Ghana (and other parts of western Sudan) would have provided a splendidly apt foundation for the subsequent cultivation of this trade, we might wish to discount reports of such a conquest for precisely this reason. The fact that the "conquest hypothesis" never clearly emerged in Africa, despite the attractive justification it could have offered for indulging in the slave trade, does lend support to contention that the conquest never occurred. Neverthless, these reservations expressed by the 17th-century scholars of Timbuktu remained - understandably - unknown to their European counterparts, among whom Leo Africanus especially continued to exercise profound influence. In fact, until the early 19th century, nearly all that was known in Europe of early West African history, before the Portuguese discoveries, was based on Leo Africanus, with some supplementary information drawn from Mármol, and the "Nubian Geographer", or al-Idrisi, whose Nuzhat al-mushtaq had appeared as an abridged Latin translation in 1619 [41]. In 1738, Francis Moore, for example, an English trader in the Gambia, compared the conquest of the blacks by the Libyan tribes, as described by Leo, to the Spanish conquest of the New World, a metaphor which could certainly have bred among his readers an image of widescale destruction and violence: «The Libyan tribe of Sanhagia ... coming amongst these harmless and naked people, made as rapid conquests as the Spaniards did in America, and contemned the Natives as much, looking upon them as brutes, because they were not Mahometans, nor instructed in avarice, nor ruled by tyrants; for which the Moors upbraided them as wanting the knowledge of religion, property and government.» [42] Leo's picture was still
repeated almost verbatim as an historical fact
by Mungo Park in 1799 [43]. Park's narrative of his first
expedition to the Niger rapidly became as popular as
Leo's decription in the previous centuries [44]. In this way, Park not only
carried over the old idea of Libyan domination of blacks
to the 19th-century European literature on Western
Africa, but he also gave it extra credibility with his
own authority and experiences. Invention of the Hypothesis by Cooley The next step in the evolution of the conquest hypothesis in European literature was taken in the availability of sources: within barely more than a decade, 1820 to 1831, three hitherto unknown sources, each a vital component for the hypothesis-to-be, came into the public domain of European learning. The first was a Latin translation of al-Maqrizi's fragment on the races of the Sudan, including this passage: «Etiam Rex al-Ganae Principum (harum regiorum) quondam fuerat maximus; deinde, ingruentibus Al Molatstsemiis, omnis eorum potestas evanuit; occupavit que familia Sousou, tum gens Mali quae interim invaluerat, hos tractus, ibique regnavit.» «The king of the Ghana was the greatest of kings but then the Veiled Men overcame them and their authority dwindled away. The people of Susu [then] conquered them. After this the people of Mali became powerful and ruled over them.» [45] The second was a Portuguese translation of Ibn Abi Zar's Rawd al-Qirtas [46]. Ibn Abi Zar nowhere mentions Ghana, but he presents a marvellously vivid canvas, for the southern Sahara, of white triumphalism over blacks, and of swashbuckling Islamic adventurers, including the following information about the death of Almoravid amir Abu Bakr b. Umar: «Concluida esta exhortação, despediu-se Abu-Bacar, e marchou para Sahara, na qual se conservou fazendo a guerra aos cafres da Ethiopia, aonde foi martyrizado em huma das suas gazuas, traspassado de huma seta hervada, de que morreu no mez de Xaaban do anno 480 (1087), depois de haver subjucato ao seu Imperio o paiz de Sahara at‚ Iabaledahb (monte de ouro), paiz da Ethiopia, cujos estados revertao depois para Iussof, filho de Taxefin.» [47] «Then he [Abu Bakr] bade him [Yusuf] farewell and left for the desert. There he remained for some time waging Holy War on the unbelievers from among the Sudan until on one of his expeditions he was struck by a poisoned arrow and met martyrdom. This was in the month of Shaban 480/November 1087 after all the land of the desert as far as Jabal al-Dhahab (Mountains of Gold) in the land of the Sudan had come under his sway. After his death authority became vested exclusively in Yusuf b. Tashfin.» And finally, in 1831, the following note appeared to Etienne Quatremère's French translation of al-Bakri's Kitab al-masalik wa'l-mamalik: «Si l'on en croit un historien de Maroc (manuscrit 825, p.5), "les habitants du pays des Noirs qui ont Gânah pour capitale, professèrant la religion chrétienne jusqu'à l'année 469 de l'hégire. A cette époque, ils embraisèrent l'islamisme."» [48] «If we can believe a Moroccan historian (manuscript 825, p.5), "the inhabitants of the Land of the Blacks whose capital is Gânah professed Christianity until the year 469 of the Hijrah. At that time, they embraced Islam."» The document cited is again the 14th-century al-Hulal al-mawshiyya, which slightly misquotes here al-Zuhri in the 12th century [49]. But the meaning - real or alleged - of these three sources was not yet understood, for the early history of West Africa was still undiscovered in the late 1830s. This changed significally in 1841, when William Desborough Cooley, in his pioneering work, The Negroland of the Arabs examined and explained; or, An Enquiry into the early history and geography of Central Africa, opened up scientific study of the area, based on Arabic sources, and (as far as our special interest here is concerned, though it is not the main element in his argument) began to piece together, albeit tentatively, the conquest evidence [50]. All the three works mentioned above, and many others, are cited by Cooley, although it was profoundly important that he, himself an Arabist, was able to read al-Bakri, Ibn Khaldun and other yet unpublished sources in Arabic manuscript, instead of translations [51]. But exactly which of these sources, and each in exactly which way, contributed to the development of the conquest hypothesis is sometimes difficult to decipher. For Cooley's key summary statement on the "conquest" is: «Ghánah merged in the empire of the Morabites, an event which may be assigned, with much probability, to the year of the Hijra 469, when the Mohammedan faith was forcibly imposed on the pagan nations of Negroland contiguous to the Western Desert.» [52] Three sources are cited. They are, in successive order, al-Hulal al-Mawshiyya (this is the note, quoted just above, to Quatremère's al-Bakri), the passage of Mármol we have already quoted above (n.22) and the Portuguese translation of Ibn Abi Zar ("Moura, Hist. do Soberanos, &c, p.146"). The date, 469 AH, for the conversion is clearly taken from al-Hulal and Mármol, as well as the link to the Almoravids, whereas the use of force and "conquest" seems to come from Ibn Abi Zar, and from Ibn Khaldun, whom Cooley quotes at length as his principal source for the history of Mali [53]. We shall return to Ibn Khaldun in a moment. But first, we must look at a rather strange witness for the "conquest" - al-Bakri, clearly the outstanding early source for Ghana, and very justly the major authority for Cooley. For our own special topic here, however, al-Bakri is a strange witness indeed, for he nowhere mentions the "conquest" of Ghana by the Almoravids in 469/1076-77, nor could he, since he was writing in about 460/1067-68, nearly a decade earlier. Nevertheless, Cooley apparently believed that there is an account of the conquest of Ghana, after all, in al-Bakri. The key is Awdaghust. An important trading terminal on the southern side of the Sahara, Awdaghust is elaborately described by al-Bakri. Inhabited by Arabs and Berbers, it was clearly in al-Bakri's time a white town, with some blacks among its slaves [54]. But Awdaghust had recognized the authority of the black ruler of Ghana, and this, according to al-Bakri, was the reason why the Almoravids conquered the town in 446/1054-55, treating its inhabitants with considerable severity [55]. Awdaghust had for centuries been a Berber city [56]; this is confirmed by reports from the 9th and 10th centuries. Yet al-Bakri says that this town "used to be the residence of the king of the Sudan who was called Ghana before the Arabs entered Ghana" [57]. Al-Bakri is our only witness for this royal black residency in Awdaghust. The Corpus editors, rightly observing that al-Bakri is here "far from being clear", and presumably calculating how far back one would have to go in order to pre-date the first mention of Awdaghust as the residence of a Berber ruler (this is al-Yaqubi, in the 9th century) [58], suggest that the first Arab entry into Ghana may have been at the time of the Arab conquest of North Africa, in the 8th and 9th centuries [59]. Perhaps, that long ago, Awdaghust was the capital of Ghana. But Cooley, repeatedly and clearly, implies that the Almoravid conquest of Awdaghust was in fact the conquest of the capital of Ghana [60]. This must have been simply carelessness on Cooley's part, for al-Bakri gives us a date for the sack of Awdaghust (1054-55), and he tells us more than once when he himself was writing (1067-68), and both these dates are well before the date traditionally proposed for the fall of Ghana in the "conquest hypothesis", which is the same date assigned by Cooley to that alleged event (1076-77). Not only are the dates askew: the entry of the Arabs into Ghana is most unlikely to refer the Almoravid conquest of Awdaghust, as the Almoravids are nowhere, to the best of our knowledge, referred to as Arabs, in spite of the belief in their Himyaritic origin [61]. This, the first misunderstanding, is Cooley's own: if his readers followed him in it, they have at least some excuse. The second flaw concerns the interpretation of Ibn Khaldun's notoriously brief and vague mention of these events, written centuries afterwards - and here the blame may rest more with Cooley's readers than with him. A modern English translation of the relevant passage is this: «Later the authority of the people of Ghana waned and their prestige declined as that of the veiled people, their neighbours on the north next to the land of Berbers, grew (as we have related). These extended their domination over the Sudan, and pillaged, imposed tribute (itawat) and poll-tax (jizya), and converted many of them to Islam. Then the authority of the rulers of Ghana dwindled away and they were overcome by the Susu, a neighbouring people of the Sudan, who subjucated and absorbed them.» [62] Cooley's own rendering of the same passage keeps the attention of the veiled people fixed more specifically on Ghana: «The people of Ghánah declined in course of time, being overwhelmed or absorbed by the Molaththemún (or muffled people - that is, the Morabites), who, adjoining them on the north towards the Berber country, attacked them, and, taking possession of their territory, compelled them to embrace the Mohammedan religion. The people of Ghánah, being invaded at a later period by the Súsú, a nation of Blacks in their neighbourhood, were exterminated, or mixed with other Black nations.» [63] It is evidently this passage which prompted Cooley to piece together the fragmentary evidence from al-Hulal, Mármol and Ibn Abi Zar into a solid conquest hypothesis, since it offered the wider historical context, into which such information as the conversion of Ghanaians to Islam in 469/1076-77 and the large-scale conquests of Abu Bakr b. Umar in the Land of Blacks, "as far as the mountains of gold", could easily be merged [64]. Yet the problem here is that the rulers of Ghana seem to be subject to a double whammy: first the Almoravids assail them, yet they seem to survive for a further century and more, before finally succumbing to the Susu, customarily dated soon after 1200 AD [65]. Probably influenced by Ibn Khaldun, Cooley frequently applies dramatic terms to the final demise of Ghana: "...what were the revolutions, it may be asked, which caused Ghánah to disappear?" (p.43) - "the extinction of Ghánah" (p.61), and so on. Such doom-laden remarks, however, often refer more to the Susu conquest, than the Almoravid conquest over a century earlier. Cooley himself clearly distinguishes between these two events - the religious conquest by the Almoravids and the more political by the Susu - in his summary table [66]: GHANAH Compelled by the
Morabites to relinquish Idolatry
and embrace the Mohammedan faith 469 / 1076 But the distinction is not always quite clear in his pages (see for example p.138), and it is not surprising if some of his readers have misapplied remarks about the Susu to the earlier Almoravid period. Two contemporary German
texts show how scholars, looking at the available
original sources, but uninfluenced by Ibn Khaldun and
Cooley's Negroland, could write about the events
of the eleventh century Western Africa without mentioning
either any Almoravid/Ghana confrontation, or any enforced
conversion. J. E. Wappäus, in 1842,
mentions the Almoravids and their conquests in the north,
but he says nothing of their relationship with Ghana,
save the conquest of Awdaghust, which he does not regard
as the capital of Ghana [67]. Friedrich Kunstmann,
in 1853, citing al-Hulal as quoted in Quatremère's
note to al-Bakri (see above), says merely that the
conversion of Ghana, from Christianity, occurred in 1075;
he makes no mention of forced conversion, or conquest, of
Ghana by the Almoravids [68]. Consolidation of the Hypothesis: from de Slane to Delafosse The following step in the textual genealogy of the conquest hypothesis was important in terms of availability, rather than of the introduction of new source material. Cooley had used Ibn Khaldun in Arabic manuscript [69], but a partial Arabic edition of the Kitab al-Ibar, Ibn Khaldun's universal history, was prepared by Baron William MacGuckin de Slane, a well-known French Arabist, and published in two volumes in Algiers, in 1263-7/1847-51. A complete Arabic edition of the Kitab al-Ibar appeared in 1284/1867, at Bulaq in Cairo, in seven volumes, but de Slane had already made all the material relevant for Almoravid/Ghana affairs easily available to western scholars, and in a printed text much more clearly legible, and in many respects more reliable, than the Bulaq edition. In 1852-6, immediately after the completion of his Arabic volumes, de Slane published a French translation of these, Histoire des berbères et des dynasties musulmanes de l'Afrique septentrionale (4 volumes, Paris) [70]. De Slane's rendering of the celebrated passage concerning the veiled people and Ghana corresponds in general with Cooley's, with one significant exception: de Slane says that this people "étendit sa domination sur les Noirs, dévasta leur territoire et pilla leurs propriétes" ("extended their domination over the blacks, devastated their territory and pillaged their property") [71]. De Slane's Arabic, however, has the veiled people pillaging the property and the territory of the blacks. We do not know what made de Slane add devastation in his translation, since it does not exist in his Arabic version, or in the Bulaq edition [72]. Nevertheless, given the prestige and popularity of de Slane's French edition, his "devastation" became an important contributory factor to the development of the "conquest hypothesis" in West African historiography. After Cooley and de Slane, the task of unveiling early West African history was continued by Heinrich Barth [73], the famous German explorer whose magisterial Travels and Discoveries in North and Central Africa ... in the years 1849-1855 appeared in three editions in 1857-59 [74]. Barth's work deals mainly with his own adventures and observations, but there is also a large chronological table of the history of western Sudan, reaching from 300 AD to Barth's own time. Concerning the history of Ghana Barth wrote: «A.D. 1076. A.H.469. Ghanáta conquered by the Senhája, and great part of the inhabitants, as well as the neighbouring districts of Negroland, compelled by the Mórabetin to embrace the Mohammedan faith. A'bú Bakr ben Omar takes up his residence in this part of Negroland.» [75] Barth's interpretation of the Almoravid/Ghana relationship hardly carries the substance of Cooley's original argument much further, but does change the mood. While Cooley seems comfortable enough with a violent Almoravid assault upon Ghana, when transferring al-Bakri's details about the sack of Awdaghust over the capital of Ghana, yet at other times, when discussing specifically the conversion of Ghana in 469/1076-77 to Islam, he seems to hesitate, perhaps worried (consciously or unconsciously) by the fact that both al-Hulal and Mármol, his sole sources for the conversion, do not mention any force at all. In fact, Cooley remarks that "[t]he kingdom of Ghánah remained little changed by the loss of its independence", and he nowhere claims that the Almoravids displaced the native rulers of Ghana [76]. Barth, however, tightens all this up, with a clear affirmation of simultaneous conquest and conversion, though the heightened devastation introduced by de Slane is still missing. This is perhaps because Barth, like Cooley, was able to read Ibn Khaldun's Kitab al-Ibar also in Arabic [77]. There, as noted, the destructive element is muted, and not precisely targetted on Ghana in particular, even though Barth's many references to de Slane's translation indicate that this too was well known to the German traveller [78]. But how widely noticed was Barth's comment? On the one hand, it is very short and tucked away among the numerous meteorological and ethnographical appendices which are attached at the end of each of the five volumes; but on the other, the three complete versions published simultaneously in the period 1857-59 in Britain, Germany and United States made his material unusually widely available [79]. Also, his worldwide reputation as an outstanding explorer meant that he enjoyed considerable authority among later western Africanists [80]. The next step was taken also in the better availability of sources. Ibn Abi Zar had been one of the first relevant sources (for the conquest hypothesis) to be translated into European languages: first into Portuguese, and later into Latin, in 1843-46, by C. J. Tornberg, who published his scholarly edition with the Arabic text, just too late to be used by Cooley [81]. The crucial turning point, however, for Ibn Abi Zar as a high-profile source, amongst European Africanists, came in 1860, with the appearance of the French translation, Roudh el-kartas: histoire des souverains du Maghreb, by A. Beaumier, in Paris. The earlier editions had seemingly been quite inaccessible to most scholars, but Beaumier's translation soon spread widely, and it began to be used as the leading source for Almoravid history alongside de Slane's French edition of Ibn Khaldun. Ernest Mercier, for example, faithfully repeated the descriptions of Ibn Abi Zar and Ibn Khaldun, together with additional material derived from al-Bakri and Ibn al-Athir, in his articles on the Almoravids and Almohads, published in Revue Africaine, in 1868-69 [82]. Or, almost faithfully. For Mercier repeats - without reference to Cooley - the implication that Awdaghust, on the eve of the Almoravid conquest, was still the capital of Ghana [83]. Mercier also intensifies the violence: for Cooley and Barth, the Ghanaian converts were not the survivors of a massacre. Mercier, manifestly building, and embroidering, upon al-Bakri [84]- and evidently influenced here by de Slane's translation of Ibn Khaldun - described how after the conquest of Sijilmasa, «the Almoravids, who seemed to have interrupted their conquests in the rich areas of the north, returned to the south and went to attack the land of the Negroes, which was called in those times by the name R'ana [85]. They conquered this country easily, and entered victoriously into Aoudar'ast, a flourishing city and capital of the country and a residence of the ruler who was also called R'ana by the inhabitants. The victors destroyed and pillaged the peaceful population, and these acts of violence were legalized, in their eyes, by the idolatry of the vanquished; they massacred the men, violated the women, and took a considerable booty which they carried into their country.» [86] At first Mercier's highlighting of violent destruction was not widely adopted. Most works dealing with Almoravid history merely repeated the superficial descriptions of Ibn Abi Zar and Ibn Khaldun concerning the withdrawal of amir Abu Bakr to the desert after he had handed the leadership of the movement in the north to his cousin Yusuf b. Tashfin. Abu Bakr is only said to have conquered large parts of the Land of Blacks, without any specific mention of Ghana [87]. Even Mercier himself, in his later work, Histoire de l'Afrique Septentrionale, published in Paris in 1888, omits the name Ghana, referring only to a part of "Nigritie" having Awdaghust as its capital, although there is still a hint of the destruction: «[After the conquest of Sijilmasa] the Almoravids set off to destroy the lands of the extreme south where lived rich and harmless Negro peoples; the religion gave the pretext for their violence. They brought under their rule that part of Nigritie, the capital of which was a great city called Aoudaghast, and settled the Berber race on the upper Senegal.» [88] Also in 1888, René Basset, a well-known French orientalist, published a brief article on the history of the Sudanese empires [89]. This article did not actually contain much new information, but Basset was highly influential in the contemporary development of broader contextual ideas, for example concerning relations between Muslim white pastoralists and animist black sedentaries. Basset, who took the idea of conquest directly from Cooley, and from the French translations of Ibn Abi Zar and Ibn Khaldun, did not actually mention the year 1076 - except in a footnote [90] - but through his authority the concept of forced conversion (and conquest) of Ghana became rooted in the French literature [91]. The conquest was repeated more clearly, with the year 1076, by Louis Binger in 1892. Binger cited, for some unknown reason, al-Bakri as the principal source for the conquest: "El-Békri added that after this conquest [of Ghana] a great part of the population was forced to embrace Islam by the Merabétin, as well as many of the neighbouring areas." [92] In fact, Binger's work is full of similarly mysterious claims: he says, for example, that al-Bakri wrote in either 1153 or 1203-4 - or even 1607! [93] All this suggests that he had hardly seen all his Arabic sources himself, even as translations, deriving his ideas mostly from previous writers, chiefly from Basset but also from Cooley and Barth [94]. Binger had, however, great influence because his book was widely read in France. Another early French propagator for the conquest hypothesis was Louis Tautain, who was also the first writer to identify the Wagadu of oral tradition and the Ghana of Arabic sources, in his article published in 1895 [95]. Yet a third influential analyst, beside Basset and Binger, was Alfred Le Chatelier, whose L'Islam dans l'Afrique occidentale was published in Paris in 1899, being a revised edition of a text written already in 1888. Le Chatelier had an unusually broad perspective of the extent of Almoravid conquests: Abu Bakr b. Umar, continuing the task undertaken earlier by the Berber race, reconquered the Songhaï of Ghana [96], and at his death, in 480/1087, all the countries bordering the Senegal and the Niger to the north were to be found under the yoke of the Lamtuna. In addition to Ghana, they occupied the states of Djenné, Zanfra, Zegzeg, Ouangara, all of Nigritie as far as Gogo, where the Songhaï succeeded in remaining independent [97]. Le Chatelier does not, however, mention the conversion of Ghana in 469/1076-77, neither is his conquest explicitly destructive in the same sense as Mercier's, in his 1868-69 articles. Even in the early 20th century, Mercier's fire and sword have not completely caught on. In 1905, Flora B. Shaw (Lady Lugard), published her A tropical dependency: an outline of the ancient history of the western Soudan with an account of the modern settlement of Northern Nigeria, a remarkable achievement at that time, and still readable today though gravely dated. She is certainly an adherent of the conquest hypothesis, but in her view this is evidently not a destructive affair, nor does it seem explicitly linked with the conversion of Ghana. "In 1076," she wrote, «Abu Bakr carried the vengeance of Audoghast to the gates of Ghana, and, overthrowing the reigning black dynasty, placed a Berber on the throne. The life of the country does not seem to have been profoundly affected at the time by this revolution. El Idrisi, writing nearly a hundred year later, still speaks of it as being the greatest kingdom of the blacks. He mentions the fact that it is ruled by a king of Berber descent, who "governs by his own authority, but gives allegiance to the Abbasside Sultan of Egypt", and that the king and people are now Mohammedans; but he does not speak of it as having become in any respect a Berber kingdom.» [98] All these and other sources suggest that the articles of Mercier in 1868-69 were not widely noticed, and the idea of the Almoravid conquest of Ghana was, instead, still derived from Cooley and Barth, with an additional flavour of violence from de Slane [99]. The crucial turning point, the full flowering of the conquest hypothesis as a throughly destructive event, came finally in 1912, when Maurice Delafosse published his massive, three-volume Haut-Sénégal-Niger [100] Delafosse, who was a great compiler rather than an original researcher, finally cemented the history of ancient Ghana: his chronology became (and to a considerable extent still remains) accepted, as well as his idea of the Sudanese empires as nation-states, succeeding each other in hegemony, with fixed capitals and boundaries [101]. And his circumstantial account of the Almoravid conquest acquired almost canonical status. "The conquest," he writes, «seems to have been complete: not only did the Almoravids take the city, pillage the goods of the inhabitants, massacre a part of the Soninke population, forcing the rest either to flee or to embrace the Muslim religion, but they obliged the emperor to recognise the suzerainty of Abu Bakr and to pay him tribute, and they annexed to their political domain all the dependencies of Ghana, as far as and including the auriferous mountains of Bambuk.» [102] What was his evidence for this claim? This passage has no footnote, and thus it is difficult to trace exactly Delafosse's sources, but the frames for this picture were certainly, again, the all too familiar accounts of Ibn Abi Zar and Ibn Khaldun. And, according to his bibliography, Delafosse knew all the previous European works where the conquest hypothesis appears: Cooley, Barth, Basset, Mercier of 1888 (and probably his articles of 1868-69, too), Binger, and Tautain. This picture of Delafosse ruled virtually unchallenged in West African historiography for half a century. In 1962, J. Spencer Trimingham, a prolific and encyclopaedic scholar of Islam in black Africa, who did a great deal to lay the foundations of serious western study in this field, offered the following description: «The energies of the Murabit movement were directed primarily towards the north and it was only when Abu Bakr b. cUmar returned to the southern desert in 1062 that, in order to absorb the restless tribesmen, he directed their energies against the Negro kingdom. Even so a long struggle ensued before the Soninke capital was finally captured in 469/1076-7. Many of its people were massacred and those who survived were forced to join Islam. We do not know [103] whether the king was left on the throne, but it seems clear that his widespread empire broke up and its constituent chieftaincies gained their independence... The fortunes of the Murabitun quickly declined through tribal rivalries, Abu Bakr was killed in 1087, and the whole organization of the desert tribes collapsed.» [104] Two books published at almost exactly the same time as Haut-Sénégal-Niger help to illustrate this historiographical progress. The first appeared in 1911, Chroniques de la Mauritanie Sénégalaise: Nacer Eddine, by Ismaël Hamet, himself a western Saharan Muslim. His volume, devoted to the Islamic history of the area, includes considerable historical material in Arabic, and Mauritanian oral tradition concerning Abu Bakr b. Umar [105]. He does not mention, however, any explicit Almoravid conquest of Ghana in 1076: nor, if we are right in arguing that the conquest hypothesis was an European invention, is there any reason why he should have. Hamet merely says that the Almoravid empire reached from the Senegal river to Andalusia [106], that Abu Bakr fought against the blacks [107], and that the Almoravids took Awdaghust "and brought the kings of Ghanata under their law. Some years afterwards, following their success, the Lemtouna took by force the country and spread Islam into Nigritie, beyond the great rivers" [108]. And in 1913, the second edition of T. W. Arnold's The preaching of Islam: a history of the propagation of the Muslim faith, appeared in London. The first edition, written in 1896, had in this connection briefly drawn upon Leo Africanus: «The reign of Yusuf Ibn Tashfin, the founder of Morocco (1062 A.D.) and the second Amir of the Almoravide dynasty, was very fruitful in conversions and many Negroes under his rule came to know of the doctrines of Muhammad. Two Berber tribes, the Lamtuna and the Jodala, whose habitat bordered on and partly extended into the Sudan, especially distinguished themselves by their religious zeal in the work of conversion.» [109] The second edition preserves this sentence, but adds a new one following: «In 1076 the Berbers who had been spreading Islam in the kingdom of Ghana for some time, drove out the reigning dynasty, which was probably Fulbe, and this ancient kingdom became throughout Muhammadan; at the beginning of the thirteenth century it lost its independence and was conquered by Mandingoes.» [110] So much for the precise
historiographical genealogy, which may be briefly
summarized thus: in 1841, Cooley, piecing together
fragmentary evidence, enlivened by the misapplication of
al-Bakri's Awdaghust to the Ghanaian capital, presents
rather tentatively a conquest hypothesis; with Barth and
others, this becomes more firmly accepted; Mercier, in
1868-69, encouraged by the translation of de Slane (and
independently repeating Cooley's Awdaghust/Ghana
confusion), adds a destructive element to the conquest
hypothesis; for some years the destructive element fails
to catch on, but in 1912 Delafosse wins almost universal
acceptance for it, which reigns undisputed for fifty
years. This is our main argument, but there is also
another, more contextual and hidden, strand in the
development of the conquest hypothesis, and to this we
must now turn, though very much more briefly. Impact of the Ideological Environment and Colonialism Earlier, in the context of Francis Moore and Mungo Park, we suggested that before the history of the Sudanese empires was properly discovered, there existed among early European Africanists a widespread - although often vague - idea of Muslim conquest in the West African interior, inspired by Leo Africanus. However, this idea received more importance at the end of the 19th century as Europeans' attitudes towards Africa and Africans changed. The age of colonialism introduced many new elements to the historiography of Africa, originating not from the actual sources, but from various presuppositions, some of them racist, which so long distorted many aspects of the European encounter with other cultures. One of the first signs of these ideological changes (in relation to our topic) is the worsening image of the Almoravids in European historiography since the mid-19th century. A reason for this deterioration was that the Almoravids began to be widely identified with Tuaregs [111]. This seemed plausible inasmuch as in the Arabic sources the Almoravids are often called the "veiled ones", al-mulaththamun [112], and the use of the tagulmust-veil is an important part of Tuareg culture [113]. Consequently the Almoravids inherited all the vices of the Saharan nomads who had an utterly negative image in European eyes. Already Mungo Park had reported that the "Moors" - referring here the Mauritanian nomads in general - were "at once the most vainest and proudest, and perhaps the most bigotted, ferocious, and intolerant, of all the nations on earth; combining in their character the blind superstition of the Negro, with the savage cruelty and treachery of the Arab" [114]. Furthermore, the murders of European travellers in the Sahara - Daniel Haughton in 1791, Alexander Gordon Laing in 1826, Alexine Tinne in 1869, and the massacre of the Flatters expedition in 1881 - all certainly reinforced the opinion that the Saharan nomads considered it to be "as lawful to murder a European as it would be to kill a dog" [115]. There were some travellers, like Henri Duveyrier, who tried to dispel the European mistrust of the Tuaregs, but these attempts were fruitless [116]. On the other hand, the Tuaregs were also respected as brave and dangerous warriors, capable of causing considerable losses to the French army in the Sahara [117]. Transferred into historiography, this conception meant that the Almoravids of the past appeared quite as intolerant and aggressive as the Tuaregs of the present. In the works of Cooley and Barth this idea is not yet so visible, but it was manifested especially by Reinhard Dozy in his classic work, Histoire des musulmans d'Espagne, published in 1861. According to Dozy, the Almoravid period had meant a severe regression for the prosperous culture of Andalusia, and many scholars followed him in this dismissive judgement [118]. In 1917, for example, the Finnish Orientalist Knut Tallqvist described Almoravid rule in Andalusia thus: «The Berber rule was a serious setback for the Arab culture of Islamic Spain. Compared to the sophisticated Arabs of Spain, the African Berbers were half-barbarians, and though they had adopted Islam, they felt no sympathy towards their Arab coreligionists. It was, therefore, natural that the Arabo-Spanish civilization was not favoured and supported by the Almoravids.» [119] Taking this stance, it was more than reasonable to suppose that the Almoravid influence had been equally disastrous on the southern edge of the desert, too, just as de Slane's translation of Kitab al-Ibar suggested. On the other hand, the position of Saharan nomads was elevated by the strengthening idea of racial hierarchy in European thought. After the elaboration of the Hamitic hypothesis at the turn of the century, the Saharan nomads were counted among those advanced African peoples who had drops of civilizing "white" blood in their veins, being thus superior to the blacks [120]. According to Le Chatelier, for example, the blacks of the Western Sudan owed their progress to contacts with the neighbouring Berber race, "more elevated in the normal order of development of humanity" [121]. The Almoravid conquest of Ghana fitted well into this picture, according to which all cultural progress in sub-Saharan Africa had taken place after waves of conquest and cultural influence from the north: by Egyptians, Phoenicians, Greeks, Romans, Berbers, Arabs, and finally by Europeans [122]. It is hardly sheer coincidence that the more destructive and complete the Almoravid conquest of Ghana becomes in western literature, the more fantastic the ideas the same writers have on the origins of Ghana. According to Delafosse, for example, Ghana was founded by Jews who had escaped from Roman revenge in Cyrenaica after their unsuccesful revolt in 116 AD [123]. Another equally widespread assumption (as the brutality and bellicosity of the Almoravids) among the early 19th-century writers was that actual power in the Sahel rests primarily with the nomads [124], as Mungo Park had affirmed in 1799: «The enterprizing boldness of the Moors, their knowledge of the country, and, above all, the superior fleetness of their horses, make them such formidable enemies, that the petty Negro states which border upon the Desert are in continual terror while the Moorish tribes are in the vicinity and are too much awed to think of resistance.» [125] René Caillié's description of the situation around Timbuktu about thirty years later was no better: «The Tooariks have terrified the Negroes of their neighbourhood into subjection, and they inflict upon them the most cruel depredations and exactions. Like the Arabs, they have fine horses which facilitate their marauding expeditions. The people exposed to their attacks stand in such awe of them, that the appearance of three or four Tooariks is sufficient to strike terror into five or six villages.» [126] Besides the European travel literature [127], positive evidence for nomadic superiority was gained from the Arabic sources which stress the warlike nature of the nomadic peoples [128]. Evidence of the opposite imbalance, with nomads dependent upon, or subordinate to, their settled neighbours, was not emphasized by early western authors - but it was certainly not absent, and surfaced from time to time [129]. Nevertheless, the later writers went even further in stressing this confrontation. Ren‚ Basset, for example, saw the entire history of Western Sudan as a continuous struggle between Muslim "white" pastoralists and pagan "black" sedendaries [130]. This view was shared by other French writers [131]. Historiographically this idea meant that the Sudanese empires had been equally helpless against the Muslim invaders from the north, as the Almoravid conquest of Ghana in 1076 (in imagination), and the Moroccan attack on Songhay in 1591 (in reality), proved. Furthermore, emphasizing the historical and cultural passivity of the black Africans gave Europeans a moral justification for their own colonial conquests [132]. The third element coming from outside the sources was the image of Islam in European thought. In this respect, western attitudes were again very ambivalent. On the one hand, Islam, and especially what appeared as Islamic fanaticism, was seen as a barbarous and backward force preventing all modern development - often tantamount to Europeanisation - in Asia and Africa [133]. Such a view seemed demonstrated by experiences in European colonies, like the Indian mutiny (as the British called it) in 1857-58 which was attributed to Muslim conspiracy, the revolts in Algeria, and the strengthening in the Sahara of the Sanusiyya brotherhood which was believed to be extremely hostile to westerners [134]. In Western Africa, Islamic fanaticism was associated with the contemporary jihad-movements which the French believed were opposed especially to their advance in Senegambia and Upper Niger [135]. From this point of view, the conquest (and destruction) of Ghana by cruel Almoravids could be seen as an historical example of the negative impact of Islamic fanaticism, justifying European "protection" in sub-Saharan Africa [136]. And the French truly believed that part of their mission in Western Africa was to save the local black population from Muslim slavery and oppression: in Upper Niger they deliberately sought the assistance of the animist Bambara in their fight against Ahmadu of Segu and Samory [137]. We should also notice that historiography had an important role in late 19th-century France in raising the national spirit, and the enemies of France, whether they were fanatic toucouleurs or bloodthirsty boches, could not expect to receive much sympathy from the pens of French writers [138]. On the other hand, Islam was seen also as a positive force in African history. In "darkest" Africa Islam was believed to work as a kind of kindergarten training blacks to receive the blessings of European civilisation [139]. This idea, which strengthened especially in the early 20th century, after the active colonial conquest was over in Western Africa, is logical, for Islam is, after all, more understandable in European eyes than African animism [140]. Thus conversion to Islam meant a step up on the ladder of progress: from oralcy to literacy, from superstition to monotheism. In this context, the Almoravid conquest of Ghana could be seen as a civilizing act, painful but necessary, giving birth to the sophisticated Islamic civilization in Western Sudan which was admired by men like Delafosse and many other French colonial administrators [141]. But how do all these divers and sometimes
contradictory points of view contribute towards the
development of the conquest hypothesis? The answer is
that they set the context in which the early European
Africanists worked. We should remember that all
historians (including ourselves) are bound to their own
society and its values which guide their work, sometimes
consciously and sometimes without their noticing [142]. In order to recognize these
values, we need to deconstruct the textual and
ideological environment in which the earlier writers
worked, regardless of whether they were 14th-century (or
16th-century) Arabic writers or 19th-century Europeans.
One reason for the invention, and popularity, of the
conquest hypothesis is certainly that the ideological
atmosphere in early 20th-century Europe favoured such
explanations in African history. On the other hand, the
gradual and hidden infiltration of the various
ideological elements - the identification of Almoravids
with Tuaregs, the belief in nomadic superiority, racism,
and the ambivalent image of Islam - into the body of
research was possible because of the methodological
decline of African historiography during the colonial
period. The widely accepted conception that Africa had no
history of its own did not mean that interest in the
African past completely ceased; it meant that African
historiography was dominated, not by academic scholars (such
as Cooley and Barth had been), but by ethnographers, self-taught
amateurs, and colonial officers who often worked without
any training for proper historical research and were thus
more open to the ideological tendencies of their own age,
and whose purpose was, after all, to serve the interests
of colonial rule [143]. Criticism and Reconsideration In the foregoing discussion, we have endeavoured to probe into the origins of some of the strands which would be gradually woven into the conquest hypothesis. Some of those origins were tainted by presupposition and prejudice. But that is not at all to imply that those who later accepted, and passed on, the conquest hypothesis were aware of such suspect, half-obscured, origins. Far less still, that they shared in the ideological orientation of some of those origins. For those who have remained loyal to the hypothesis, even for those who helped in the protracted business of building it up, it was credible, well-recommended historiography, not at all a mythology embodying a subtle, subversive hidden agenda. Only later, and initially hesitantly, did reservations emerge. The first modern scholar who began to express any doubts about the historicity of the conquest hypothesis was Paul Semonin, in an article published in 1964 [144]. It is again hardly sheer coincidence that criticism of the conquest hypothesis started only in the early sixties. The primary stimulus for Semonin was indeed the finding of a new source for the Almoravid history, Ibn Idhari [145], which required scholars to reconsider some of their earlier conceptions concerning the chronology of Almoravid history, for example the founding of Marrakesh [146]. But the early 1960s meant also great ideological changes in African historiography in general, as decolonisation forced Europeans to change their attitudes towards Africans. The old undervaluation was no longer acceptable, nor was the idea that all progress and change in the African past had always been caused by northern invaders [147]. And, finally, we should not forget contemporary methodological development, as African historiography became a respected academic discipline not only in the former metropolitan countries, but also in the former colonies, which were eager to create for themselves an historical identity. Semonin did not go so far in his criticism as to deny that any conquest of Ghana had ever occurred, but he did noticeably reduce its importance. Noting that "there has been a lack of critical attention to the problems presented by the source material", he concluded: «The Sack of Ghana is so poorly documented that it is questionable what precedents it may have set, or what real impact it may have had, beyond that of a calamitous but momentary disturbance in the Sahil with limited resources for a continuing influence upon the Islamization of the area.» [148] The next was Paulo Fernando de Moraes Farias, who in 1967 spoke of the "supposed destruction" of the power and capital of Ghana by Abu Bakr b. Umar, adding that "it is well known that the date of the Almoravid 'victory' over Ghâna is poorly documented" [149]. Some years later, in 1972, a further two major scholars expressed their doubts. Jean Devisse noted "the futility of pursuing the question of Ghana's conquest given the doubtful historicity of the evidence" [150]. The other was Harry T. Norris, who expressed his misgivings by referring to Abu Bakr's disappearance from history (apart from his very briefly reported death) after his retirement to the desert [151]. All these comments, on the inadequate documentation of the alleged conquest, indicate that European scholars had gradually began to notice the methodological weaknesses of the conquest hypothesis. Yet they did not dare to challenge the entire tradition; monuments like Haut-Sénégal-Niger do not lose their authority overnight. In 1974, in an article far too little known, Mahamadou Coulibaly was perhaps the first scholar to attempt an inclusive study of the conquest hypothesis [152]. He began by emphasizing, quite rightly, that al-Zuhri - the source for the conversion of Ghana in 1076 - makes no reference to any military victory. There is no evidence for the destruction of Ghana, not even of its dismemberment, nor any substantive political or military domination of Ghana by Muslim proselytes [153]. But Coulibaly was no root-and-branch enthusiast in his attack on the attack thesis. He suggested that confusion may have arisen between Ghana as the name of a country, and Ghana the name of the capital city [154]. He also accepted the existence of an original Almoravid ribat as a physical building, while at the same time noticing that some modern scholars doubt this [155]. Like Semonin, Coulibaly attributed his difficulties in trying to define l'attaque de Ghana to the insufficiency of sources and of documents; but he was certainly very dubious about the whole thing, and in his conclusions spoke of the myth of Almoravid conquest from Ghana to the gold-bearing areas of Bambuk, and of an Almoravid empire streching from Muslim Spain to the banks of the Senegal River [156]. Coulibaly also introduced a new element into the discussion by using the results of the archaeological excavations in Tegdaoust (the putative site for Awdaghust) and Kumbi Saleh (the putative site for Ghana) as arguments against the conquest. He pointed out, correctly, that no indisputable remains of Almoravid presence have been found at either site [157]. In the same year, 1974, Farias returned to the fray in a review article of Ancient Ghana and Mali, by Nehemia Levtzion [158]. Farias argued against the hypothesis in both general and specific terms. In general, he pointed out that it is too much of an oversimplification to think that the relationship between the sedentary Sudanese and the pastoral Berbers would have been constantly violent forever (as some of the 14th-century Arabic and 19th-century European sources suggest). Quite the contrary, he stressed the co-operative and symbiotic aspects of their relationship [159]. More specifically, Farias rejected Levtzion's interpretation of the Wagadu legend in oral tradition, with its killing of the sacred snake. This killing might appear to correspond all too well with the idea of a bloody Almoravid conquest of Ghana, the snake myth thus being automatically and rashly treated as actual historical evidence [160]. But, said Farias, "the interpretation of the mythical killing as a dramatic representation of an event is not easy to accept" [161]. In the case of the Almoravids and Ghana, this meant that Farias - like Semonin before him - rejected the idea of the destruction of Ghana by the Almoravids in 1076, as manifested by Delafosse; and instead of conquest, Farias used the term "influence" [162]. This angle was taken up again by John Hunwick, in 1980, who argued that the Almoravids did not actually conquer Ghana, but had great influence on its affairs, helping a Muslim faction of the ruling house rise to power [163]. This idea had been introduced in 1970 by Jean Devisse, who speculated that the relationship between the new ruler of Ghana and the Muslim population might have changed for worse after the death of King Basi in 455/1063, who had been famous for his friendship with Muslims, eventually causing an Almoravid intervention [164]. Similar thoughts have been expressed also by Daniel F. McCall in 1971, Harry Norris in 1982, and recently by Dierk Lange [165]. All this suggests that several scholars were prepared to give up the Delafossean conquest hypothesis, which they realised is indefensible or at least weak and all but undocumented, and the formulation of the "influence hypothesis" was an attempt to give a more reasonable answer to the question, why the ruler and people of Ghana adopted Islam in 1076. There seems to us little reason to believe that the Almoravids exercised any kind of determining political influence over Ghana, be it by force, or by insinuation. The story of a forceful intervention originates (though it was subsequently much embroidered and enlarged) from the unreliable, for Ghanaian affairs, Arabic chronicles of the 14th century and their modern interpretations in colonial historiography. Remove the element of force, and we are left with no reliable peaceful alternative: our sources, even after we have excised the use of force, remain as reliable, or unreliable, as they were before. Furthermore, the hypothesis of Almoravid "influence", or Almoravid support for a Ghanaian coup-d'état, is linked to the existence of a strong Almoravid "southern wing" led by Abu Bakr after the division of the movement in early 1070s. There is, however, little evidence for this: Ibn Idhari and other sources mention Abu Bakr's return to the desert, and his death there, but with very little additional detail indeed [166]. In fact, the idea of "southern wing" may itself be understood as a by-product of the conquest hypothesis, the historicity of which requires the presence of a considerable Almoravid force in the south [167]. The reconsideration of the conquest theory received its most radical expression in 1982-3, first in Humphrey J. Fisher's review article "Early Arabic sources and the Almoravid conquest of Ghana", and then in a two-part joint article by David C. Conrad and Fisher, "The conquest that never was: Ghana and the Almoravids, 1076" [168]. They build a rather complex argument upon a relatively basic premiss: given that Ghana is by far the most often cited name in Arabic writing about western Africa from the beginning to about 1500; given that the Almoravids were also a popular topic in this writing, and that often both Ghana and the Almoravids are mentioned by the same author; and given that, despite this interest (sustained over centuries), there is nowhere even a single direct reference to any such destructive conquest of Ghana by the Almoravids as described in the western historiography on Africa - then the logical conclusion seems clearly to be that no such cataclysmic event occurred: "there was no Almoravid conquest of ancient Ghana." [169] Reaction to this anti-conquest argument
has, predictably, been mixed [170]. There was no serious attempt to defend
the conquest hypothesis in the 1980s: the whole issue
only boiled over again ten years later, in 1992, with the
appearance of Sheryl Burkhalter's
article, "Listening for silences in Almoravid
history: another reading of 'The Conquest that never was'"
[171]. She strongly criticises Conrad
and Fisher, particularly for ignoring the contexts in
which the Almoravids and Ghana are mentioned in the
Arabic sources [172]. A little surprisingly, Burkhalter
nowhere affirms that the conquest did actually take place.
On the contrary, her phrasing seems sometimes quite
compatible with the anti-conquest case as presented by
Conrad and Fisher - she speaks of "more traditional
interpretations of Almoravid history", and of "conventional
wisdom concerning an Almoravid 'victory' over Ghana"
[173]. Her principal objection to
Conrad and Fisher seems to be methodological - the
problem of individually contextualising all the dozens of
sources which might be expected to mention the conquest,
but do not - with some additional condemnation of Conrad
and Fisher's suggestions, very subsidiary in the context
of their article, about what might actually have been
happening between Ghana and the Almoravids, if we
eliminate the idea of a conquest. Nevertheless, it seems
that in the most recent histories of Africa the sceptical
attitude towards the conquest and destruction of Ghana by
Almoravids has achieved a lasting foothold [174]. One of our principal reasons for surveying in such detail the gradual process of osmosis, by which first the conquest hypothesis, and later the criticism of it, have become established in West African historiography, was to test the extent to which these changes have been a response to changing ideological conditions in European thought. We have described in this paper important periods of historiographical transformation, which were more or less, we have suggested, ideologically led. First there was the Moroccan interpretation of Almoravid history, embedded within Arab and Berber attitudes towards the blacks; both this interpretation and the attitudes were carried over to the Renaissance European literature of Africa. Later there were Europeans' own experiences in early 19th-century Africa, which also affected their conceptions of African past. And finally there was the rise of a new science, and a new philosophy to go with it, in Europe itself, towards the end of the same century, contributing to the new politics of colonialism in Africa, and stimulating changes to the historical record of Almoravid days in particular, calling up the conquest of Ghana. Has the same sort of thing - a radical shift in politics and ideology, and a radical shift in historical understanding to go with it - been happening since the 1960s, with the new demands which decolonisation makes upon European scholars, forcing them to take up a more conciliatory attitude towards Africans and their past? Is one reason for the increasingly positive reception given to the anti-conquest hypothesis that it corresponds better with current ideological tendencies both in European thought and in the historiography of Africa than does the earlier idea of destruction? It is, we confess, very hard to accept that the changed perception, in the specific cases with which we are concerned (was there an Almoravid conquest of Ghana? and if not, where did the notion came from?), is just a result of evolving ideology; we believe rather that it is an expression of the methodological progress which has greatly accelerated in the historiography of Africa from the early 1960s, and which is now leading to a new situation in African historical research, with emancipation from the burden of colonial tradition. What is needed is to re-examine early West African history, from the very beginning, not by picking out only such evidence as corresponds with the knowledge we have inherited from earlier researchers and which we consider justified as such since it has been repeated for a century, or as corresponds with the attitudes of our time towards the African past; but by using all the sources we have available to us now, not only textual, and by interpreting them without any fixed presuppositions for or against. In this task, we must, if necessary, dare to challenge all the previous hypotheses, even if they were the holy of holies. In a short story entitled "One Hundred Views of Mount Fuji", the Japanese author Osamu Dazai contemplates his relation to this famous volcano which appears to him hopelessly vulgar. Yet he knows that most people, including westerners, are against his view. To them, Fuji is "wonderful", «simply because they've heard so much about it and yearned so long to see it; but how much appeal would Fuji hold for one who's never been exposed to such popular propaganda, for one whose heart is simple and pure and free of preconceptions? It would, perhaps, strike that person as almost pathetic, as mountains go.» [175] Are we right in
believing that this is what we have been, and are, doing
with this paper? That we are those pure in heart and free
of preconceptions? There seem to be at least three
alternative answers. The first is the worst: no, we are
not right, for we are as much at the beck and call of the
ideological tide surging all around us, as Leo Africanus
and Maurice Delafosse were representives of their own
ages, and as someone in the future will demonstrate us to
have been. The second is less sombre, but still very
sober: we are partly right, inasmuch as we have done what
we have could, forwarding the discovery of historical
truth to the best of our ability, just as did earlier
European Africanists of this century and the preceding
one, and just as did the Arab Africanists of the
fourteenth century, all of us making some progress, but
still much more shackled by ideology and misunderstanding
than we can possibly realise at the time. The third is,
quite likely, far too triumphalist: yes, we are right,
and we, unlike our predecessors here or in Africa, have
been able to strip away all the ideological veils, and to
search for, and partly to see, what actually happened. [1] The earliest version of this paper was
presented in May 1994 to the African History Seminar at
SOAS, and a more elaborated version in March 1995 to the
International Conference on Mande Studies in Leiden. We
are grateful to the participants in these two events for
their comments, especially to Michael Brett, Paulo F. de
Moraes Farias, Lansin‚ Kaba, Pertti Luntinen, Harry
Norris and Ed van Hoven for their more extensive and very
precious help. Surviving errors and eccentricities
remain, of course, our own. [2] Daniel F. McCall, "Islamization of the Western and Central Sudan in the Eleventh Century", Boston University Papers on Africa, 5 (1971), 7; Carmen Navarro, "Los Almorávides y la islamización de Bilad al-Sudan", Studia Africana, 4 (1993), 139-40. On effects of the alleged event, see for example Nehemia Levtzion, Ancient Ghana and Mali (London 1973), 45; and Joseph Ki-Zerbo, Histoire de l'Afrique Noire d'hier à demain (Paris 1978), 117-18. [3] See Michael Doran, "The Maritime Provenience of Iron Technology in West Africa", Terrae Incognitae, 9 (1977), 89-98. [4] E. Gozalbes Cravioto, "Comercio y exploraciones del Sahara en la Antigüedad Clásica", Estudios Africanos, 12-13 (1993), 13-9; for an assertion for the Carthaginian sailings, see Werner Huß, "Die Antike Mittelmeerwelt und Innerafrika bis zum Ende der Herrschaft der Karthager und der Ptolemaier", in H. Durchhardt et al. (eds.), Afrika. Entdeckung und Erforschung eines Kontinents (Cologne 1989), 4. [5] On the origins of the Periplus, see P. E. H. Hair, "The Periplus of Hanno in the History and Historiography of Black Africa", History in Africa, 14 (1987), 43-66; also Gabriel Germain, "Qu'est-ce que le Périple d'Hannon? Document, amplification littéraire ou faux intégral?", Hespéris, 44 (1957), 205-48. [6] See David Henige, "The Race is not
always to the swift. Thoughts on the Use of Written
Sources for the Study of Early African History", Paideuma,
33 (1987), 55: "A shockingly high proportion of work
done on early African history between, say, 1955 and 1975,
can now only be termed quaternary in that it was content
to rely on secondary and tertiary accounts, because few [historians
of tropical Africa] recognized the need to seek out
primary sources at whatever cost and base their work on
them." [7] We know only that Islam was spreading in
Ghana by the time of Almoravids (1054-1147), which is
confirmed by Arabic sources, like al-Bakri (writing ca.
1068), al-Zuhri (ca. 1137), al-Idrisi (1154), al-Sharishi
(before 1222), and the anonymous authors of Kitab al-Istibsar
(1191) and al-Hulal al-Mawshiyya (1381). See the
relevant passages in J. F. P. Hopkins & N. Levtzion (eds.),
Corpus of Early Arabic Sources for West African
History (Cambridge, 1981), 79, 98, 109, 146, 153 and
310. On the Arabic sources, see also Dierk Lange, "The
Almoravids and the Islamization of the Great States of
West Africa", Res Orientales, 6 (1994), 65-67. [8] On Leo's life and travels, see for example Raymond Mauny, "Note sur les 'grands voyages' de Léon L'Africain", Hespéris, 41 (1954), 379-94. [9] "Descrittione dell' Africa, et delle cose notabilis che iui sono, per Giouan Lioni Africano", in Giovanni Battista Ramusio, Delle Navigationi et Viaggi (3rd ed.: Venice 1563), vol. I:ff. 1-95. [10] For bibliographical information about these various editions, see Oumelbanine Zhiri, L'Afrique au miroir de l'Europe: Fortunes de Jean Léon l'Africain à la Renaissance (Geneva, 1991), 227-28. [11] See for example P-A. Lattreille, Recherches
géographiques sur l'Afrique centrale, d'après les écrits
d'Edrisi et de Léon L'Africain comparés avec les
relations modernes (Paris 1824), 4. [12] Ramusio (3rd edition, vol. I, 1563), fo. 77. See also Leo Africanus, The History and Description of Africa, tr. J. Pory in 1600, ed. R. Brown (London 1896), III, 819; Jéan-Leon L'Africain, Description de l'Afrique, tr. A. Épaulard, eds. A. Épaulard, Th. Monod, H. Lhote and R. Mauny (Paris, 1956), II, 461. Leo calls Ghana Cano, the name which he gives elsewhere to the city and kingdom of Kano in Hausaland (1563, fo. 79; 1896, III, 829-30; 1956, II, 476). The respective contexts, the first with references to "il Bichri" [al-Bakri] and "el Meshudi" [al-Masudi] and the second to Casena [Katsina] and other landmarks in Hausaland, make clear that these two places are entirely different, though confusion between Ghana and Kano was widespread in western geographical literature well into the 19th century. See for example Lattreille 1824, 4 and 24; and Friedrich Stüwe, Die Handelszüge der Araber unter den Abbassiden durch Afrika, Asien und Osteuropa (Berlin 1836), 101-2. [13] In Ramusio 1563, fo. 77: "Tutti adunque questi paesi sono habitati da huomini, che viuono à guisa di bestie, senza Re, senza Signore, senza republiche, & senza gouerno & costume alcuno, & appena sanno seminare il grano." For translations, see 1896, III, 819; 1956, II, 461. [14] Ramusio 1563, fo. 5; 1896, I, 7; 1956, I, 8. These names resemble the list of the principal Sanhaja tribes given by Ibn Khaldun: Gudala, Lamtuna, Masufa, Watzila, Targa, Zaghawa and Lamta (see Corpus, 327 and 321). [15] In Ramusio 1563, fo. 77: "Giuseppe
Re & edificator di Marocco del popolo di Luntuna,
& i cinque popoli di Libia dominarono questi Negri,
& alloro insegnarono la legge di Macometto, &
l'arte necessarie al viuere: et molti di loro si tecero
Mahumettani". See also the translations: 1896, III,
820 ("These Negros were first subiect vnto king Ioseph
... and afterward vnto the fiue nations of Libya...");
1956, II, 462 ("Joseph ... et les cinq peuples de
Libye dominèrent ces Noirs..."). [16] Leo also, it seems, associates the conversion of Mali with the Almoravids (1563, fo. 78; 1896, III, 823; 1956, II, 466). Much of this is highly implausible - al-Bakri, for instance, whom Leo himself cites, gives a circumstantial account of the first conversion of Mali by a Muslim trader, free from any Almoravid involvement. Or perhaps Leo confused Mali, which was widely called Takrur by the 14th-century North African writers, with that Takrur described by al-Bakri which was a close ally of the Almoravids, though here too al-Bakri does not mention the Almoravids in connection with the initial conversion of Takrur (see Corpus, 73 and 77; also Ibn Abi Zar, in ibid., 239). [17] See Ibn Abi Zar, Ibn Khaldun and al-Maqrizi, in Corpus, 248, 331, and 355. [18] On this attitude, see Bernard Lewis, Race
and Color in Islam (New York 1971), 37-38. [19] The first and second volumes were printed in Granada in 1573; the third, which contains Mármol's principal description of sub-Saharan Africa, in Malaga in 1599. A French translation, L'Afrique de Marmol, by Nicholas Perrot d'Ablancourt, appeared in Paris, also in three volumes, in 1667. A modern Spanish reprint was published in Madrid in 1953. [20] Mármol mentions several times, for
instance, such Arabic writers as el Moçaudi/Mucaudi
(al-Masudi), el Bebquer/Bubquer (al-Bakri),
Ibni Alraquiq ("an ancient African writer",
Ibn al-Raqiq), Abdul Malic ("a Moroccan
historian", Ibn Abd al-Malik al-Marrakushi?), and Aben
Gezar ("an African geographer", Ibn
Juzayy?). On the other hand, he mentions Leo Africanus by
name only once (see 1573, I, fo. 17; 1667, I, 36). On Mármol's
life, see James T. Monroe, Islam and the Arabs in
Spanish scholarship (Leiden 1970), 16-17. [22] 1599, fo. 21; 1667, III, 57: "En la
descripcion que hazen el Mucaudi, y Bubquer, y Aben
Gezar, y otros Geografos Africanos, solamentely hazen
mencion en la tierra de los negros de Guequin, y Cano,
porque no deuieron tener tan particular noticia della,
como se tiene agora. Todas las provincias que confinan
con la Zaara, o cerca della son el dia de oy Mahometanos,
por que reynando los Almorauidas en Africa, y el pueblo
de Lumtuna. Cerca de los trezientos y ochenta años de la
Hixara, que fueron nouecientos y ochenta y dos de Christo
redemtor nuestro, vuo entre ellos muchos morabitos, y
alcoranistas predicatores de la maldita seta de Mahoma,
que la enseñaron a quella gente barbara, y los traxeron
a su opinion. Y despues metiendose por la Etiopia entre
aquellos pueblos negros Hagin hijo de Abdulmalic, en el año
de quatrocientos y sessenta y nueue de la Hixara, les
començo a enseñar sus ritos y ceremonias, y otro
setaria, llamado Yahaya hijo de Ali Benbucar, acabo de
conuertir todos los que caen en la ribera del rio Niger,
y cerca del, que..." [23] Mármol calls Almoravids with their Spanish denomination "Almoravidas", whereas the "morabitos" were properly a sect established by "Mahamat Mohaydin", the last descendant of "Ali Hussein", son of Caliph Ali. Adherents of this sect resemble the Turkish "dermisios", or dervishes. According to Mármol, the Almoravids received their name because their founder and other leaders were morabitos (1573, I, fo. 59ff and 149; 1667, I, 125ff and 282). [24] See Corpus, 310; al-Hulal is here quoting, not entirely accurately, from the Kitab al-Jughrafiya of al-Zuhri (see Corpus, 98). [25] 1573, I, fo. 45; 1667, I, 96. [26] See 1573, I, fo. 149ff; 1667, I, 283ff. [27] 1573, fo. 152; 1667, I, 286. Mármol's
source for this information is Abdul Malic,
"choronista de Marruecos" (1573, II, fo. 27;
1667, II, 50). Mármol's ignorance concerning the early
Almoravid history, before the reign of Yusuf, suggests
that he did not actually know the text of al-Bakri well. Abu
Texifien seems to be a fictitious character, which
embraces all the Almoravid leaders who lived before Yusuf.
In reality Abu Bakr b. Umar was not the father, but a
cousin of Yusuf b. Tashfin. [28] 1599, III, fo. 23; 1667, III, 62. The association of Almoravids with the Lamtuna is strong in Mármol: he calls Yusuf b. Tashfin "el Rey Iuzef Lumtuna". [29] We should also notice that this information was published eight years after the actual Moroccan invasion in Timbuktu took place, yet we may only guess, whether this event affected Mármol. There are no references to al-Mansur's campaign to the Western Sudan in Mármol, but it is not impossible that he had heard of it. There were Spaniards in Marrakesh when news of the victory arrived, in June 1591, and they very probably passed the news to Spain. See H. de Castries, "La cônquete du Soudan par El-Mansour (1591)", Hespéris, 3 (1923), 433-4. [30] 1599, III, fo. 23 (1667, III, 62): "Quando el Xerife Mahamet estaua en su prosperidad, comoidado de las ofertas de los pueblos de Libya, quiso yr a conquistrar estos pueblos de negros, como lo auian hecho antiguamente los Lumtunas." [31] Nehemia Levtzion, "The western
Maghrib and Sudan", in Cambridge History of
Africa, vol.3 (Cambrige 1977), 400-1 and 410-11. [32] According to Mármol, "el Xerife Mahamet" did send his troops to the south, but having encountered the "King of the Blacks", referring here to Askiya Ishaq I of Songhay, with his army of 300,000 warriors, the Moroccans decided to retreat without fighting (1599, III, fo. 23; 1667, III, 62). [33] On the 14th-century Marinid historiography, see Maya Shatzmiller, L'Historiographie mérinide: Ibn Khaldun et ses contemporains (Leiden 1982). [34] See Abderrahman es-Sadi, Tarikh es-Soudan,
tr. O. Houdas (Paris 1898-1900; reprinted 1964 under
UNESCO auspices), French tr., 163-4; Arabic text, 99-100.
Yet the Moroccan ruler here referred to is "Mawlay
Ahmad the Great", meaning most likely Ahmad al-Araj,
whom his brother Muhammad al-Mahdi superseded in 1544.
Ahmad's counterpart was Askiya Ishaq I (1539-49), and
their dispute was about the possession of Taghaza, an
important salt mine in western Sahara, and by then under
Songhay control. [35] The Moroccans kept on demanding Taghaza still after the unsuccessful attempt of Muhammad al-Mahdi. According to al-Sadi, Sultan Mawlay Ahmad al-Mansur (1578-1603) sent his troops to the south in 1584 to capture "all the cities which they meet on the banks of the River [Senegal] and elsewhere, and then continue their way into Timbuktu". But "it was God's will" that the Moroccan army perished in the desert. Later another expedition was sent, which eventually occupied Taghaza, but they had to return to Marrakesh in autumn 1585, since the oasis was abandoned. But here too, al-Sadi mentions no Almoravid pretext to justify this sneak raid (1898-1900, French tr., 193; Arabic text, 120). Neither does he mention such when he describes the content of the documents, which were sent to Askiya Ishaq II by Ahmad al-Mansur in late 1589, and in which the latter repeated his demand for the possession of Taghaza, because he protects Songhay from Christian attacks (although Ahmad al-Mansur's real aim was the subjugation of entire Songhay empire). It is particularly interesting that al-Sadi specifically emphasizes that he has seen the originals of these documents himself. Had they contained any references to a previous Almoravid conquest to justify Ahmad al-Mansur's claims, it is quite plausible to suppose that al-Sadi would have mentioned it (see ibid., French tr. 215-6; Arabic text, 137; also Levtzion 1977, 413). [36] Al-Sadi knew al-Hulal al-mawshiyya,
which he cites as a source for the history of the Sanhaja.
Al-Sadi says that they make holy war against the blacks,
and he also mentions amir Abu Bakr b. Umar by name, but
he nowhere says that the Almoravids had conquered the
blacks, or even converted them to Islam in the year 469/1076-7
(1898-1900, French tr. 42-44; Arabic text 25-26; compare
to al-Hulal in the Corpus, 310-11 and 313-14). [37] For the date of composition, see J. O.
Hunwick, "A new source for the biography of Ahmad
Baba al-Tinbukti", Bulletin of S.O.A.S, 27 (1964),
592. For an English translation, with the Arabic text, of
al-Kashf, see Bernard Barbour & Michelle
Jacobs, "The Miraj: a Legal Treatise on Slavery by
Ahmad Baba", in J. R. Willis (ed.), Slaves and
Slavery in Muslim Africa (London, 1985), vol. I, 125-59. [38] Al-Kashf, in Barbour & Jacobs 1985, English tr. 137; Arabic text, 158. [39] Ibid., English tr. 128; Arabic text, 142. Cited also in Uthman ibn Fudi, Bayan wujub al-hijra ala'l-ibad (The exposition of the obligation of emigration upon the servants of God), ed. and tr. F. H. El Masri, (Khartoum and Oxford 1978), English and Arabic texts, 51. [40] Al-Kashf, English tr. 129; Arabic
text, 143-44. [41] Geographia Nubiensis, id est accuratissima totius orbis in septem climata divisi descriptio, continens praesertim exactam universae Asiae, & Africae... tr. Joannes Hesronita and Gabriel Simonita (Paris). This translation was based on an abridged Arabic edition, published in Rome in 1592 (De Geographia Universali...). See Vincent Monteil, "L'oeuvre d'Idrisi", Bulletin de l'I.F.A.N, 4 (1939), 837-57. [42] Travels in the Inland Parts of Africa (London), page v in "Letter to the Editor". The Spanish atrocities during the conquest of their American colonies were widely known in Europe: the original source, Brevissima Relación de la Destruyción de las Indias, published by Bartolomé de Las Casas in Sevilla in 1552, was rapidly translated in Dutch, French, English, German, Italian and Latin - for the bibliographical information about these editions and their availability, see Juha Pekka Helminen, "Bartolomé de las Casas in History, or an Example of How Historical Persons can be Used for Different Purposes", in Antero Tammisto et al. (eds.), Miscellanea, (Studia Historica, no. 33; Helsinki 1989), 84-85. [43] Travels in the Interior Districts of Africa (London), 112. [44] Mary Louise Pratt, Travel Writing and
Transculturation (London and New York 1992), 74.
French and German translations were both published
already in 1799, but Park's Travels soon became
available in minor European languages, too. A Swedish
translation, for example, was published in 1800. [45] The Latin text is found in H. A. Hamaker, Specimen catalogi codicum mss. orientalium bibliothecae academiae Lugdono-Batavae (Leiden 1820), 207-9. The English, including brackets, is from the Corpus, 355. On the origins of al-Maqrizi's fragment, see Dierk Lange, "Un text de Maqrizi sur 'Les races des Sudan'", Annales Islamologiques, 15 (1979), 187-90. [46] Historia dos Soberanos Mahometanos das primeiras quatro dynastias, e de parte da quinta, que reinarao na Mauritania, escripta em arabe por Abu-Mohammed Assaleh, filho de Abdel-halim, natural de Granada, tr. Jozé de Santo Antonio Moura (Lisbon). Extracts had appeared earlier, in José Antonio Conde's Historia de la dominacion de los árabes en España, vol. II (Madrid 1820). [47] Historia dos Soberanos, 147. The
English, excluding brackets, is from the Corpus,
248. [48] "Notice d'un manuscrit arabe contenant la description de l'Afrique [man. de la Bibliothèque du Roi, no.580]", Notices et Extraits des Manuscrits de la Bibliothèque du Roi, 12, p.642 n.1. This note is attached to the passage describing the ruler of Ghana and his court (see al-Bakri, in the Corpus, 79-80). [49] See Corpus, 98 and 310. [50] (London; reprinted by Frank Cass in 1966). [51] Both al-Bakri and Ibn Khaldun were
available to him in the British Museum (see 1841/1966, pp.4
n.5 and 61 n.104), while a copy of al-Zuhri, "An
anonymous Arab geographer", and other Arabic mss.
were lent to him by his close friend Pascual de Gayangos,
the famous Spanish Arabist (pp.19 n.33 and 127 n.210). [52] 1841/1966, 66. [53] Cooley's reference to Moura (p.146) points actually to the passage where Abu Bakr hands his authority as the true leader of Almoravids to Yusuf, but he was clearly thinking about the following page and the passage which we have quoted. [54] Corpus, 73-74 and 68. [55] Ibid., 73-74. [56] Ibid., 22, 45-49, 62 and 69. [57] Ibid., 73; the editors insert "[the city of]" before the last word. [58] Ibid., 22. [59] Ibid., p.385 n.25, referring also
to Ibn Khaldun on p.332. [60] See 1841/1966, 6, 27 and 29. [61] See for example Ibn al-Athir, in Corpus, 158. [62] This translation, including brackets, is
from the Corpus, 333. Both the Arabic sources
cited by the Corpus include "their property
and their country" after "pillaged". [63] 1841/1966, 62. [64] It is a bit strange that Cooley makes no reference to al-Maqrizi in this context. He knew Hamaker's Specimen catalogi (see p.29 n.51) and his Negroland in fact contains an English translation of al-Maqrizi's fragment, however omitting, for some reason, the last passage concerning Ghana and the Almoravids (see pp.119-20). Perhaps there was no need for al-Maqrizi, since Ibn Khaldun was the more authoritative witness. [65] On the Susu conquest, see Stephan Bühnen, "In Quest of Susu", History in Africa, 21 (1994), 4-7. [66] 1841/1966, 69. [67] Untersuchungen über die Negerländer der Araber und über den Seefandel der Italiener, Spanier und Portugiesen im Mittelalter (Göttingen; reprinted 1966 by Meridian Publishing Co., Amsterdam), 74. Wäppaus knew Cooley's Negroland for he mentions it in his preface (p. v), but he had apparently received the book too late to be used by him, because he nowhere cites it. The principal sources of Wappäus for Ghana were Leo Africanus, Quatremère's translation of al-Bakri, and an English translation of al-Idrisi published in Annals of Oriental Literature, vol. I (London 1820). [68] Afrika vor den Entdeckungen der
Portugiesen, p.28, being the publication of a Fest-Rede
which Kunstmann had held in Munich, at the "Königlichen
bayerische Akademie der Wissenschaften", on the 29th
March 1853. Al-Hulal does specify conversion from
Christianity, but this is apparently a mis-reading of al-Zuhri's
text which specifies kufr, "unbelief" (see
the Corpus, 98, 310, also 389, n.16). Kunstmann
knew Wappäus's work (p.37 n.1), but nowhere mentions
Cooley's Negroland, although he was probably aware
of it. [69] Cooley calls this ms. "Prolegomena" (1841/1966, p.61 n.104), or the Muqaddima; yet all the quotations are clearly taken from Kitab al-Ibar. [70] Re-edited by Paul Casanova and reprinted in 3 volumes (Paris, 1925-34). De Slane published also French translations of al-Bakri (Journal Asiatique, 5. série, vols. 12-14, 1858-9) and Ibn Khaldun's Muqaddima (Notices et Extraits des Manuscrits de la Bibliothèque Imperiale, vols. 19-21, 1862-68). [71] Ibn-Khaldoun, Histoire des berbères (Paris 1852-6), II, 110; italics added. [72] Ibn-Khaldoun, Histoire des berbères
et des dynasties musulmanes de l'Afrique septentrionale
(Alger 1847-51), Arabic text prepared by M. de Slane, vol.
I, 263, lines 16-17. [73] If Cooley had "discovered" Ghana and Mali, Barth did the same for Songhay, for he was able to find in Timbuktu a copy of al-Sadi's Tarikh al-Sudan. An abridged German translation, based on Barth's own notes, was published already in 1854 (C. Ralfs, "Beitäge zur Geschichte und Geographie des Sudan, eingesandt von Dr. Barth", ZDMG, vol. 6). A complete version of Tarikh al-Sudan was found by a French traveller, Felix Dubois, in Jenne in 1896 (Tombouctou la mysterieuse, Paris 1897, 356). [74] Barth's Travels and Discoveries was published simultaneously in five volumes in English (London 1857-8), and in German (Reisen und Entdeckungen in Nord- und Central-Africa, Gotha 1857-58). The English version was very soon afterwards published in three volumes in New York, in 1857-59. The modern reprint, "Centenary edition", published in London in 1965 by Frank Cass, is based on the American edition with additional maps from the English edition. For reasons of availability, our references are to this reprint. [75] 1965, III, 660. In the original English and German editions this table is included in the fourth volume. [76] 1841/1966, 66. [77] See 1965, I, 365 and II, 22. [78] See the footnotes in 1965, III, 658-61. [79] Dutch, Danish and French translations appeared, but all are incomplete, missing the appendices where Barth introduced his ideas on West African history (see A. H. M. Kirk-Greene's "A Bibliographical Note", in Barth 1965, I, xvi). [80] Barth's triumphant homecoming from Tripoli to London and Berlin in autumn 1855 was keenly observed even in Finnish newspapers, which did not then contain much foreign news (and hardly any from Africa), except concerning the ongoing Crimean war. See for example Åbo Unterrättelser, 21.9.1855, 30.10.1855 and 27.11.1855. [81] Annales Regum Mauritanie (Uppsala). [82] "Notice sur les Almoravides et les Almohades d'après les historiens arabes", nos. 69, 71, 76, 77 (in vols. 12 and 13 of the Kraus reprint of 1968). [83] 1868-69, no. 69, p.222. Mercier nowhere refers to Cooley's Negroland, and the fact that he dates the Almoravid conquest of Ghana shortly before the death of amir Yahya b. Umar in 448/1056-57 (see al-Bakri, in Corpus, 73) - and not in 469/1076-77 - proves clearly that he had not been influenced in this by Cooley. Furthermore, Mercier says nothing of Abu Bakr b. Umar's campaigns in the south, except repeating Ibn Abi Zar's superficial account of his death (p.224). [84] See 1868-69, no. 69, p.217 n.1. [85] Mercier's spelling reflects the old-fashioned, but nonethless identifiable, rendering of the Arabic consonant ghayn with r', which is nowadays represented with "gh". [86] 1868-69, no. 69, 222. [87] See for example the article "Les Almoravides" by O. Houdas in La Grande Encyclopédie (vol. 2, Tours 1886, 426); also A. Müller, Der Islam im Morgen und Abendland (Berlin 1887), 116. [88] Vol. II, 25. [89] "Mélanges d'histoire et de littérature orientales II", Le Muséon, vol. 7 (1888), 49-60 and 137-151. Also cited under the title "Essai sur l'histoire et de la langue de Tombouctou et des royaumes de Songhaï et Melli", which is the opening sentence of the article. [90] Ibid., 51. In the main text Basset says that Islam was adopted in Ghana in the year 1000 AD, referring in this context to an article published by General Faidherbe in Revue Scientifique 15.11.1884 ("Tombouctou et les grandes voies commerciales du nord-ouest de l'Afrique", pp.609-13). In the footnote (n.4) he, however, says that the date 1076 given by Cooley is more exact. [91] See ibid., pp.51 n.4, 56 n.1,
and 57 n.1. [92] Du Niger au Golfe du Guinée, (Paris), vol. II, 383. This sounds like a distant echo of Mercier 1868-69. [93] Ibid, II, 369, and 391. The date 1607 appears on p.383, and may be a misprint, especially since on the opposite page Binger says that al-Bakri was writing in 1067-68. However, the mention of Ouqaïmagha - the Wakajamaga in Ralfs' edition of Tarikh al-Sudan (1854, 526) - on p.383 suggests that Binger may have here confused al-Bakri with al-Sadi; elsewhere (p.369) he is apparently confusing him with al-Idrisi. [94] Binger 1892, I, 386; II, 379 and 381. [95] "Légende et traditions des soninké
relatives à l'empire de Ghanata", Bulletin de géographie
historique et descriptive, no. 2 (1895), 472-80. The
first version of the Wagadu legend had been published as
a French translation already in 1879 (see L. B-J. Bérenger-Féraud,
Les peuplades de la Sénégambie, Paris 1879, pp.
169-72, "Légende du serpent de Bambouk"). [96] Barth had supposed that Ghana had been founded by the Fulani, though its population were the Aswanék, or the Soninke (Barth 1965, III, 657 and 703). In the 1890s some French scholars, however, began to claim that Ghana had been the first of the Songhay empires; the second was the empire of the askiyas (see Le Chatelier 1899, 79; and O. Reclus, "Songhaï", La Grande Encyclopédie, vol. 30, Tours 1901, 271). [97] Le Chatelier 1899, 45. Le Chatelier (p.
46) extends the rule of the Lamtuna over Ghana until the
Susu conquest. [98] p. 110. This sounds like Cooley, certainly known to Shaw though she uses footnotes only seldom and includes no bibliography. According to her biography, Shaw wrote her book in England using "Spanish archives and translations of Arab works dealing with the occupation of Negroland" (E. Moberly Bell, Flora Shaw, London 1947, 253). [99] See for example Paul Meyer, "Erforschungsgeschichte
und Staatenbildungen des Westsudan", Petermanns
Mitteilungen, Ergänzungsheft no. 121 (Gotha 1897),
61: "The fanatic Meràbetin from Morocco, the
Almoravids of the Spaniards, pillaged Audaghost in 1052
... in 1076 [they] conquered Ghanata, which became now
completely islamized and a subject of contention between
the Berbers of the north and the Negroes of the south and
east." Within this passage, Meyer refers to Barth. [100] Paris; reprinted, with a preface by Robert Cornevin, in 1972 (3 volumes, Paris). [101] Delafosse was chosen to write Haut-Sénégal-Niger because he had already gained fame with his various linguistic works. Haut-Sénégal-Niger became immediately popular, and its author was decorated with three gold medals. Very soon Delafosse was regarded as the highest authority in the early West African history not only in France but also in the Anglophone world, though Haut-Sénégal-Niger was never translated into English (Ed Van Hoven, "Representing social hierarchy. Administrators-Ethnographers in the French Sudan: Delafosse, Monteil, and Labouret", Cahiers d'études africaines, no.118, 1990, 181 and 185). See also Delafosse's biography, written by his daughter Louise Delafosse (Maurice Delafosse: le Berrichon conquis par l'Afrique, Paris 1976). [102] Delafosse 1912, II, 54. [103] It may be too cynical a thought, but it is at least possible that there was an inclination (surely subconscious), by admitting that we do not know this particular detail, to divert attention from the fact that we do not necessarily know any of the described details. [104] A history of Islam in West Africa (London 1962; reprinted 1970), 55; see also 25, 29-30, 33. [105] (Paris), see p. 25. Hamet's principal source for the Almoravid history is de Slane's translation of Kitab al-Ibar (see p.4). [106] Hamet 1911, 1-2. These limits for the Almoravid empire were rooted in western literature already much earlier (see for example Léon Godard, Description et histoire du Maroc, vol. I, Paris 1860, 314). [107] Ibid., 24. [108] Ibid., 6. Hamet gives no dates
for these events, except 1052 for the beginning of
Almoravid movement. Yet it sounds that he is here
repeating Mercier's (and Cooley's) confusion of Awdaghust
as the capital of Ghana. [109] (Westminster), 262. Arnold's references for this passage are "Leo Africanus (Ramusio. Tom. i. pp.7, 77)" - see our n.15 - and "Chronik der Sultane von Bornu, bearbeitet von Otto Blau, p.332 (Z.D.M.G. vol. vi. 1852)". [110] pp. 317-8. Arnold cites Meyer's paper of
1897 for this addition. [111] See Cooley 1841/1966, 29; Godard 1860, I, 307; Henri Duveyrier, Les Touaregs du nord (Paris 1864), 324-5. [112] See for example al-Bakri and Mafakhir al-Barbar, in the Corpus, 75-6 and 233. [113] George F. Lyon, A Narrative of
Travels in Northern Africa in the years 1818, 19, and 20
(London 1821), 109; Henri Duveyrier, "Note sur les
touareg et leur pays", extracted from Bulletin
de la Société de Géographie (February 1863), 15-16. [114] "An Abstract of Mr. Park's Account of his Travels and Discoveries, abridged from his own Minutes by Bryan Edwards, Esq. 1798", Proceedings of the African Association, 1810/1967, I, 349. [115] Ibid.; see also G. Mollien, Travels in the interior of Africa to the sources of the Senegal and Niger, ed. T. E. Bowdich (London, 1820; reprinted 1967), 8; F. De Lanoye, Le Niger et les explorations de l'Afrique Centrale depuis Mungo-Park jusqu'au Docteur Barth (Paris, 1858), 15; Josef Chavanne, Die Sahara oder von Oase zu Oase (Wien, 1879), 28; Basset 1888, 56. Further support for the negative image was gained from Arabic literature where the Tuaregs were treated in no better (for a critical Arab view of the Tuareg, see Muhammad bin Uthman al-Hashashi, Voyage au pays des Senoussia à travers la Tripolitaine et les pays Touaregs, tr. Serrer & Lasram, Paris 1912, 177-84). [116] For example Hanns Vischer, who had crossed the Sahara in 1906, complained that Duveyrier's high opinion of the Tuaregs "has been the cause of many disasters to Europeans, who did not know that under his outward dignity the Tuarek hid the most treacherous character" (Across the Sahara, London 1910, 164-65). [117] For a record of Tuareg conflict with
French colonialists in the southern Sahara, see for
example A. Hacquart, Monographie de Tombouctou (Paris,
1900), passim. [118] Dozy 1861, IV, 252, 266; see Neville Barbour, Morocco (London 1965), 58-9. [119] "Arabialaiset ja islamin aikaisemmat vaiheet", in J. W. Ruuth et al. (eds.), Maailman historia, vol. II (Helsinki 1917), 583; see also G. Faure-Biguet, Histoire de l'Afrique septentrionale sous la domination musulmane (Paris 1905), 126. [120] On the use of Hamitic hypothesis in European literature, see Edith Sanders, "The Hamitic Hypothesis: Its Origin and Functions in Time Perspective", Journal of African History, 10 (1969), 521-32. [121] 1899, 35. [122] Philip de Barros, "Changing
Paradigms, Goals & Methods in the Archaeology of
Francophone West Africa", in Peter Robertshaw (ed.),
A History of African Archaeology (London 1990),
160; Seppo Rytkönen, "Katsaus nykyiseen Afrikan
historiantutkimukseen", Historiallinen
Aikakauskirja, 87 (1989), 6; Pekka Masonen, "Länsi-Afrikka
ja Välimeren maailma ennen löytöretkiä", in
Tuomo Melasuo (ed.), Vieras Välimeri. Kulttuurien ja
politiikan kohtauspaikka (Tampere 1994), 37-38. [123] Delafosse 1912, II, 22. Other favoured
candidates were for instance Carthaginians, Garamantes,
Egyptians, and Persians. Common to all writers was,
however, steadfast faith in white founders of Ghana.
Evidence for this belief was gained both from Sudanese
oral tradition, which attributes white ancestry to many
islamized dynasties, including the Sise of Ghana (see
Fatimata Mounkaïla, "Ancestors from the East in
Sahelo-Sudanese Myth: Dinga Sonink‚, Zabarkâne
Zarma, and Others", Research in African
Literatures, 24, 1993, 13-21), and from al-Sadi who
claims that the forty-four rulers of Kayamagha,
customarily identified with Ghana, were whites, though
their origin was unknown to him (Tarikh es-Soudan,
French tr. 18; Arabic text, 9). Besides the whites, there
was one African candidate, the Fulani. Yet the Fulani
were believed to have a Semitic ancestry, and thus they
were not pure blacks but a more advanced Hamitic people.
For European opinions of the Fulani, see for example
Mollien 1820/1967, 157 and 164; Oskar Lenz, Timbuktu.
Reise durch Marokko, die Sahara und den Sudan (Leipzig
1884), II, 258-59 and 266; Le Chatelier 1899, 107-10 and
123. [124] See for example Wappäus 1841/1966, 45; De Lanoye 1858, 118 and 122. [125] 1799, 159. Furthermore Park claimed that the Moors ruled all the areas reaching from the Senegal to the borders of Abyssinia (pp. 112-13). [126] Travels through Central Africa to
Timbuctoo; and across the Great Desert to Morocco,
performed in the years 1824-1828 (London, 1830;
reprinted 1968), II, 65. The French original (Journal
d'un voyage à Tombouctou et à Jenne dans l'Afrique
centrale...) was published in Paris in the same year. [127] See also Mollien 1820/1967, 2-4; Lyon 1821, 112. [128] Eduard Conte, "Herders, Hunters and Smiths: Mobile Populations in the History of Kanem", in J.G. Galaty and P. Bonte (eds.), Herders, Warriors and Traders. Pastoralism in Africa (Boulder 1991), 221-2; Also Lange 1994, 64; for Arabic sources, see for example Ibn Khaldun, The Muqaddimah, tr. Franz Rosenthal (New York 1958), I, 252-61; also Ibn Abi Zar, in the Corpus, 236-37. [129] See Ibn Khaldun, in Corpus, 331; Richard Jobson, The Golden Trade (London, 1623; reprinted 1933), 48; for an historical perspective of the nomad/sedendary relationship in Western Sudan, see Nehemia Levtzion, "Berber Nomads and Sudanese States: the Historiography of the Desert-Sahel Interface", in Islam in West Africa: Religion, Society, and Politics to 1800 (Aldershot, 1994). [130] 1888, 49-50. [131] See for example Jean Bayol, Voyage en Sénégambie (Paris 1888), 80 and 90-1; Le Chatelier 1899, 34-5; Jules Brévié, Islamisme contre "naturisme" au Soudan Français (Paris, 1923), 229-30. [132] Y.U.Zotova, "Dva podhoda k
izucheniyu istorii Afriki", Narodni Azii i
Afriki, 27 (1987), 103. [133] The Austrian explorer Oskar Lenz, for example, described the condition of Morocco in 1879 thus: "The Moroccan people lives in a semi-cultured condition, which approximately correspond with our Middle Ages ... Regarding the present situation, Islam is identical with regression and barbarism, while the Christian powers represent civilized life and progress." (1884, I, 430). [134] Eric Stokes, The Peasant Armed: the
Indian Rebellion of 1857 (Oxford 1986), 7-9; A. S.
Kanya-Forstner, The Conquest of the Western Sudan
(Cambridge 1969), 66-67; Knut S. Vikør, "Mystics in
the Desert: the Sanusiya and the Sahara", in H.
Palva & K. S. Vikør (eds.), The Middle East -
Unity and Disunity. Papers from the Second Nordic
Conference on Middle Eastern Studies, Copenhagen 22.-25.
October 1992 (Bergen, 1993), 144-5. [135] Kanya-Forstner 1969, 20; Christopher Harrison, France and Islam in West Africa 1860-1960 (Cambridge 1988), 12-14; see also Brévié 1923, 164-66. [136] See Brévié 1923, 147-48. [137] Kanya-Forstner 1969, 149 and 195-98; Harrison 1988, 50-51; Van Hoven 1990, 186-87; see Basset 1888, 49-50; Binger 1892, II, 345; Brévié 1923, 234. [138] Jean Suret-Canale, Essays on African
History from the Slave Trade to Neocolonialism (London,
1988), 179. [139] C-A. Walckenaer, Recherches géographiques sur l'interieur de l'Afrique septentrionale (Paris, 1821), 11-12; Lenz 1884, II, 266; Brévié 1923, 117 and 297. Richard Burton, though he wrote often as much to shock as to inform, puts this point of view persuasively: "I would record my sincere conviction that El-Islam has wrought immense good in Africa; that it has taught the African to make that first step in moral progress, which costs so much to barbarous nature; and it thus prepares him for a steady onward career, as far as his faculties can endure improvement. What other nation, what other faith, can boast that it has worked even the smallest portion of the enduring good done, and still doing, to Africa by El Islam? Granting that ill temper, polygamy, domestic slavery, and the degration of women are evils; yet what are they to be compared with the horrors of cannibalism and fetishism, the witch tortures, the poison ordeals, and legal incest, the 'customs', and the murders of albinos, of twins, of children who cut their upper teeth first, and of men splashed by crocodiles? Surely the force of prejudice cannot go beyond this!" (Wanderings in West Africa from Liverpool to Fernando Po, London 1863, 180-81). [140] Brévié 1923, 231-34; Kanya-Forstner 1969, 137; Harrison 1988, 103; see Maurice Delafosse, "L'état actuel de l'islam dans l'Afrique Occidentale Française", Revue du monde musulman, 9 (1910), 32-53. [141] de Barros 1990, 162. [142] See the Preface in John W. Blake, West Africa. Quest for God and Gold (London 1977); also Ernst Opgenoorth, "Historians and Written Sources: General Problems", Paideuma, 33 (1987), 107-14; and Finn Fuglestad, "The Trevor-Roper Trap or the Imperialism of History. An Essay", History in Africa, 19 (1992), 309-26. [143] Lansiné Kaba, "Histoire africaine
et idéologie", Afrika Zamani, no.2 (1974),
9-10; see also A. E. Afigbo's chapter "Colonial
Historiography", in Toyin Falola (ed.), African
historiography. Essays in honour of Jacob Ade Ajayi
(Harlow 1993), 39-51. [144] "The Almoravid movement in the western Sudan. A review of the evidence", Transactions of the Historical Society of Ghana, vol. 7 (1964). In the same year, Vincent Monteil (L'Islam Noir, Paris 1964, 58-62) pleaded for the emancipation of the history of Ghana, insofar as it is discernible at all, from the unsubstantiated hypotheses of Delafosse - but Monteil gives little attention to the specific conquest question, despite being cautious about it. His second edition (1980, 81-6) changes nothing from 1964, save to add a carbon-14 date. [145] A previously unknown part of a
manuscript concerning the Almoravids, written by Ibn
Idhari in early 14th century, became familiar to western
scholars in the mid-1950s, although the Arabic text was
not published until 1961 by Ambrosio Huici Miranda
("Un fragmento inédito de Ibn Idhari sobre los
Almorávides", Hespéris-Tamuda, 2). The
author's full name is Abu'l-Abbas Ahmad b. Muhammad Ibn
Idhari, and the full title of his work Al-Bayan al-mughrib
fi akhbar al-andalus wa'l-maghrib (The Amazing
Exposition on the History of al-Andalus and the Maghrib).
A French translation of Bayan by E. Fagnan (Histoire
de l'Afrique et de l'Espagne intitulée al-Bayano'l-Mogrib,
2 vols, Algiers) had been published in 1901-4, but it
lacks the important account of the Almoravids. [146] Relying on the authority of Ibn Abi Zar and Ibn Khaldun, western scholars had believed that Marrakesh was established by the Almoravids in 454/1062 - although it is unnoticed by al-Bakri, writing some six years later! According to Ibn Idhari, it took place on 23 Rajab 462/8 May 1070. See E. Levi-Provençal, "La fondation de Marrakesh (462-1072)", in Mélanges d'histoire et d'archéologie de l'occident musulman. Hommage á Georges Marçais (Algiers 1957), vol. II, 117-20. [147] Caroline Neale, Writing "independent"
history. African historiography, 1960-1980 (Westport,
Conn. 1985), 3-4. [148] Semonin 1964, 42, 59. Semonin (42n.2) derived this idea from Huici Miranda who had expressed his doubts on Ibn Abi Zarc's reliability in this matter already in 1961 (see "El Rawd al-Qirtas y los Almorávides", Hespéris-Tamuda, I, 523-24). See also Monteil 1964, 62: "Some years later the Almoravids islamized the land of Ghana, in 1076 (469 AH), a date given by one author only, az-Zuhri." [149] "The Almoravids: some questions concerning the character of the movement during its periods of closest contact with the western Sudan", Bulletin de l'I.F.A.N, s‚r. B, vol. 29, 848-49. [150] "Routes de commerce et échanges en Afrique occidental en relation avec la Méditerranée: un essai sur le commerce africain médiéval du XIe au XVI siècle", Revue d'histoire économique et sociale, vol.50, 57. See also "La question d'Audagust", in D. Robert, S. Robert & J. Devisse (eds.), Tegdaoust I. Recherches sur Aoudaghost (Paris 1970), 153. [151] Saharan myth and saga (Oxford),
p.108 and n.2. [152] "L'Attaque de Ghana (XIe siècle)", Afrika Zamani, no. 2 (1974), 55-77. [153] Ibid., 57-8. Coulibaly used fragments of al-Zuhri (Arabic text with French translation) published in Youssouf Kamal's Monumenta Cartographica Africae et Aegypti (vol. III, fasc. 3, Cairo 1933, pp.801-3), although a more complete version had been published by Mahammad Hadj-Sadok in 1968, in Bulletin d'études orientales, vol. 21. Yet in Hadj-Sadok's edition the date for the conversion is 496, instead of 469. [154] 1974, 60-61. [155] Ibid., 67-69. The idea of Almoravid ribat as an island fortress originates from Ibn Abi Zar and Ibn Khaldun (see Corpus, 239 and 329). It was incorporated within Almoravid historiography at a very early phase (see for example Godard 1860, 308; Farias 1967, 821-34; also Humphrey J. Fisher, "What's in a name? The Almoravids of the eleventh century in the Western Sahara", Journal of Religion in Africa, 22, 1992, 290-317). [156] Coulibaly 1974, 73. [157] Ibid., 60-61 and 73. Actually
some traces of destruction have been found in Kumbi Saleh
which are dated in the 11th century; but conclusive
proofs are still lacking. We do not even know whether the
ruins of Kumbi Saleh are really the city of Ghana
described by al-Bakri (see I. Hrbek & J. Devisse,
"The Almoravids", in Unesco General History
of Africa, vol. III, London 1988, 359). [158] "Great states revisited", Journal of African history, 15 (1971), 3; see particularly 482n.22, and 484-85. [159] Ibid., 480-81. Farias' criticism in this respect certainly corresponds with the contemporary shift in the research of nomad/sedentary relations in the Sahel. Farias cites Claude Meillassoux (The Development of Indigenous Trade and Markets in West Africa, London 1971) but similar thoughts were expressed by P.E. Lovejoy and Stephen Baier in 1975 ("The Desert-side Economy of the Central Sudan", International Journal of African Historical Studies, 1975, 551-81), and Charles Frantz ("Shifts in Power from Nomads to Sedentaries in the Central Sudanic Zone", in Y. Fadl Hassan & P. Doornbos (eds.), The Central Bilad al-Sudan. Tradition and Adaptation, Sudanese Library Series no. 11, Khartum 1979, 171-91). On the other hand, we cannot deny that there have been conflicts between the Saharan nomads and Sahelian agriculturalists, although it is not plausible to suppose that the nomads had always been the stronger side. These conflicts have, however, a political and economic focus, rather than religious motives (see Nigel Cross & Rhiannon Barker, eds., At the Desert's Edge. Oral Histories from the Sahel, London n.d. [1991?], 56, 63, 68, 144 and 152; also James L. A. Webb Jr., Desert Frontier. Ecological and Economic Change along the Western Sahel 1600-1850, Madison 1995, 22-26). [160] Levtzion 1973, 47. [161] Farias 1974, 484. [162] Ibid, 480-81. [163] "Gao and the Almoravids: a
Hypothesis", in B. Swartz Jr and R. Dumett (eds.), West
African Cultural Dynamics (The Hague, 1980), 420. [165] McCall, 1971, 14; Norris, The
Berbers in Arabic Literature (London and Beirut 1982),
133-34; lange, "Les rois de Gao-San‚ et les
Almoravides", Journal of African History,
32 (1991), 275. [166] See Corpus, 231; also Hanna E. Kassis, "Observations on the First Three Decades of the Almoravid Dynasty (A.H.450-480 = A.D. 1058-1088). A Numismatic study", Der Islam, 62 (1985), 318. [167] Lange 1994, 66. [168] Journal of African History, 23 (1982), 549-60; Part I: "The external Arabic sources", and II: "The local oral sources", History in Africa, 9 (1982), 19-59, and 10 (1983), 53-78. [169] Conrad and Fisher 1982, 45. [170] Supporters of the new view include Ann
McDougall ("The Sahara reconsidered: pastoralism,
politics and salt from the ninth through the twelfth
centuries", African Economic History, no.
12, 1983, pp.273-4), Michael Brett ("Islam and trade
in the bilad al-sudan, tenth-eleventh century A.D.",
Journal of African History, 24 (1983), p.439),
and David Henige (1987, pp.65-69). Those who still
hesitate include Peter B. Clarke (West Africa and
Islam, London 1982, pp.17-22), and the late Mervyn
Hiskett (The development of Islam in West Africa,
London 1984, 23, 25-26, 302-3). Among the apparently
unpersuaded are John O. Hunwick (Sharia in Songhay:
the replies of al-Maghili to the questions of Askiya al-Hajj
Muhammad, Oxford 1985, p.15 n.4), Nehemia Levtzion (see
his chapter "The early states of the western Sudan
to 1500", in J.F.A. Ajayi and M. Crowder, eds., History
of West Africa, vol. I, 3rd edition, London 1985),
Vincent Lagardère (Les Almoravides, Paris 1989,
pp.89-90, which still repeats Delafosse almost verbatim),
and Roland Olivier (The African experience: major
themes in African history from the earliest times to the
present, New York 1991, 87-88). The ambivalent
situation is well visible in UNESCO General history
of Africa, vol. 3 (London 1988): see especially pp.359-60
and 460. [171] History in Africa, 19 (1992), 103-31. [172] Ibid, 105 and 120. [173] Ibid, 104 and 103. [174] See for example Philip Curtin, Steven
Feierman, Leonard Thompson & Jan Vansina, African
History from Earliest Times to Independence, 2nd
edition (London and New York 1995), p.94; John D. Fage, A
History of Africa, 3rd edition (London 1995), 71;
John Iliffe, Africans: The History of a Continent
(Cambridge 1995), 53; Kevin Shillington, History of
Africa, Revised edition (London and Basingstoke,
1995), 86-87. [175] Self Portraits, tr. Ralph F.
McCarthy (Tokyo: Kodansha International 1992), 71. |
||