In the past twenty to twenty-five years valuable contributions have been made to southern African agrarian history. Stanley Trapido's publications, for example, opened up stimulating perspectives on the processes and forces inherent to nineteenth-century Transvaal agrarian history. Although he was modest in his 1980 chapter, “Reflections on Land, Office, and Wealth in the South African Republic, 1850-1900,” and referred to it as “a tentative and preliminary attempt to outline some important aspects of these social relationships,” it has provided historians and others with an important instrument of analysis.

However, there are still themes, regions, and periods that need attention, one of these being the central districts of the Transvaal before the industrial revolution. In this regard a little-known source which may contribute to our knowledge of the pre-industrial history of the Transvaal, and which will be published this year as an annotated source publication, should be taken note of. This is the 1871 Commission on African labor in the Zuid-Afrikaansche Republiek (ZAR). Despite the valuable information contained in its documents on agrarian history and various aspects of race relations, especially with regard to the central districts of the Transvaal, it has been neglected by historians in the past. Of the few historians who refer to the 1871 Commission, most have merely utilised the report of the commission and have probably missed the important testimonies, correspondence, and minutes. Very few have managed to locate these documents, which are concealed among the supplementary documents of the State Secretary for 1871 in the Transvaal Archives.

Visual strategy begins on the first page. The most pretentious form of first page differentation is the “lead-in” quotation whereby the author prefaces the main body of the text with a quote from an esteemed scholar, a famous decision, or some other prestigious source. … The objective of the “lead-in” quote is to spark immediate attention with a titillating example of erudition, humor, or impertinence… Ideally the lead-in quote should be obscure—oriental sources are recommended—and should not have a substantive link to the subject matter of the article. … This technique can generate guilt among readers who suspect the game but lack the nerve to speak out.

His books positively clank and groan under the weight of apparatus. Very good it is too.

As indicated in the first part of this paper, I adopt a generous definition of “annotation” in this discussion. There the traditional forms, footnotes, and other textual apparatus were discussed. Here I want to concentrate on a number of forms of annotation that are not usually treated under that rubric. Included (in roughly the order in which they are likely to appear in a given work) are titles, tables of contents, prefaces, epigraphs, graphs and charts, maps, quoted matter, facsimiles, appendices, glossaries, bibliographies, and indexes. Each of these is an occasion—and an opportunity—to provide access to the text, to the author's own sources, or to the author's mind. While every work will not use all of these, certain of them (prefaces, tables of contents, bibliographies, and indexes) should be a part of every substantial scholarly study.

Mainly as a result of the work of E. E. Evans-Pritchard, the Azande are among the best-known African peoples. In anthropological theory they have become indissolubly associated with the study of religion and magic. Also remarkable is their expansion under the leadership of the dynasties of the Avungara and the originally Ngbandi-speaking Abandia. Starting from a small core area in the basin of the lower Mbomu, where the ancestors of the Avungara and Abandia had established themselves as rulers over parts of the resident, mainly Zande-speaking, population around the middle of the eighteenth century, the Abandia extended their rule into the region of the lower Mbomu and lower Uele, while the Avungara and their Azande followers swept eastward in a vast movement and in less than one hundred years conquered a huge area reaching as far as the upper Sue and upper Uele, integrating the population into their system of rule.

One of the reasons for the speed of this expansion is that individual members of the Avungara dynasty (who all claimed descent from Ngura, the first historical ruler in the lower Mbomu area) repeatedly founded principalities of their own in new territories. This led to the existence of a varying number of polities under numerous, more or less, powerful rulers who descended from several dynastic branches, thereby preventing the formation of a single kingdom, stable in time and place. Through the integration of numerous groups of different linguistic and ethnic origins, the population cluster was formed for which the collective name Azande has become established. The history of Azande expansion thus provides a very interesting example of a society being created through political processes, which raises questions concerning the origin, acceptance, and characteristics of centralized political organizations, as well as assimilation and acculturation processes (besides the Mangbetu in the Uele-Bomokandi area, the Azande were the only group in the region to develop centralized political structures on a wide scale).

Samori Toure (d. 1900) is celebrated, both in written history and oral tradition, in Mali and Guinea because of the empire he founded and his fierce resistance against the French, as they sought to occupy their future colony of the French Sudan. Recently published anthologies of African epic (Johnson/Hale/Belcher 1997; Kesteloot/Dieng 1997; Belcher 1999) attest that an orally transmitted Samori epic exists in these countries. In this paper the texts hitherto presented as the Samori epic will be compared to some oral sketches about Samori which I recorded during two years of fieldwork conducted in southwestern Mali and northeastern Guinea. I will hypothesize that a Samori epic may be in the making, but does not yet exist. The texts hitherto presented as the epic of Samori are largely oral narratives produced more or less in concord with expectations about what an epic should look like. The focus is on Samori as a hero on the battefield, and this is not representative for the present-day oral narrative on Samori. Therefore, an epic of Samori, if it ever does come into being and takes the form of a standardized oral narrative, might deal with different issues than one might expect from reading the texts presented in the anthologies.

In this paper I take into account informants who oversee fieldworkers in local arenas. My intention is not to question the importance of participant-observation as the hallmark of field research but to elevate the recognition of informants in fieldwork. The argument centers on local overseers who are integral to participant-observation by controlling more of what the researcher observes, and with whom one participates, than they are usually given credit for in the literature. I intentionally use “overseer” rather than the usual “gatekeeper” label to convey a strong sense of informant involvement in ethnographic research (cf. Ellen 1984 where “gatekeeper” is used to describe a range of actors who control access, channel activities, and so on). The case study is drawn from my experiences with a host family in southwest Madagascar with whom I lived for 10 months in 1997.

There are different ways of representing informants in the literature. One way is to draw a wide line between informant and participant-observer. Informants are defined as local people who willingly provide information to participant-observers. Participant-observers are non-locals involved in intense interaction with members of a community over a long period such as a year. The two are not to overlap. Evans-Pritchard found informants vital to his Azande research, participant-observation central to his Nuer fieldwork. “Azande would not allow me to live as one of themselves; Nuer would not allow me to live otherwise. Among Azande I was compelled to live outside the community; among Nuer I was compelled to be a member of it” (Evans-Pritchard 1940:15). This tradition, of treating each as a separate method, is widespread (see, for example, Bernard 1995:136-79; Casagrande 1960; Ellen 1984; Freilich 1970:541; Herskovits 1955:376; Jackson 1997:188; Malinowski 1961:1-25; Pelto and Pelto 1973:242-43; Spradley 1980:177).

In an earlier issue of this journal I published the text of a letter to King George I of England written in the name of King “Trudo Audati” (better known under the name which he is given in in local tradition, Agaja) of the west African kingdom of Dahomey. Although dated 1726, this letter was received in England only in 1731, when it was belatedly delivered to London by Bulfinch Lambe, a former employee of the Royal African Company of England, who had spent some time in captivity in Dahomey, and who claimed to have written the letter at King Agaja's dictation. Lambe was accompanied to England by an African interpreter called “Captain Tom,” who vouched for the letter's authenticity; this man's African name was given as “Adomo Oroonoko Tomo,” though the middle name “Oroonoko” at least was surely not authentic, but borrowed from the popular romantic novel by Aphra Behn, Oroonoko: or, the Royal Slave (1689). An official enquiry by the Board of Trade decided that the letter itself was a forgery, though on grounds I at least find unpersuasive; but it was acknowledged that Lambe had been charged with some sort of message from King Agaja, and arrangements were made for the repatriation of the interpreter “Adomo Oroonoko Tomo” to Dahomey, which was effected in the following year, 1732.

Before I start, I would like to stress that the following is in no way meant as a complaint. Neither would I like you to think that I am disgusted by fieldwork in general, and fieldwork in Mali in particular. The point that I want to get across is that the personal emotional experience and state of mind during fieldwork have an impact on the way fieldworkers, in being their own instruments, practice their research. We would like to believe that this emotional state is one of euphoria, but in reality it includes negative experiences and feelings. These feelings should be accounted for—not just personally, but also professionally.

There exists a dichotomy between open and hidden discourse about experiences in the field among social scientists practising fieldwork. This dichotomy is as old as fieldwork itself and gained some acknowledgment of existence after the publication of Malinowski's Diary in the Strict Sense of the Term (Malinowski 1967). This dichotomy consists of a representation of fieldwork experiences towards colleagues and the larger audience as ideal, unproblematic, and among the best experiences a person can have in life. It is almost like a holiday. It is only in an informal setting, usually the pub, that some dare to speak of the danger, nuisance, despair, and general inconvenience of it all. Even talking about severe illness or other dangers is often done with the self assured attitude of someone capable of facing Armageddon and getting out alive. Apparently, Indiana Jones does exist.

His voice is difficult to understand; he is an old man. Sitting on a mat placed on the sand beneath his chair, I listen carefully, even though I know that my tape recorder will later help me in reconstructing his speech. His narrative is not unfamiliar and in fact it is almost surprising to me the similarity to WoDaaBe origin narratives that I have seen on print.

My discussion here focuses on WoDaaBe origin stories from the perspective of histories as being socially meaningful, integrated into political contexts. Ethnographic research among the WoDaaBe has reflected rather similar stories of origin as the one presented here, focusing on the WoDaaBe as originating in a mythical way, along with their cattle. My own research in the Tchin-Tabaraden area in Niger in 1996-98 indicated, however, that even though the narratives documented earlier by travelers and anthropologists are known, people were generally more interested in reciting and discussing stories strongly grounded in time and space. In my approach, I follow Jan Vansina's comment that “all messages have some intent which has to do with the present, otherwise they would not be told in the present” (Vansina 1985:92). Similarly, Ellen Basso has pointed out in relation to narratives, that it is not enough to ask, “what is a story about;” why it was told must be questioned as well (Basso 1995:x). I also contribute from theorists that have emphasized history itself as a political phenomenon formulating relations between past and present (Popular Memory Group 1982:240), making how the past is understood relevant to the present political-cultural context, without reducing it to merely functional justification for the present (Werbner 1998:2). History and memory thus have a texture that is both social and historic (Bommes et al., 1982:256), intertwined with power and meaning.

Writing in his Les Bambara du Ségou et du Kaarta, the French colonial administrator and ethnographer Charles Monteil considered the family name, or jamu, to

Monteil was right, to a certain degree. In the Western Sudan, family names are weighted with history and significance. Yet what Monteil characterized as evidence of stability and tradition, Charles Bird has more recently called a “ticket to mobility.” The fluidity and mobility that had come to characterize the jamu eluded Monteil entirely, just as its mutability often eludes contemporary historians.

A jamu represents both an all-important identity marker and an instrument of “mobility.” Yet it is also highly contigent, even aleatory. This mobility has a double sense, signifying both the mutable nature of the name itself and its potential for “making outsiders insiders” by creating an immediate link between people who would otherwise be strangers. Jamuw—the plural takes a ‘w’—also have a deep historicity. Embedded in them are history and myth, along with suggestions of family occupational category—commonly referred to as ‘caste’—and social status. Epics such as Sundiata often provide etymologies and legendary origins of family names, and scholars have sought—misguidedly—to use these to understand the historical processes of ‘caste’ formation and other aspects of the distant Mande past.

Lake Rudolf, which is also known as Lake Turkana, lies in the eastern arm of the great Rift Valley. It is primarily fed by the Omo River, which flows south from the Ethiopian highlands and sits in an inhospitable landscape of dormant volcanoes, wind-driven semidesert, and old lava flows. During the morning hours, strong gusts of wind usually blow from the east down the slopes of Mount Kulal and across the surface of the lake. This unrelenting wind creates large, white-capped waves on the lake's surface and makes navigation almost impossible. It also gives the lake a bluish color, reflecting the clear sky above. However, when the wind dies down in the afternoon, the lake takes on the color of green jade, due to algae that rise to the surface when the waters are calm. It is because of this afternoon and evening color that the lake has long been known as the Jade Sea (Imperato: 1998:3).

By the end of the eighteenth century the Baga, or peoples of the seaside (the bae raka), were already long-term residents of mangrove islands located between Guinea-Bissau and Iles de Los—the northern half of Guinea-Conakry's current coast. That fact is well known and accepted by everyone interested in the ethnohistory of this part of Guinea. What is less clearly documented or understood is how far they extended inland, the context of that residence which allowed them to operate economically and socially with guests upon their lands, and perhaps how outsiders influenced their characteristics, within a historically definable period. Drawing on observations from first-hand accounts written between 1793 and 1821, this study attempts to answer these questions and to suggest assumptions about the Baga experience and conclusions for particular Baga groups by 1821.

This paper flows from a longer and broader study of peoples and commercial activities in the Rio Nunez to Iles de Los section of coastal Guinea-Conakry for the 1750 to 1865 period, that period just prior to establishment of French suzerainty. It also emerges from a historian's desire to assign ethnic groups or at least ethnic types to identifiable regions and to understand relationship traits that may have operated during a specific period. The Baga were likely minor players in political and commercial transformations that accompanied the rise of slave trading or establishment of new markets for commodities' traffic occurring with its decline in the late 1830s. As producers of rice, salt, and livestock, and collectors of cola, however, and as ‘first-comers’ with claim to land through which commerce with the interior inevitably flowed, the Baga favorably served trades that found their basis in both trades and that attached interior markets to those of the coast.

Europe and the Americas have long dominated studies of transatlantic exchanges and much more is known about European participation in the Atlantic world than of its African counterpart. Current knowledge of how those parts of Africa located a few miles away from the African littoral contributed to the early modern Atlantic World is particularly sparse. This is despite the fact that the slave trade was the largest branch of transatlantic migration between Columbian contact and 1870, and that it is becoming apparent that Africans and indigenous Americans helped shape the new political and economic power structures, as well as the post-Columbian worlds of culture and labor.

Assessments of the impact of any group on the global stage must begin with the nature of the group itself, and thus efforts to raise the African profile in Atlantic scholarship and to focus on the agency of Africans must quickly face the contentious issue of ethnicity. From the broadest perspective, it is odd that the way the ancestors of the Atlantic World defined themselves should have become so much more contentious among Africanists and Afro-Americanists than among those scholars who study Europe and Europeans overseas. At the outset of the repeopling of the Americas, the European state existed in nascent form in only Spain, Britain, and France. The predominance of the nation-state in the way the world is organized in the twenty-first century—rather than its status in 1492—has perhaps led scholars to stress the contrasts between Africa and Europe on issues of early modern nationhood, and, more generally, human identity.

The use of colonial census data on religious conversion in historical analysis is often fraught with methodological problems. What, at first glance, might appear to have been profound structural transformations of society may in part be illusions created by changes in the census apparatus, or data collection process, itself. In this paper I will discuss how a colonial census was conducted and how local rural peoples were active agents in enumerating their own (official) religious identities. Thus, how did changes in this central tool of bureaucratic knowledge and power distort history? As an overarching critique, I suggest that the problem of the census as an exercise of colonial bureaucratic power in Africa has hardly elicited comment. Thus I suggest that the census may be a potentially fruitful avenue of historical inquiry in examining in closer detail the primary “enumerative modality” of colonial forms of knowledge.

I will present examples from the history of French colonial southern Mali during the postwar period in arguing that the colonial census was a complex social and political process that must be ‘deconstructed’ if we are to make critical use of such state-generated numbers. Southern Mali occupied what was believed, in the French imagination, to be an active frontier of Islamization, separating the “Muslim” Sahel (and Sahara) from the “fetishist” forest regions to the south. This religious frontier status elicited mixed views. As Jean-Loup Amselle has argued, colonial officials in southern Mali adopted a highly ambiguous stance toward Islam, “going from one extreme to another, sometimes seeing Islam everywhere, sometimes nowhere.”

The following bibliography is intended to supplement the excellent one (largely) of secondary sources compiled by Thomas Spear and published in History in Africa 27(2000). Research for a forthcoming monograph on the East African coast in the ‘middle’ period has taken me in recent years into a number of libraries and archives in India, East Africa, and Europe. There I have been able to build an extensive listing of source material and oral informants interviewed in East Africa. While this compilation includes many of the titles in Spear's list, study carried out in Goa and Lisbon afforded me the opportunity of viewing primary sources not included in Spear's collection. Despite the fact that this is still a work in progress, I submit this supplementary list hoping it might prove useful to other scholars interested in East Africa and the western Indian in the pre- and early-modern period.

Readers also will note that I have included some secondary listings not included in Spear's bibliography. This is due to the fact that my ideas concerning what is relevant to coastal history appear to be somewhat broader than Spear's. Consequently, this list includes some titles on southern and central Africa, as well as of coastal literature, which I have found to be useful and apposite to coastal studies. Naturally, I have tried not to duplicate titles found in Spear's list.

The intense scholarly debate concerning the shift from orality to literacy has not often directly concerned African historians in spite of the fact that many work closely with oral sources. In the process of publishing a series of locally-written histories, I discovered that transforming oral tradition into written form is ultimately political. It raised a number of important ethical dilemmas for me as a scholar and brought to my attention the power relations inherent in these transactions. Oral knowledge and its transformation is not neutral or entirely benevolent. I found out that the change from an oral to a written knowledge base takes power out of the hands of community elders and puts it into the public domain, where literate men have the advantage and where community security may be vulnerable. As scholars we need to face our own involvement in these political processes, even if we cannot ultimately control them.

A cornerstone of the Western intellectual heritage is the fervent belief in the power of the written word to transform man and society. In this tradition, the existence of writing serves as a hallmark for civilization and a marker to separate history from prehistory. While a great deal of scholarly work has dispelled many myths about literacy, thus bridging “the great divide” between the written and the oral, our intellectual and emotional attachment to writing persists. This appears to be especially the case in reference to the origins of writing systems, many of the latter being claimed and reputed to have been “independently invented.” For those peoples most involved historically in such developments, the invention and use of original scripts are points of pride, and hence claims for the “authenticity” of the scripts, that is, for their invention and coming into use having been an entirely indigenous undertaking, are passionately guarded.

Historians of writing, however, are cautious of claims for independent invention. From ancient to modern times, the history of the development of writing has been characterized by a balance between “independent invention” and “stimulus diffusion.” While epigraphers and paleographers attempt to unravel the inevitably obscure origins of certain ancient scripts possibly devised in environments free from external influence, no script devised in the last two thousand years is likely to have emerged totally independent of the stimulus of some diffused knowledge of the previous history of scripts—at the very least, the mere idea of writing. Nonetheless, for many modern observers, any suggestion of an outside stimulus on the development of such scripts is considered virtual heresy, tantamount to an attack on the intellectual ability of the peoples who claim to have single-handedly devised the scripts.

One problem facing any historical anthropologist or anthropological historian of Madagascar is locating sources that are scattered around the globe. A recent guidebook on African source materials in the United States makes important advances toward resolving this problem, notably for South African materials in the United States, but Madagascar is poorly represented there. Researchers of Madagascar can help each other connect the gaps between recognized research materials by sharing with colleagues what they have found at less well-known archives.

In the United States there are numerous Malagasy archival resources that may not be evident from a quick search through RLIN, Committee on Institutional Cooperation (CIC), or Center for Research Libraries (CRL) databases. Small archives may fall outside the range of microfilm projects, such as the Cooperative Africana Microform Project (CAMP), and therefore remain off the researcher's radar screen. Yet quite good materials, such as those at the Luther Seminary in St. Paul, Minnesota, may be relatively nearby and welcoming to the researcher.