A major theme in the historiography of the Rift Valley region of east Africa has been the series of raids and wars during the nineteenth century between groups of Maa-speaking peoples who dominated the plains from northern Kenya to central Tanzania. Since the 1840s European and African observers have tended to divide the combatants into two factions, usually called the Maasai on one hand and the Iloikop (or Kwavi) on the other. Since the 1880s European administrators and western scholars have tended to designate the groups they have called Maasai as “pastoralists” or sometimes “pure pastoralists” and the groups they have called Iloikop/Kwavi as “agriculturalists” or “semi-pastoralists.” According to this interpretation, the “Iloikop Wars” or the “Wars between the Maasai and the Iloikop” of the nineteenth century pitted agricultural Maa-speakers against pastoral Maa-speakers. In surveying the relevant literature and in analyzing the European descriptions in light of explanations of my Maasai informants, it became clear that this orthodox dichotomy rests on a mistakenly static perception of socio-economic groups and denies the precariousness of pastoral life in the Rift Valley. Scholarly acceptance of the Maasai-Iloikop (Kwavi) dichotomy as the basis of interpretation of nineteenth-century Maasai history has resulted in a serious distortion of that history and an avoidance of more complex and important issues. In this paper I will review the literature on the “identities” of the Maa-speaking peoples -- identities attributed to them by outside observers -- and subject those interpretations to the perceptions and explanations of the Maa-speaking peoples themselves.

The Rwandan revolution of 1959-1962 marked an important watershed not only for the history of the country but also for its historiography. Within Rwandan historical studies these political changes encouraged the development of a more broadly-based analysis, one which went beyond the earlier tendency to focus on the Nyiginya royal court. The effects of this new historiography were not limited to Rwanda alone, however, since historical perceptions ultimately derived from Rwandan studies had dominated much of the earlier research on pre-colonial history both east and west of the lake. Combined with the emergence of new analytic assumptions elsewhere in Africa, the shift in Rwanda during the 1960s therefore freed studies west of the lake from the constraints of a particularly sterile historical model and opened the way for the initiatives of a new generation of Zairean historians conducting research there in the 1970s. This changing context of research initiatives and the continuing process of reassessment within Rwanda has also aligned perceptions of Rwandan history more closely with those of other areas. We can look forward in the future to more fruitful regional historical perspectives which transcend present political boundaries; it may well be that Rwandan research of the 1980s will draw increasingly on the concepts and conclusions of Zairean research of the 1970s.

Even while this new approach has perforated political boundaries on the ground, it has also dissolved the earlier rigid disciplinary boundaries of research. The major influence in this new approach came from the work of anthropologists (or historians with anthropological training).

The earliest European accounts of the Lower Guinea coastlands are fewer and less informative than those available for the coastlands of Upper Guinea. This is not surprising. The exploration of the 900 or so miles of the Guinea coast to about as far as Cape Mesurado was a more deliberate process, over some eighteen years, than was that of the nearly 2000 miles of coast between Cape Mesurado and Cameroun, which seems to have been undertaken essentially in the five years, 1471/75. It was also probably a more open process, involving sailors and merchants of many nations besides the Portuguese, men who were full of a Renaissance wonder at what they saw and keen to communicate their new knowledge. Some of these men or their followers were soon – certainly by about 1500 – residing more or less permanently in African communities on the Upper Guinea coasts, subject to little or no effective control from the Portuguese authorities, becoming lançados. On the other hand, by the later 1470s the discovery of the gold wealth of Mina had led the Portuguese crown to seek to establish a royal monopoly over sea trade with Lower Guinea, and to confine it to a few posts over which it sought to assert its direct control.

The first comprehensive account of the Upper Guinea coasts to have survived is to be found in the Esmeraldo de Situ Orbis of Duarte Pacheco Pereira, almost certainly written between the years 1505 and 1508. It is constructive to compare this work with its contemporary, the Description of the West Coast of Africa written by Valentim Fernandes, probably in 1506 or 1507, which describes the Guinea coast only as far as Cape Mount (although it also has fascinating accounts of the islands of São Tomé and Annobón in the Gulf of Guinea).

The Bantu expansion is one of the most important large-scale problems in African culture history -- an epic enacted over two or three thousand years and ten million square kilometers, by a cast not merely of thousands, but of many millions. By definition, the problem is primarily linguistic, but it cannot fail to engage the interest of other Africanists. The evidence arising from the comparative study of the Bantu languages has to be collated with evidence derived from other sources -- especially from archeology -- and extra-linguistic factors have to be invoked as soon as we raise the question of explanation. Bantu-speaking communities did not expand by virtue of the fact that they spoke Bantu: this at least we may safely take for granted.

Until a few years ago, the argument revolved around the names of two linguists -- about as different in temperament and training as any two linguists could be -- who had both by chance turned their attention to Bantu in the late 1940s. An American linguist, Joseph H. Greenberg, working towards a genetic classification for all African languages, arrived at a controversial conclusion regarding the relationship between Bantu and the so-called “Semi-Bantu” languages of Nigeria.

In 1964 the Spanish archivist and scholar, D. Vicenta Cortés, published summaries of a large number of entries in the Crown records of Valencia relating to the arrival of slaves between 1479 and 1516. Although about two thirds of the entries relate to lighter-skinned slaves - Guanche from the Canary Islands, Muslims and Jews from North Africa and other parts of Iberia, “blancos” apparently from the eastern Mediterranean - about one-third relate to black slaves or “negros.” Cortés described the legal and administrative context of the records but did not attempt a systematic analysis of the entries. However, she later usefully examined those relating to “negros” and offered a preliminary identification of stated provenances. In the present paper I consider the data on some 3,000 negros of stated Black African provenance, paying special attention to some 260 individuals whose provenance was given in terms of an ethnonym or narrowly located toponym and some 150 whose personal name in an African language was recorded. I note many instances where the Valencian data provide the earliest recorded documentation of an African ethnonym, while the anthroponyms are tabulated for study by field linguists.

As summarized and published by Cortés, the entries refer to some 5,400 Black Africans - the total of negros, apart from a handful stated to be from India or Brazil. This figure gives only a rough indication of the extent of the trade in Black Africans at Valencia during the period to which these entries apply, 1482-1516. About one hundred of the Africans were obtained, not more or less directly from Black Africa and normally via Lisbon, but from localities in North Africa or Iberia where it is likely that they had spent the greater part, if not all, of their lives.

Holders of Established Chairs occupy seats of some antiquity and renown, and on the occasion of their inaugural they are expected to indicate how they propose to maintain or alter the course charted by their distinguished predecessors. Holders of Personal Chairs escape this formidable task. On the other hand, they face a question which is appropriately personal: namely, why they come to be seated at all. An audience less charitable than this one, which has been hand-picked as far as possible for its generosity as well as for its wisdom, might wish to pursue this question with enthusiasm. Fortunately, a further safeguard against being unseated at the moment of investiture exists in the admirable tradition whereby I am allowed to provide my own answer without hindrance or interrogation. I have chosen to explain my presence here today in the manner commonly adopted by those who find themselves elevated beyond their customary station; that is by creating a tradition to authenticate my ancestry and hence to validate my title.

“There is a story that the Prophet appeared to Abdu Rahaman in a dream and said to him, ‘Get up and go west and establish Islam.’ Abdu Rahaman got up and took a handful of the soil of Medina and put it in a cloth and brought it to Hausaland. Whenever he came to a town, he took a handful of the soil of the country and put it beside that of Medina. If they did not correspond, he passed that town. So he journeyed until he came to Kano. And when he compared the soil of Kano with Medina soil they resembled one another and became as one soil. So he said, ‘this is the country that I saw in my dream.’” [xx]

I wish in this paper to treat the Kano Chronicle (henceforth KC) as a document of intellectual history, and not just as a mine from which to dig valuable ‘facts.’ The aspect of intellectual history I will discuss is the meaning of historical metaphors - or analogical geography - of which the above story is a rather special example. But first I will try and show that the first ‘edition’ of KC was completed in the mid-seventeenth century and was compiled from materials which had been developed since the fifteenth and sixteenth centuries - thus locating the intellectual history in a specific period. The texts used are discussed in the appendix.

In writing this essay I am treading where many have trod before. Abdullahi Smith's work on the Sayfawa and on the origin of the Hausa states and Mervyn Hiskett's publications on the Kano Chronicle and on the Song of Bagauda are the most notable examples.

Historians recognize that the perception and organization of time are fundamental to the internal ordering of all human cultures. However, the history of pre-colonial Africa has largely been written to conform to the calendrical rhythms of an imposed European chronology. Regret over this discrepancy has been tempered by recognition of the very real problems involved in rectifying it. Chief among these is the fact that linear chronology per se -- in purist interpretation requiring numbered series derived from a fixed base, and therefore the mnemonic support of a graphic record -- is beyond the technological competence of any pre-literate society. However, the inability to maintain chronologically precise memorials of the past by no means precludes the existence of sophisticated mechanisms for ordering and dividing temps courant. That is, a historical sense disordered or dissolved through the agency of unassisted, and thus all too fallible, human memory may happily co-exist with an exact (and often symbolically charged) calendrical time. Broadly speaking, the foregoing was the situation obtaining-in Asante in the nineteenth century. Time in nineteenth-century Asante, in a number of its aspects, is the subject of this paper.

The establishment of a historical chronology in nineteenth-century Asante was severely inhibited, and in ultimate terms negated, by the absence of a literate culture. It is the case that rare and isolated individuals like oheneba Owusu Ansa (ca. 1822-1884) and oheneba Kwasi Boakye (1827-1904) acquired in foreign exile skills in European languages. However, Akan Twi was not effectively reduced to writing until the mid-nineteenth century, and then not in Asante. Thus, at the time of the British usurpation in 1896 Asante was still a pre-literate culture.

The fall of the Roman empire and the rise of its successor states in continental Europe is reasonably well-documented and royal pedigrees are both brief and among the least of our sources of information. In the islands Roman administration ceased in Lowland Britain in 410, and had never obtained in northern Highland Britain or in Ireland, so that continuing and successor states existed side by side. It is likely that their secular influences on one another are masked by the common process of the conversion to Christianity in the fifth, sixth, and seventh centuries; certainly the “waves of saints” are prominent both in accounts of the period and in present notions about it. This is no doubt one reason why the various bodies of royal pedigrees have remained until recently (or still) only partially published and little studied.

The matrilinear Picts, however, became a subject for historical and anthropological consultation a century ago. Unfortunately, as we shall see, this conversation has not continued, and it is only of recent years that historians of the Insular Dark Ages have once more become aware of the possibilities of learning (both positively and negatively) from the insights and procedures developed by anthropology. The present moment, when both disciplines are grappling with the problem of what constitutes historical evidence for pre-archival periods, may be a propitious time for recommencing joint consideration of our common interests.

The foundation for the argument that follows is the demonstration in Part I of this essay of Ilesha's steady physical expansion over at least three hundred years, as manifested in the establishment of new quarters. Behind this lay Ilesha's success in exploiting the geo-political possibilities of her situation midway between Oyo and Benin, twenty-five miles back from the savanna/forest divide and so safe from Oyo's cavalry. Other aspects of Ilesha's growth as a capital were her loose hegemony, periodically reasserted, over the smaller neighboring kingdoms to the north and east; her effective continuous domination of other communities (including some earlier centers of the Ijesha Kingdom) within a 20 to 30 mile radius; and her establishment of further rural out-settlements in this area, many of them, as has been mentioned, ruled by members of her royal lineage. In what follows, we will be less concerned with the interlocking means by which this increase was brought about - slaves taken from communities defeated in war and incorporated, revenues from the trade drawn to the markets and routes which Ilesha was able to dominate, free immigrants attracted from other towns - than with how this growth was managed and affected the structure of the community. Our evidence takes two principle forms; itan told about many of the Owa, such as formed the raw material for the Itan Ilesa of Abiola, Babafemi, and Ataiyero; and the system of chiefly titles, each with distinctive attributes and traditions, which defines the political structure of the community.

Tribes no longer exist, at least in the writings of Africanists, but they tend to be replaced by very similar looking entities, going under a variety of aliases. Perhaps purely as a result of laziness of thought, there remains a tendency of write of the Zulu, the Tswana, or the whomever, and not to probe the assumption that these units have some actual existence. Nevertheless, like all working assumptions, this has to be continually re-examined. In this paper, therefore, I return to one of the classic problems in southern African history (in the broadest sense), namely that of the relationship between the Khoisan and the Bantu-speaking Africans. It has of course been widely studied, especially by linguists and physical anthropologists, and, to a lesser extent by archeologists. These disciplines show complementary results. Linguistics shows that many loan-words crossed from Khoisan to the Bantu languages - though not, apparently, in the reverse direction - and that this occurred to its greatest extent in the various Nguni languages of the south-east coastal belt. Physical anthropology has shown that the proportion of Khoisan genes in the Bantu speaking population increases from north to south. The archeological evidence is less clear. For eastern Zambia, Phillipson has argued that “the Early Iron Age folk and their late stone age neighbours… can be shown to have existed in several regions” and that “throughout the first millenium the two populations clearly to a large extent maintained their own separate identities”. This situation seems to have lasted until “four or five centuries ago.”

George McCall Theal, most prolific of South African historians, deserves a modern biography. In the few, very inadequate, accounts of his life that are available there is no mention of his work as a labour agent in the western Cape. Bosman mentioned that he visited Stellenbosch and Tulbagh in 1878, but did not say what he was doing there. Babrow admits that she did not know what his first formal civil service post was. Immelman misleadingly suggests that it was a post in the Colonial Treasurer's Department, when in fact he did not join that department until March 1879.

This ignorance is not altogether surprising for, while Theal himself in his later published work and in the evidence he gave to parliamentary select committees in 1895 and 1906 provided a fair amount of information about his career before May 1878 and after March 1879, he did not say what he had done between those dates, almost as if he did not wish to remember it. In his History of the Boers in South Africa, indeed, he wrote that “when the war was over [referring to the Cape-Xhosa war, which ended in May 1878] I asked for and obtained the charge of the Colonial Archives preserved in Cape Town,” which is totally inaccurate. In the relevant volume in his History of South Africa, he mentioned that he was asked to superintend Oba's Xhosa in the Victoria East district in December 1877, and later that he had served as special magistrate at Tamacha in the King William's Town district in 1881, but there are no other references to his own career.

[This continues the comparison of texts of the English and Dutch versions of Bosman. For earlier instalments see History in Africa 2(1975), 185-216; 3(1976), 91-126; 4(1977), 247-273; 5(1978), 225-256; 6(1979), 265-85. Procedural matters are discussed in the first instalment, to which the reader is referred.]

Interest in the question of Bantu expansion rose dramatically in the 1950s as historians, archeologists, and anthropologists all joined in the fray. This reflected both the rise of Africa in world affairs and the expansion of research in general. The scholars involved were typically a new breed of professionals, and as such more dependent than their predecessors on universities or research institutions. The School of Oriental and African Studies in London achieved overwhelming dominance from about 1950 until the late 1960s, so that opinions held by its staff found the widest audience. The new scholars also were, for the most part, anti-racist, sympathetic to African nationalisms, and of liberal or socialist persuasion. They tended to reject the notion of “conquest,” believing in gradual change rather than abrupt cataclysmic mutation, perhaps because they were repelled by their recent experiences during the war. As had happened earlier, these extraneous circumstances left a deep imprint on the speculations that were now proposed. Early in this period a new paradigm almost achieved consensus, but after 1968 this fell apart and during the last decade two new trends have appeared: the single-minded quest for a new paradigm and the search for better understanding through the study of analogous processes, coupled with a more radical skepticism.

Poor William Smith must be squirming in his grave. After 250 years, his New Voyage to Guinea has been judged a fraud. Reviewing the reprint edition published in 1967, H.M. Feinberg writes that

the reprint publisher wasted its money, and the African history community is no more enlightened because of their effort. A fraud, however ancient, has been uncovered, and another old, “classic” description can now be struck from bibliographies of “valuable” works.

Feinberg gives a number of examples of how Smith plagiarized from Bosnian's New and Accurate Description of Guinea, published nearly forty years earlier. Each of the passages he quotes is evidently a rephrased version of a passage in Bosman. He makes no mention, however, of the many occasions when Smith has something new to say. Smith's personal observations on Conny's Castle, Dixcove, Sekondi, Komenda, Cape Coast, Elmina, Tantumquerry, Winneba and Accra are passed over in silence. In fact Feinberg himself has elsewhere cited Smith as an independent source on Elmina.

Probably Feinberg is most unfair to Smith by concentrating only on Smith's account of Gold Coast affairs. While this is admittedly Feinberg's area of expertise, it is precisely in areas of the coast to the windward of the Gold Coast that Smith is most valuable and least derivative. Here, too, some material is taken from Bosman -- for instance, on Cape Mount, Cape Mesurado and the “Quaqua Coast”. But almost everything else records Smith's own impressions of events which he himself witnessed. Among the most valuable passages are eight pages on the River Sierra Leone and a 48-page narrative of events at Sherbro, where the Royal African Company was losing its foothold on York Island.

The papers of James Augustus Grant, until recently in family ownership and inaccessible to most scholars, were sold at Sotheby's in London on 13 March 1979 and purchased (with the exception of photographs and annotated printed books) by the National Library of Scotland. They are now available for consultation in Edinburgh by registered readers; microfilm of selected portions can be supplied, also at the moment photocopies of the unbound letters (which will be bound in due course). They have been assigned the numbers 17901-26 in the National Library's sequence of manuscripts, and a description follows; an index is also available at the National Library, and this together with the description will appear in one of the future volumes of the Library's Catalogue of Manuscripts acquired since 1925. Of particular interest to African historians are the letters of explorers and others in MSS. 17909-10, of which those of Gordon, Kirk, Murchison, Speke, and Stanley are especially important; the African journal, MS. 17915, and sketches, MSS. 17919-21, both extremely significant not only for the Grant-Speke expedition, but also for the culture and history of East Africa; and (from a later period) the papers of Grant's son in South and Central Africa, MSS. 17907 and 17918.

As foundation money for overseas research grows more difficult to obtain, historians of Africa will perforce seek archival resources closer to home. A rough listing of African materials in American archives has been published, the Handbook of American Resources for African Studies, but the catalog evidently relied in part on reports written by American archivists who had little or no training in African history. As a result many available sources have been inadequately described. Take the case of the Southern Historical Collection, a repository for private manuscripts at the University of North Carolina Library at Chapel Hill. Although several collections with African papers in this repository are briefly noted in the Handbook, there was no indication of important materials like the forty-page eyewitness description of the court of the Mijjerteyn Sultan written in 1878. Nor, for that matter, is this account noted in the unpublished description of the papers available at the Collection; in fact, even if a researcher should ask, the archivist probably could not readily locate the account unless the researcher already had a name and date for it.

The Southern Historical Collection holds three kinds of documents relating to Africa: (1) The Khedive of Egypt hired several former Confederate officers to conduct mapping expeditions in Sudan, Ethiopia, and Somalia when he began his military conquests to the south in the 1870s. These officers' papers consist of letters, diaries, and printed material concerning their explorations and their daily lives in Cairo. (2) Several groups of papers contain information on missionary activities, mostly by Episcopalians, in Liberia from 1829 to 1880, and (3)there are a number of scattered items in various collections, mainly travelers' accounts of brief visits to Africa.

In 1846 Pope Gregory XVI established the Apostolic Vicariate of Central Africa, aiming at converting Africans to Christianity, bringing Christian assistance to European traders and officials in Africa, and combating the slave trade. The participation of the Verona Fathers dates from 1872, when Propaganda Fide authorized the Institute for African Missions of Verona to direct the mission to Central Africa and appointed Daniele Comboni as Pro-Vicar Apostolic.

Conditions in Africa, official mission goals, and Comboni's own policy shaped the observations and reactions of the missionaries to their African surroundings. Difficulties of supply and transportation, as well as their self-imposed obligation to minister to the Europeans, forced the early missionaries to associate with the motley assortment of local traders, especially along the White Nile and in Khartoum. Keen observations and sharp indictments of much of European activity, particularly the slave trade, resulted. Early in his career, Comboni had formulated the concept of converting Africa by using Africans as proselytizers. In their work in the mission field, Comboni and his followers generally reacted to African societies neither by ignoring them nor by objectively collecting ethnographies, but rather by evaluating their possibilities and potential in providing the environment and personnel for mission activity.

Records concerning the activity of the Verona Fathers have been collected at Missionari Comboniani, via Luigi Lilio 80, CP. 10733, 00143 Rome. Most of the material is in Italian and German, with occasional English translations, and deals with the southern and central Sudan, Uganda, Ethiopia, and Egypt.

The archives of the Republic of Cabo Verde contain the official records of Portuguese rule up to 1975. I did not see any documents that dated from before the early eighteenth century. At independence in 1975 the jubilant crowd broke into the main administrative building in Praia, the capital, and threw the records into the street. The records of the Praia municipal administration and of the Instituto de Trabalho were also ransacked. Eventually the dispersed documents, together with several thousand volumes from the Praia Public Library, were gathered up and packed haphazardly into wooden crates. In 1978 a large municipal warehouse was allocated as an archives store. Shelves were installed and the documents were slowly disinterred from the crates, where they had accumulated thick layers of dust.

The earlier documents, about 1500 manuscript volumes of the eighteenth and earlier nineteenth centuries, have been put on shelves, but not in any systematic order. Many are defective or fragile. The documents from the later nineteenth century up to 1949 are preserved in about 1300 metal boxes. They too have been put on shelves, but not in chronological order (if only because many of the boxes have lost their labels). The documents from 1950 to 1975 were enclosed in cardboard file covers. Some have been put on shelves, others are stacked on the floor. None are in order. Many have come loose from their covers and have been tied up arbitrarily in bundles, along with documents from the municipal and Instituto de Trabalho archives. There are also many bundles of miscellaneous municipal records.

The first members of the Congregation of the Holy Ghost arrived in the Gabon estuary in 1844. Their activities were largely confined to the northern shore of the estuary and nearby Cape Esterias until 1878. When they began to establish posts on the coasts from Loango to Rio Muni and at various points in the Ogooué and N'Gounié valleys. The establishment of New Kamerun and the coming of the First World War delayed their penetration into the Fang areas of the northern interior until the 1920s. Prior to the early 1880s all of Gabon belonged to the Vicariate Apostolic of the Two Guineas with its bishop at Libreville. After that time the southern coasts were assigned to the new Vicariate Apostolic of Loango. Loango itself formed part of the colony of Gabon until 1918 when it was definitively attached to the Middle Congo.

The kinds of records deriving from the Spiritan presence in Gabon, which are housed in the mother house of the French province, include: (1) the Bulletin Général de la Congrégation, handwritten from 1857 to 1885 and then printed. The Bulletin summarizes the activities of each vicariate and each mission station, annually at first and then for periods of two to four years. (2) Annual and five-year reports to the Sacred Congregation for the Propagation of the Faith in Rome; annual reports to the Oeuvres de la Propagation de la Foi and the Oeuvre de la Sainte Enfance, French lay organizations which provided the bulk of the funds for missionary activities; reports to the superior-general of the congregation and later to the secretary for African missions.