The interdisciplinary task of reconstructing early African history demands more than a passing familiarity with many fields of knowledge. Not least among these is geography, in its broader conception as the pursuit of an understanding of physiographic processes and the role(s) played by people in shaping and responding to these processes. A standard beginning in many a history book takes the reader through annual rainfall figures, topographic generalities, and, perhaps a cursory pedological overview of the region in which the people(s) under study live. Rarely are these introductory remarks integrated into the historical narrative. We are too often left to ourselves in determining the relevance, for example, of a bi-modal rainfall to the local agriculture in a subregion of open grasslands and woody savanna ecotones on black clay valley bottoms. We trust these are data relevant to our understanding of the region's agriculturalists.

Many readers may know that such an area will be subject to heavy soil waterlogging during the wetter months. They may also realize that mound and ridge construction are excellent responses to the risks of root-drowning such waterlogging poses to crops such as beans, and as well, that the indigenous African grain crops Sorghum spp. and Pennisetum spp. can withstand waterlogged soils for brief periods without suffering significant loss of yield (Purseglove 1972:269-71). What is likely to be poorly understood are the forces of ecological and cultural change that have been at work in such an environment transforming the very nature of its vegetational cover and soil profile. In short, a historical perspective is necessary so that we do not run the risk of presenting Africa's ecological background in the same unchanging light as colonial ethnographers did Africa's past.

In a previous number of this journal, a note on the development of a computer-based Arabic manuscript finding aid described its application to a manuscript microfilming project in Boutilimit, Mauritania. The filming of that collection was then (at the time the note was written in the spring of 1989) in its second phase, and the work was concluded in December of that year. Thanks to the technology developed for the finding aid to that collection, the catalog was completed six months later, and it is now possible to do database searches on that material (and soon on other collections now being entered in the same format). This note comes as a description of the Boutilimit collection, the film of which is now available to researchers at the University of Illinois Archives, in the University Library.

Regularity of phonological change is considered a basic principle of historico-linguistic reconstruction. In 1876 A. Leskien of the Leipzig school of neogrammarians formulated the postulate of sound change occurring without exception, meaning “that if one assembled all the facts, and analyzed them accurately and thoroughly, one could state exceptionless principles or laws for the development of language” (Lehmann 1973:87). However, it has since been argued on the basis of empirical research that languages are never in complete balance and that, therefore, change is not necessarily constant, nor parallel among groups speaking different languages. Various reasons have been found to account for irregularities in language: the impact of one language on another, the effect of children acquiring their language, etc. Yet before we enter into any etiological discussion of irregular or unstructured phonological change, the area of investigation for the present case study must be introduced, and the characteristic of the problematic nature of the case sketched.

On 24 November 1989 the Library at the University of Swaziland in Kwaluseni formally received a large portion of the papers and manuscripts of the eminent Swazi writer, historian, statesman, and current chief executive of the Swaziland National Trust, J. S. M. Matsebula. The deposit of these papers has formed the nucleus of a small, but interesting and rapidly expanding, Swaziana Collection and this contribution, it is hoped, will inspire other prominent Swazis and scholars of Swaziland similarly to deposit their materials. The preeminent anthropologist, Hilda Kuper, has recently offered her large and valuable collection of manuscripts and field notes to the University of Swaziland, and the elder statesman, J. J. Nquku, founder of the Swaziland Progressive Party (among other accomplishments), is preparing his personal papers for the Collection as well. Additionally, it has just been learned that the Swaziland Oral History Project materials will be transferred to the University of Swaziland from the University of the Witwatersrand.

James Shadrack Mkhulunyelwa Matsebula was born in 1918, outside Mbabane, the capital of Swaziland, into a family which has traditionally played a significant role in ritual Swazi royal life. The King's first wife is chosen from the Matsebulas and is known as laMatsebula. He was educated in Swaziland and South Africa and has written on a wide variety of topics in several genres and languages.

In 1974—the same year I ventured into the field to begin collecting oral data for my doctoral thesis, a precolonial history of a Mandinka state at the mouth of the Gambia River—Robert W. Fogel and Stanley L. Engerman published their now much maligned work on African American slavery, Time on the Cross. With publication of the book, Fogel and Engerman did something few historians had done before or have done since: they made public their evidence—all of it, data and statistical methodology—so others could determine how they had arrived at their conclusions. Perhaps it was because their interpretation of slavery was so different from those preceding it that historians used Fogel and Engerman's published evidence to dismantle, piece by piece, their arguments and the way they had arrived at them.

But making available otherwise inaccessible evidence seemed to me the right thing to do. So, in the field and afterward, I offered up my oral data. (The written evidence I used was already available, either published or in archives at various places on three continents.) I deposited copies of cassette tapes of my interviews, with copies of transcribed translations, in the Gambia and in the United States. Also, within a few years of finishing the dissertation I published two volumes of translated, transcribed, and annotated oral traditions from the collection in an inexpensive series that I thought would be accessible to most interested parties. If people wanted to test my hypotheses, attack my methods, or berate my conclusions, they at least had the materials for doing so.

This paper needs to be read against the background of the critique which has gradually been gathering force over the last half-dozen years or so of the concept of the mfecane. By the mfecane is meant the idea that in the 1820s much of the eastern half of southern Africa was thrown into turmoil by a series of wars and population migrations set in motion by the explosive expansion of the Zulu state under Shaka. Ever since Theal first popularized it in the late nineteenth century, this idea has remained one of the bedrock concepts around which the history of southern Africa in the later eighteenth century and the first half of the nineteenth century has been written. The term mfecane itself did not become widely used until as recently as 1966, when, in his widely influential book, The Zulu Aftermath, Omer-Cooper repackaged what had previously been called “the wars of Shaka” for an emerging Africanist readership. Since then the concept of the mfecane has permeated the literature, both popular and academic, inside and outside southern Africa, to the point where it is regarded as a fixed fact of the sub-continent's history.

When in 1985 I issued an English translation of the 1684 version of the Description of Guinea by Francisco de Lemos Coelho, I referred to the 1973 annotated French translation of one chapter by Nize Isabel de Moraes, and I noted that this included a brief statement to the effect that “une publication de la seconde version, en portugais actualisé, fut entreprise en Lisbonne en 1937 (Inéditos Coloniais, sér. A, num. II).” This statement puzzled me, since the 1953 edition of the Portuguese texts of both versions, by Damiãlo Peres, said nothing about a previous edition, leaving the reader to suppose that the 1953 edition represented the first time these accounts by Lemos Coelho had found their way into print. However, since the reference by Nize Isabel de Moraes was a little imprecise (why “fut entreprise” rather than “fut imprimé,” for instance?), and since I have never been able to see her thesis, which originally contained the French translation and may give additional information about the 1937 enterprise, it has taken me some years to confirm that there was indeed an earlier edition and an earlier editor—albeit in rather curious circumstances, about which I would like to learn more.