The old Oyo ‘empire’ was the largest and best-known of the Yoruba kingdoms. Located in the savannah below the bend of the river Niger in the Bussa-Jebba region of southwest Nigeria, it achieved prominence during the seventeenth and eighteenth centuries but collapsed and disintegrated in the early years of the nineteenth. Its origins and early history are imperfectly known because the traditions dealing with this period are enmeshed in myth and legend. This state of affairs has led one writer to conclude that the history of this period “is beyond meaningful’ enquiry.

Two major problems confront anyone attempting to reconstruct early Oyo history. The origins of the kingdom are linked both to the process of the settlement of the Yoruba people in their present location and to that of state formation among them. Furthermore, information about these processes is to be found in traditional accounts that seem to have been fossilized since the publication of Samuel Johnson's The History of the Yorubas in 1921. Indeed, many subsequent ‘traditions’ seem in no small measure to be derived from this work. It is therefore appropriate to begin this paper with a discussion of the influence of Johnson's work, followed by an analysis of Johnson's sources and motives, insofar as these can be determined. In 1901 an Iiebu man found it necessary to make an emphatic declaration on Yoruba history:

In the pages which follow, the traditions surrounding the first eight (of a total of thirty-five) kings of Buganda's traditional history will be investigated from a perspective that has not been utilized before. I hope to demonstrate that these traditions—beginning with the establishment of the first king, Kintu, in Buganda proper and ending with the death of the eighth king, Nakibinge, at the hands of the Banyoro—form an interrelated set which can be analyzed structurally as myth. While so doing, I do not wish to imply that the traditions in question demand this type of analysis and no other; different perspectives have been and will be useful for understanding early Buganda and its traditions. But I do want to show that such a structure can be discerned in the traditions and that the existence of this structure should be taken into account regardless of the type of analysis to be undertaken.

One task which will further these ends will be a discussion of the reasons for setting off the particular traditions chosen from the larger context of the set of traditions surrounding all the kings of Buganda. Following this, the selected traditions will be subjected to a structural analysis. In the process, however, I will digress frequently from the dominant mode of analysis to discuss: (1) variant or additional traditions which seem to remain outside the particular structural themes dominating the traditions as a set; and (2) alternative interpretations of the traditions, especially historical ones, which have been or might be offered. Finally, the structural themes discussed in the body of the paper will be represented in a comprehensive chart.

The Mpongwe people of the Gabon estuary live today in the immediate area of Libreville, the capital city of the Gabon Republic. Libreville is built on Mpongwe ancestral lands, and its history is only a small and comparatively recent chapter in the longer story of the Mpongwe and their neighbors. In the nineteenth century the expression “les Gabonais” or “the Gaboon people” had only one meaning—the Mpongwe of the estuary who were the coastal trading aristocracy.

The Mpongwe are only one of the six peoples belonging to the Myèné-speaking group of Gabon. The other five are the Orungu, Nkomi, Galoa, Adyumba, and Enenga. Only the Mpongwe are patrilineal. Myèné is purely a linguistic classification, a subdivision of the Bantu language. All six of these societies fit into a circle whose circumference includes the three largest cities in Gabon today—Libreville, Port Gentil (formerly Cape Lopez), and Lambaréné (Map 1). From the seventeenth to the nineteenth centuries, Myèné was the coastal lingua franca between the southern Cameroun and Cabinda. The Myèné societies in general, and the Mpongwe in particular, have played a key role in Gabon's past, and continue to be an influential minority in modern Gabon. In the early nineteenth century, and for an unknown previous period, the closest non-Myèné neighbors of the Mpongwe were the Benga and the related societies to the north, and the Shekiani and Bakélé to the east. The Shekiani were the couriers in the Mpongwe trade with the Bakélé and other interior societies.

Few peoples of the Middle East have produced as many historical works as the Armenians: their historiography dates back at least to the fifth century A.D. While most medieval Armenian historians have concerned themselves with contemporary history and the immediate past, there have been some who have attempted to trace Armenian history from the earliest times. It is to two of these, Pseudo-Sebeos and Pseudo-Moses of Khoren, that we owe the survival of the body of historical memories now generally referred to as the Primary History of Armenia.

This Primary History has come to us in two redactions, a long and a short. The shorter version is attributed to the earliest known Armenian historian, Agathangelos (fourth century A.D.?) and is presented in the opening section of a seventh-century work ascribed-probably wrongly-to a certain bishop named Sebeos. The longer version, much expanded and edited, is contained in Book One of the compilation of Armenian antiquities known as the History of Armenia by Pseudo-Moses of Khoren. While the date of this work has been much disputed, it appears now to be a product of the late eighth or early ninth century.

According to Pseudo-Sebeos the short redaction of the Primary History was a work originally written by Agathangelos, secretary to Tiridates HI (298–330), the first Christian king of Armenia, and was based on information contained in a book written by a certain Marab the Philosopher from Mtsurn, a town in western Armenia. Pseudo-Moses, on the other hand, claims that the parallel material in his history (I. 9–32 and II. 1–9) is an extract by Marabas Katiba from a Greek translation of a Chaldean history of Armenia made by order of Alexander the Great.

“History is always written by the conquerors.”

It is a convention in the writing of Nupe political history to begin with the name ‘Tsoede.’ This convention has been current at least since the early years of colonial rule, when we find ‘Tsoede,’ or ‘Edegi,’ as he is also called, being credited with the founding of the kingdom whose successive rulers can be traced up to the present. The clearest expression of the place of both ‘Tsoede’ and the kinglists which his name heads comes from the standard study of Nupe society by S.F. Nadel, who explains that

It will be our purpose in this paper to explore the evolution of the Tsoede story and to inspect the authority of its authorship. First, let us look at the story offered in Nadel's account:

• a) At the time of Tsoede's birth Nupe had not been unified under a central government.

• b) Whatever political forms existed elsewhere in Nupe, Tsoede's homeland Bini was a confederacy of towns. The leading Bini town was Nku, at the confluence of the Kaduna and Niger rivers.

The Binis as well as some other Nupe were subordinate to the Attah of Igala.

Among the Xhosa the institution of the ‘Right-Hand House’ acts both as a political charter and as an historical explanation. As a political charter it defines the relationship between the Ngqika Paramount (the Right-Hand House) of the Ciskei ‘Bantu Homeland’ and the Gcaleka Paramount (the Great House) of the Transkei Homeland. As it presently stands, the essence of this relationship is that the Ngqika Paramount recognizes the Gcaleka Paramount as his superior in rank, but without accepting any implications of practical political subordination. This position was defined by J.H. Soga, the standard authority on Xhosa history and customs, and himself an umNgqika, as follows:

In terms of historical explanation, secondary authorities from 1846 to 1975 have singled out the privileged status of the Right-Hand House as the principal cause of Xhosa political fragmentation.

Whereas historians of Africa normally agree that institutions and their myths of origin are, at least in part, susceptible to historical interpretation and reconstruction, they may justifiably be more doubtful of an historical approach which seeks to explain historical events by imputing to the past the continuous retrogressive operation of institutions which can be seen to be operating in certain ways in the present. In this regard the present exercise has two aims.

Since the publication in English in 1965 of Jan Vansina's Oral Tradition, historians of Africa have been increasingly concerned with developing methods which confirm the historical value of oral traditions. Independent proof for the historicity of oral traditions is often lacking; consequently the historian is usually left with comparative analysis as his primary analytical method. Archeologists such as Merrick Posnansky, Frank Willett, and John Sutton have in part attempted to show linkages between oral traditions and archeological evidence.

Posnansky, especially, has contributed much to the idea that it is possible to combine the two sources to obtain a more comprehensive view of Hfeways usually referred to as ‘prehistoric’ Recent research, though, now suggests that the concept ‘prehistoric’ must be questioned, particularly in cases where there is a demonstrated tie between archeological evidence and oral traditions. When archeology affirms the accuracy of oral traditions which explain, comment on, interpret, or locate activities and sites which predate a literate tradition, then the germaneness of the concept must be critically questioned. It is my position here that when archeological evidence confirms the historical value of oral traditions about preliterate life, then those cultural phenomena in that time period should be considered historic rather than prehistoric. To retain ‘prehistory’ as a concept in this context ignores and even militates against the historiographies and historical concepts of other cultures. The ramifications of this relativist perspective are considerable both for the study of history and of prehistory. Given this logic, historians must begin to reassess and expand their concepts of what history is and prehistorians must prepare to forfeit part of the temporal domain previously considered as prehistory.

The most important development in African historiography since 1970 is almost surely the growing awareness that oral tradition alone is insufficient as a source for reconstructing the past beyond about 1750. Henceforth historians will be increasingly obliged to turn to other bodies of data, at least if they wish to avoid writing history which, in Bradbury's words, is no more than “the rationalization of myth.” Despite pleas by, among others, Bradbury, Vansina, and Lewis as long ago as a decade or more, historians have thus far shown little sign of incorporating ethnographic data into then-catalog of sources, or comparative ethnography into their methodological tool-kit. One reason for this reluctance (at least among English language Africanists of our generation) may be a vague acquaintance with the serious abuses which European Africanists— especially those associated with the Vienna ‘culture-historical school’—committed by stretching ethnographic data to fit grand, but spurious, global schemes of human evolution. Clearly, though, earlier abuses in no way justify our refusing now to exploit a body of data we badly need, provided we examine it with appropriate care and modesty. We must admit at the outset that ethnographic comparisons for historical purposes on a continental scale will be impossible until a good many regional studies are available. In turn, such regional studies will normally be limited to selected aspects of related culture, and not to cultures as holistic units. This means that points of articulation between regional studies may well be difficult to identify.

The period between the eleventh and sixteenth centuries is a perplexing one in the history of the east African interior. For the first part of the period archeology is the most reliable indicator of what really happened there. For the second part oral traditions supplement the archeological record. Regarding evidence, therefore, our period is one of transition. Our principal concern in this survey will be to consider how far this transition in evidence reflects actual transitions in history.

This is an intricate problem. Each piece of historical evidence reflects its historical origin. This applies as much to pottery fragments and spear tips as to oral traditions. However, the former are easier to handle historically than the latter; they are products of their own time. On the other hand, oral traditions tend to be products of our time in this particular area of Africa. Before the nineteenth century indigenous literacy was restricted to a very narrow strip of land along the coast, and during our period contacts between coast and interior were extremely limited. Literate outsiders did not penetrate into the interior in any number until the last half of the nineteenth century, and even then earlier history did not receive much attention from these footloose adventurers. Oral traditions relating to the early history of the interior were only reduced to writing in substance during the subsequent colonial period-sometimes by newly-arrived European missionaries and administrators, more frequently by newly-literate African intellectuals. These circumstances create special interpretative problems. Some of these problems are common to all oral traditions relating to early history, while others are peculiar to the particular colonial situations in which the oral traditions were first recorded.

What will follow in this and subsequent papers will constitute an extended gloss of the English translation of Willem Bosman's account of the Guinea coast. This translation was published in London in 1705 under the title A New and Accurate Description of the Coast of Guinea … and was based on the 1703 Dutch edition published in Amsterdam. For the English translation I have used the edited reprint published in 1967, which was an identical copy of the 1705 edition. For the Dutch version I have used the 1737 edition, also published in Amsterdam; this was in fact the last of a series of reprints of the 1709 second edition. This second edition contained numerous amendments and additions to the first (1703) edition, and these will be noted as they occur.

In addition to including all the material in the Dutch edition which was omitted from the English translation, I have included all passages in which significant differences in tone or meaning occur. Throughout I have tried to retain the capitalization, italicization, and punctuation employed in the Dutch version–styles to which the English translation generally conformed.

Citation is by page number, paragraph, and line number within paragraph, with the first paragraph presumed to begin in each case with the first line of a page. Thus, for example, P. 17 I/8 indicates that the gloss concerns the eighth line of the first paragraph on page 17 of the English translation. Shorter passages are arranged on a FOR … READ basis, while for longer passages only the Dutch original is included after SHOULD READ.

Neither Chad (or Tchad) nor the Central African Republic (or République Centrafricaine) has been much studied by American scholars. Because of this the following report discusses more than merely the archival resources of the two countries; it is hoped that the more practical aspects of research there that are discussed will render it more useful.

The Institut Tchadien des Sciences Humaines (henceforth INTSH) supervises research projects in Chad which are concerned with the social sciences—as defined by French scholarship. Research authorization is granted by the Director of the Institut, Boîte Postale 503, N'Djaména. Anyone intending to pursue research in Chad should have a letter of authorization from the Institut before leaving the United States—if possible, even before applying for a Chadian visa, since this will expedite the granting of such a visa.

Most temporary accommodations in N'Djaména (like other African countries) tend to be expensive; the only inexpensive hostel is operated by the Peace Corps for its own members. Sometimes, though, researchers have been allowed temporary accommodation pending the location of other housing. Enquiries should be sent to: Corps de la Paix, Boîte Postale 260, N'Djaména. In addition, temporary housing may sometimes be had with researchers at the Organisation pour des recherches scientifiques et techniques d'Outre-Mer, or ORSTOM.

For a long stay it is wise to rent a house. While the Peace Corps or Embassy personnel may be of some assistance, Muslim merchants near the main market should be of more help in finding a house in the African part of the town. Rents in N'Djaména are high: a mud brick house with electricity and running water will run about $175 monthly.