On the eve of All Saints' Day on November 1, 1898, a Portuguese army officer, Henrique Augusto Dias de Carvalho, observed a colorful and noisy crowd of people wending through the streets of Bolama beginning the celebration of dia dos finados (All Souls' Day), which day of supplication for the faithful departed is observed by Christians on November 2.

Some scholars interested in ancient Ghana and Mali dare to sift relevant oral traditions of the Western Sudan in search of historical evidence, while others express doubts that these sources can contain any information of value to historians. A period markedly affected by this question is that which saw the disintegration of Ghana and the rise of Mali in the twelfth and thirteenth centuries. Despite historians' general acknowledgement of the pitfalls accompanying the use of oral tradition as a source of information, much of what we know, or would like to think we know about this era, has been drawn from the legend of Wagadu and from the Sunjata epic.

Clearly a large part of the material in these oral traditions is composed of the stuff of myth and folktale, and on the face of it the prospect of trying to glean historical information from them is not an encouraging one. But woven into the patchwork fabric of these narratives are infrequent threads bearing diminishing echoes of people and events of the distant past. Vague, inaccurate, and potentially misleading as they must be, these archaic fragments nevertheless merit whatever attention is necessary to interpret their significance, in the hope that they might yield some useful historical insights.

Any pretensions to historicity in the Wagadu tradition and in the Sunjata epic may be open to question because there is so little that can be verified. While the mythical quality of some elements in the texts is obvious, there are others that could have a historical basis but cannot be independently confirmed. The material consulted here is approached with the attitude that, given the rarity of firmly documented sources, historians cannot afford to ignore the possibility that there is some information worth distilling from the oral accounts of ancient Mali and the related Soninke era that preceded it.

Before the advent of colonialism, Arabic was widely used in northern Nigeria where Islam had penetrated before the fifteenth century. The jihād of the early nineteenth century in Hausaland led to the establishment of the Sokoto Caliphate, the revitalization of Islamic learning, and scholars who kept records in Arabic. Indeed, some local languages such as Hausa and Fulfulde were reduced to writing in Arabic scripts. Consequently, knowledge of Arabic is a crucial tool for the historian working on the history of the caliphate.

For Ilorin, a frontier emirate between Hausa and Yorubaland, a few Arabic materials are available as well for the reconstruction of the history of the emirate. One such document is the Ta'līf akhbār al-qurūn min umarā' bilad Ilūrin (“The History of the Emirs of Ilorin”). In 1965 Martin translated, edited, and published the Ta'līf in the Research Bulletin of the Centre for Arabic Documentation at the University of Ibadan as a “New Arabic History of Ilorin.” Since then many scholars have used the Ta'līf in their studies of Ilorin and Yoruba history. Recently Smith has affirmed that the Ta'līf has been relatively neglected. He attempts successfully to reconstruct the chronology of events in Yorubaland, using the Ta'līf along with the Ta'nis al-ahibba' fi dhikr unara' Gwandu mawa al-asfiya', an unpublished work of Dr. Junaid al-Bukhari, Wazīr of Sokoto, and works in English. The purpose of this paper is to analyze the information in the Ta'līf by comparing its evidence with that of other primary sources which deal with the history of Ilorin and Yorubaland.

Historians of Africa have often cast a jaundiced eye in the direction of written sources; their suspicions have been particularly grave with regard to colonial documents. They have seen them as, at best, the creations of ill-informed foreign observers or, at worst, as the deliberately self-serving justifications of the ruling elite. To counter the all-too-real deficiencies of these documents, historians have laid increasing emphasis on materials produced by Africans themsleves--oral accounts. Yet these records too are problematic, and their principal weakness lies in the fact that they are often collected decades after the events and circumstances the historian wishes to understand. Each of these kinds of sources thus presents methodological difficulties. This paper advocates the use of a third type of source, African court records, which are at once contemporary and yet created by Africans themselves. These records--while not without their own particular biases--permit one to overcome some of the problems inherent in official documents and oral interviews. Courts for Africans were established by colonial regimes throughout the continent. But, although Africans participated in varying degrees in their proceedings, the records have largely remained ignored by historians. In what follows I discuss the court records in two tribunals in Bujumbura, Burundi. The paper is divided into three sections: the first is a description of the city's African community and the courts; the second is a discussion of the court proceedings and the reliability and use of the records; and the last is a brief summary of some of the data the documents have yielded.

A great deal has been written about the colonial conquest of Africa, from the perspectives of both the conquerors and the conquered. Primary resistance has come and gone as an ‘in’ topic in African Studies. Yet to the extent that such literature deals with the colonial conquest, it has been within a structural-functional framework, focusing on social, political, and economic factors. Possible cultural and psychological aspects have been relegated to the occasional vague comment. More often, these latter elements have tended to be assumed, rather than demonstrated, and then generally in the hindsight of nationalist manipulation of oral traditions in the process of decolonization. Only through the elucidation of meaning to participants of events can we transform them from the status of ‘objects of study’ to ‘subjects in action.’

This paper examines the impact of colonial conquest of the Eggon of central Nigeria in terms of a reconstruction of indigenous institutions of warfare, in particular, Eggon concepts of ritual warfare and its functions. What met in the Mada Hills were not merely two disproportionately armed fighting forces, but two different military ideologies (for want of a more apt phrase), representing two quite different perceived, experienced, and constructed realities. The impact of that confrontation was such that it has been transformed into a prototype myth encompassing the colonial experience of all Eggon, not just those directly involved in the Wulko hills campaign.

The three principal components of demography--fertility, mortality, and migration--can each provide information for reconstructing elements of the African past. Birth rate, more than anything else, determines the reproduction of the labor force, a particular concern of many historians. Death rate is an existential measure of the quality of life; any assessment of the long-run effects of colonial rule must take into account any discernable changes in mortality patterns that may have occurred. Finally, the study of migration is essential for understanding the redistribution of people and resources which resulted from the slave trade and colonization.

Although all three indicators are essential for a complete understanding of the history of a given population, they occur with substantially different rhythms. Mortality is a one-time happening for us all. Fertility depends on both partners in a sexual union, although most clearly circumscribed by menarchy and menopause in females. Migration, particularly in Africa, has affected males more frequently than females and can be, moreover, an abrupt and irregular phenomenon. Following the three activities simultaneously is thus an extremely complex undertaking.

For the developed world demographers have examined whole countries for periods as far removed as the eighteenth century and individual villages for demographic phenomena reaching back to the Middle Ages. Unfortunately for students of Africa, the materials available to us are not nearly so rich. Most colonial regimes did not conduct censuses in the current sense of the word. At most, they conducted surveys or village-by-village ‘nose counts’ which, naturally, vary enormously in their reliability.

As is well known, in his Nouvelle Relation de l'Afrique Occidentale of 1728 Père Jean-Baptiste Labat published material which led many later historians to suppose that the first exploration by Frenchmen of the furthest navigable point of the Senegal River, the Falls of Félou, was that conducted in 1698 by André Brue, chief agent of the Compagnie du Sénégal and governor of the post at St. Louis, following an abortive mission towards the falls despatched by Brue in 1697. However, in 1893 Henri Froidevaux published documents which showed that an earlier governor of St. Louis, Louis Moreau de Chambonneau, after a visit in 1686 to Galam, the “kingdom” containing the falls, sent two boats upriver in 1687 which reached the falls; and that after his return to France in July 1688, Chambonneau presented his superiors with a crude map of the falls. In 1913 Prosper Cultru published an account by a contemporary of Chambonneau, Michel Jajolet de la Courbe, a visitor to St. Louis in 1685–86. This account stated that Chambonneau had made an unsuccessful attempt to visit Galam in 1685, and that La Courbe himself, at a later date, (unstated but almost certainly 1690), during his second period in Africa, visited “the highest point in the river that can be reached.” Furthermore, La Courbe's account proved that Labat had misleadingly ascribed to Brue ca.1700 a whole series of journeys through Senegambia in fact undertaken by La Courbe ca.1685. Although the alleged achievements of Brue thus decisively invalidated do not include the 1698 visit to Galam, Brue's responsibility for the 1697 venture towards the falls is thrown into doubt; and other evidence cited by Cultru makes it less than certain that in 1698 Brue actually visited the Falls himself. In 1958 Abdoulaye Ly republished the Chambonneau manuscript map of the falls, and accepted that they had first been explored by the 1687 mission. Finally, in 1968 Carson Ritchie published two accounts by Chambonneau, written in 1677, during an earlier period of service at St. Louis, which, though not claiming that the author had yet visited Galam, stated that it had already been visited by other whites, presumably French colleagues. This indicated more gradual French exploration of the region leading up the falls than was once supposed, for even La Courbe, in a document of 1963 cited by Froidevaux, had spoken of Galam being “discovered only 7–8 years previously.”

Translations ought not to serve as sources for academic research. This precept is an ideal which has today become quite unattainable in many fields, notably in the natural sciences and medicine. In the human sciences it is still widely operative (at least tacitly), yet here too it is increasingly becoming Utopian. For any scholar of German literature, of course, command of the German language is (and, one hopes, will remain) an essential prerequisite: a treatise on Goethe's Faust based on a translation of this work could scarcely be taken seriously. Likewise it should be expected of classical historians that they study their Greek and Latin sources in the original language. The question becomes more problematic, however, when we consider the citation of modern pieces of research and other secondary literature. Admittedly, the majority of studies are still published in one of the major world languages, and it is possible to get by with a command of two or three modern languages (e.g. with English and French, with Russian and English, or with Arabic and French). But the internationalization of research is gaining ground steadily. A book written in Japanese on, say, the history of South America has little prospect of coming to the notice of western historians unless it is made accessible to them in translation, in this instance probably in Spanish, Portuguese, or English (leaving aside as a rare exception the historian with an above-average flair for languages).

While the historiography of art as an academic discipline can hardly be construed as a science, it is nevertheless governed by certain dominant paradigms in both of the senses that Thomas Kuhn intended. First, at any point in time there is a constellation of beliefs, values, and techniques shared by the community of scholars who comprise the discipline known as art history. This can be further broken down, altered, and refined for the various sub-fields, but taken together, the separate facets constitute a “way of seeing” art history which differs substantially from the “way of seeing,” say, political history.

Applying Kuhn's second and more rigorous sense, the historiography of art is dominated by certain paradigms which serve as exemplars or models of puzzle-solutions. While these change over time (it is no longer permissible to ascribe German expressionism to “national character,” for example), they are so powerful that they function as unquestioned assumptions when in force. Even more importantly, they are frequently invisible because they are rarely made explicit. In European art history, the dominant paradigms have coalesced into entities such as “The Baroque” or “Mannerism” which are largely ontological models used to simplify the otherwise intractable complexity of European art styles and movements.

The precolonial history of Yorubaland, and especially of the most powerful Yoruba kingdom of Oyo, is still dominated by the influence of the Revd. Samuel Johnson, the Yoruba clergyman whose History of the Yorubas was published in 1921. Especially for the period before continuous written documentation of Yoruba affairs begins with the arrival of the first Christian missionaries in the area in the 1840s, Johnson's History remains a source of overwhelming importance, and the framework within which we study Yoruba history is still very much that established by Johnson. But despite its great importance and persisting influence, Johnson's work has still not been subjected to much critical evaluation. In particular it has commonly been assumed, implicitly rather than explicitly, that Johnson's History can be mined as a source of ‘oral tradition’ in the belief that what he wrote down is unproblematically identical with what he heard, and what he heard with what had been retained and transmitted orally down to the time that he made his enquiries.

Johnson's History is a somewhat complicated work to characterize, since it makes use of rather different sorts of sources for different periods of Yoruba history. The history of its publication was also complex. It was originally completed in 1897, the date of Samuel Johnson's “Preface,” but the manuscript was lost. Samuel Johnson having died in 1901, it was left to his brother Obadiah to reconstitute the work from Samuel's notes and drafts, a task which he apparently completed only in 1916.

A wholly fraudulent source is an unusual occurrence in historical research. What is much more common and of much more intellectual interest is the discovery of a source that skilfully combines spurious invention with some regard for accuracy and convincing detail. This second kind of source is of course very familiar to historians of Africa, and the present note deals with a document of this type that, among other things, purports to offer a brief first-hand account of life in Kumase in 1839. I am chiefly concerned hare with the accuracy or inaccuracy of this piece of reportage; and--unusually for precolonial Africa--the document presently under review can be directly compared with a number of other precisely contemporary written accounts of life in Kumase.

In his pioneering work published over thirty years ago on the suppression of the illicit nineteenth-century slave trade from Africa, Christopher Lloyd remarked on the paucity of first-hand accounts authored by slave traders. He also offered a number of judicious observations respecting the veracity or reliability of such accounts of this type as did exist. In his treatment of the workings of the illegal West African slave trade, Lloyd relied very heavily on one of the annotated editions of the memoirs of Théodore Canot (alias Théophile Conneau). Canot or Conneau is justly famous, and in the intervening years since the appearance of Lloyd's book this important source has become something of an exegetical industry among historians of West Africa.

“Settlements of foreign, predominantly Semitic, peoples”

Strandes' gambit concerning ‘Muslim Civilization’ of the east coast of Africa is a familiar one to many Africanists. Persians and Arabs, so the stories go, settled coastal sites as part of the Islamic diaspora; they vanquished less virile African societies; they built cities which were reflections of Middle Eastern prototypes; they imposed their religion; and, they ‘founded’ coastal civilization, a civilization, therefore, which was characteristically Middle Eastern. Nineteenth- and early twentieth-century historians cannot be entirely blamed for holding such simplistic views. After all, the traditions themselves at least imply some of these things. And, given the literalist and diffusionist nature of past anthropological and historical theory, simple and biased interpretations of these traditions are not surprising.

What is perhaps remarkable, given the developments in anthropological theory since the 1940s, is the persistence of such views in some quarters. Examples include portions of two recently published papers by Saad and Wilkinson. Both deal wholly or in part with the most intriguing of the coastal origin traditions, the stories which tell how many coastal towns were originally settled by immigrants who came from Shiraz or Persia (see appendix). Their interpretations are so literal as to link the Shirazi name not merely with a particular dynasty, but with a specific family or kin group which ruled in pre-fourteenth-century Kilwa.

Precolonial African historiography has been plagued by historical reconstructions which remain in the realm of legend because events are suspended in almost timeless relativity. This kind of reconstruction seems to erect barriers which hinder or prevent synchronization with neighboring societies and regional events. Isolated from comparative material by the lack of even a relative chronology, coupled with the difficulty of incorporating undated evidence into the mainstream of regional historiography, suggests that the basis of history has not been provided. However, if we adopt the perspective that chronology distinguishes history from legend, the onus falls directly on every historian to provide a reasonably accurate relative dating progression through which the correlation and synchronization of regional data can be assessed. Development of this objective will eventually provide the basis of regional history, social history, and a realistic reappraisal of precolonial historiography. The following discussion, therefore, endeavors to present a methodology of chronological development for a particular case study, and suggests means whereby a relative yet reliable dating progression might be determined.

“Strange foreign jewels on a mournful silent shore”

Historians have frequently viewed the Swahili-speaking peoples of the East African coast as members of an Arab diaspora that spread around the Indian Ocean with trade over the last two thousand years. The interpretation flowed easily from the apparent “Arab” nature of Swahili culture--a written language using Arabic script, elaborate stone buildings and mosques constructed in urban settings, Islam, and genteel social behavior--especially when contrasted with the culture of mainland Africans, members of preliterate, uncentralized communities. Since the Swahili culture of the islands and coastal fringes bore little apparent resemblance to the cultures of the mainland, historians reasoned, its development could only have been the product of Persian and Arab merchants bringing to the “mournful silent shores” of East Africa the “jewels” of their own Muslim civilizations.

The perspective was essentially diffusionist in assuming that cultural innovation and historical development in Africa could only have come from elsewhere, and racist in assuming that race and culture were so inextricably linked that a separate “race” of immigrants had to carry these new ideas. As a result, historians failed to investigate the possible African roots of Swahili culture in their Bantu language, their religious beliefs and values, their economy, or their social structure. But this charge applies not only to European historians; Swahili oral historians have long recounted the development of their societies in essentially the same terms in involved genealogies tracing the development of different Swahili families, communities, and institutions back to Persian or Arabian ancestors. When European historians came to study the oral traditions of the Swahili (usually in written, chronicle form), they thus found ready confirmation of their own assumptions and interpretations.

This paper sets forth the methodology employed to establish and refine a dating system based on genealogies collected during field research in two areas of Africa, Agago county in Eastern Acholi and Awe district in the Benue Valley. Interview sessions with the elders normally lasted from two to four hours and many topics might be, and usually were, discussed in the course of a single interview. To demonstrate the progression of questioning relative to multiple themes would be beyond the space available or the stamina of readers. It is proposed rather to isolate the problem of chronology and show how, throughout the fieldwork period, achieving accuracy should be a major preoccupation and of greater or lesser concern in every interview. The regnal list is not collected once or twice or thrice. It forms a problem which continuously protrudes into the interviews, more so in the early stages of fieldwork, less so in the later, but seldom totally ignored.

When historical research focused on large kingdoms such as Bunyoro or Oyo, a vague and extremely relative chronology was adequate for many research objectives. One consequence of this type of research was a precolonial history of modern African states fashioned from a series of chapters, each one concentrating on spearate tribal traditions. The impression created was of isolated groups whose interaction, except for some outstanding military conflicts, awaited the coming of Arabs and Europeans. Since these aliens brought dates, integrated regional histories were forced to await their coming.

Bio-bibliographical studies do not belong, despite the initial impression that might be created by the term, to the same genre as new or faddish methodologies such as prosopography, cliometrics, or psychohistory. Rather, the term is one--at least in its most commonly used context--which describes the creation of a reference tool designed to meet specific criteria. Chief among these are the provision of evaluative descriptions of all extant, identifiable sources of material on the individual being covered, together with suggestions and/or commentary as to the utility of these sources and how they might best be employed in writing a comprehensive biography. Also relevant to such studies, and of particular interest as regards the intended audience of this paper, is some attention on the part of the author to the contribution his finished product can make to understanding the milieu in which his subject worked.

Among the various sources of information about the Africa of the Slave Trade era, one of the more voluminous and detailed is the great series of English Customs records held at the Public Record Office in London under CUST 3 and CUST 17. Each year's records are contained in a gigantic ledger. Those up to 1780, in the CUST 3 series, are organized entirely under countries, of which the whole of Africa counts as one (though each West Indian island is separately listed, and most of the North American colonies, though New England is mercifully counted as a single unit, as is “Virginia and Maryland”). In the CUST 17 series, which overlaps with the CUST 3 series from 1772 and runs until 1808, the arrangement is somewhat different, and a second series of entries is arranged under commodities. From 1796, the Cape (newly occupied) is shown separately; Sierra Leone appears in 1798, and Morocco in 1807/08. A fire destroyed all the records for 1813, and export records for the previous years, as well as an unknown number of more detailed records for earlier years, and also the records of the East India trade, and of Prize goods, and odd other parts of the records for the years before the fire. So far as Africa is concerned, it is possible to construct continuous series for Imports into England, and for re-exports, with only 1813 missing, but the export records for 1809-11 inclusive are lost. The list of countries recognized continues to vary after 1809, and from 1827 to 1845 the records divide up the coast of Africa into stretches, of which only “Gold Coast and Cape Coast” bears any relation to later divisions.

The struggle by the people of Zimbabwe to overthrow Ian Smith's Rhodesian regime in the 1970s won the support and gripped the imagination of democratic peoples around the world. As the struggle intensified, so it began to move away from nationalism towards liberation. Control of the state was no longer seen as an end in itself. For the vision of a socialist Zimbabwe to become reality, the state would have to be fundamentally restructured. The pace and direction of this historic movement has lain at the center of all subsequent encounters between the advanced capitalist countries and their local class allies on the one hand, and progressive elements inside ZANU on the other. While the former have busied themselves with reversing such progress as was made during the war, the latter have tried to avoid the snares and pitfalls which infest the pragmatic roads leading away from the compromise reached at Lancaster House.

Education no less than other state apparatuses has been a site of struggle between the forces of accommodation to the rule of capital and those of socialist transformation. By any measure, the expansion of Zimbabwe's educational infrastructure since independence in April 1980 has been phenomenal. Awarded some Z$408 million in 1982's Budget--a four-fold increase in little over three years--the Ministry of Education now caters for 1,903,917 primary school pupils, as opposed to 819,586 in 1979. Secondary school enrollment over the same period has climbed from 66,215 to 218,430. Equally impressive gains have been registered in the number of schools and schoolteachers.

The present report is intended to complement the report on the Norddeutsche Missiongesellschaft archive provided by Donna Maier in History in Africa, 8 (1981), 335-37. I worked in Bremen in the summer of 1982 and the following information may be of interest to readers. Firstly, several small organizational details have changed. The Staatsarchiv is now open from 9 a.m. to 4 p.m. on weekdays, except Thursdays, when it remains open until 8 p.m. Photocopying is available at AO Pf per exposure and, once a week, on Thursdays, the archive can provide microfilm facilities at 20 Pf per exposure.

As a social anthropologist whose principal interest is the central Volta Region of Ghana, I was most concerned to examine material pertaining to the Ho and Amedzofe areas, the sites of two of the Bremen mission stations. However, the files under the index reports for these stations are mostly concerned with internal mission business: finance, station-building, the training of and reports from native catechists (from the 1880s), and the setting up of schools. On the other hand, the sub-section “Land und Leute” (reference 7, 1035: 41/1 and 41/2) contains several attempts by the missionaries at general descriptive ethnography. Included in this section are a relatively early account by Schlegel, “Zur Geschichte der Eweer;” a long account by Spieth of Avatime; accounts by Schosser and Spieth of Akpafu ironworking and smithing among the Ewe, respectively; and an account of Anlo by Hornberger.