During the mid-1950s British administrators in the Machakos District of Kenya enlisted categories of Kamba occult “experts”—“witchdoctors” and “cleansers”—to cleanse local “witches” and migrants from Nairobi who were believed to have taken the Mau Mau oath. A compendium of colonial documents concerning the “cleansing” campaigns illustrates how and why the socio-historical context of Mau Mau-era Machakos drove the colonial administration to break with its longstanding de facto policy of not officially combating supernatural challenges to state authority with supernatural means. The overwhelming disorder wrought by Mau Mau motivated state officials to break with precedent and to identify and employ Kamba “experts” to cleanse Mau Mau adherents and witches.

The widespread and politicized nature of the violence occurring during Mau Mau, and its perceived linkages to the supernatural, precipitated the state's shift to the employment of Kamba experts to combat “witchcraft” and Mau Mau oathing. An anthro-historical approach to understanding Mau Mau in Machakos shows that, while the cleansings constituted a group of “critical moments” at which British colonial officials could argue that they had dealt with supernatural challenges to state authority by rendering them “knowable,” the cleansings also demonstrated the degree to which state authority became situated in Kamba colonial officials and the extent to which the implementation and interpretation of British colonial cleansing policies depended on these local authorities.

In the early 1920s British West Africa saw a flurry of colonial activity, in which the formation of the colonial state—originally started in this region in the 1870s—was brought to a higher plane. The introduction of Indirect Rule in the newly-amalgamated Nigeria by governor Frederick Lugard called for a rethinking of colonial political and administrative structures. Where before, the relatively small administrative units were dominated by Europeans and western-educated Africans, now the position and role of “traditional” leaders was enhanced on all levels of colonial government. Control over the economy came more and more into the hands of European businesses and business conglomerates, at the expense of African firms. As a result, relations between African elites, who had vested economic and political interests in the colonial states, and the growing European colonial establishment hardened.

In the case of the Gold Coast, the African urban coastal elite of merchants, educators, missionaries, and others faced an overwhelming onslaught of change and modernization in all parts of society. In many cases these changes undermined the elites' social status, as well as their political and economic position. One of the weapons in the battle between British colonial authorities and African urban elite society was the written word. Within this context, expert knowledge about African achievements, molded in the form of biographies, was the two-edged sword of African cultural nationalists of diverse plumage.

The documents originated by the German colonial administration in German East Africa are located in two main archives: the Tanzania National Archives in Dar es Salaam, where they are identified under the name “German Records,” and the Bundesarchiv in Berlin, where they are collected under the classification R 1001. This note aims to provide some general information regarding a part of the German Records, referred to as “German Maps,” which is collected at the University Library of Dar es Salaam.

The German Records are a part of the holdings of the Tanzania National Archives, which also include the records of the British administration and various documents of the post-independence period. The German Records are a very well-known source for the history of the German presence in East Africa and they can be divided in two main categories: the documents of the Central Administration, cataloged with the numbers G 1-G 65, and the Private Archives, with the classification G 66-G 86. These records are very well cataloged and easily accessible thanks to the work of archival reorganization done by Peter Geissler between 1967 and 1969. His work was published in 1973 in a two-volume guide with the title Das Deutsch-Ostafrika-Archiv: Inventar der Abteilung “German Records” in Nationalarchiv der vereinigten Republik Tansania, Dar es Salaam. This guide offers a very useful overview of the records of the German colonial administration and is available for consultation in the Reading Room of the Tanzania National Archives. Also available in the Reading Room is a manual catalog which, in some cases, could be helpful in finding some documents that, owing to print errors in the edited catalog, have become difficult to find.

Nearly every historian of early African history has recently encountered studies that use the history of words as a source for history more generally defined, an approach also known as words-and-things. Indeed, by now a more or less elaborate use of words-and-things has become fashionable, especially in the Anglophone literature. A number of short presentations of the overall principles underlying this approach have been published, but they all lack an extended discussion of the methodological issues involved. Perhaps this is the main reason why words-and- things analyses are almost never subjected to critical scrutiny, while the conclusions of studies based on them, however weak or strong they might be, tend to be accepted as gospel—a most unsatisfactory situation. Hence a book-length study of the methodology involved in the application of words-and-things should be very welcome.

As Klein-Arendt's book is devoted exclusively to this subject, it should fill the gap. Yet it will disorient most readers of this journal because K-A is not concerned with the solution of smaller- or large-scale problems of history as understood by such readers, but focuses on the epistemology of the central European school known as Kulturgeschichte (more or less “Culture History”), to which he subscribes and which is likely to be largely unknown to most readers. Yet when the expression “words-and-things” was first coined in 1909. it was intended to be a tool to elucidate Kulturgeschichte. Indeed, the journal from which the label stems was called Wörter und Sachen. Kulturhistorische Zeitschrift für Sprach-und Sachforschung (Words and Things, a Journal of Culture History for Research in Languages and Things) and on its very first page Rudolf Meringer stated that “[linguistics is only a portion of the science of culture … We hold that the future of Kulturgeschichte resides in the union of the science of language with the science of things” in which “things” stood for what came to be better known as “culture traits.”

Recently in this journal (Wilmsen 2003:360-375), I delineated Charles John Andersson's route to the Okavango River based on entries in his diaries and an overlooked map of the route published by Thomas Baines in 1866. noted there that, while the place, Okambombo, where Andersson first stood on the banks of the Okavango was correctly plotted on Baines' map, its geographical coordinates were given only approximately. I also noted that Andersson did not bring along his astronomical instruments on his first excursion to the river, and assumed therefore that no coordinates for the place were obtained. However, in the course of my preparation of Andersson's papers for publication, I have since worked through the 116 unnumbered pages of geographical calculations recorded in volume 10 of the papers and found latitude observations for several points along the route, including Ombongo and Okambombo; his calculations for these latter two places are shown in Figure 1.

Andersson left the river on 29 March 1859 and returned to it on 29 April (CJAP-3.3:169); on 4 May the correct latitude, 17°46′08″S, was determined by astronomical observation. About this, Andersson (CJAP-3.3:172) wrote, “[o]btained a latitude. Between Ombongo & this place (Okambombo) I am only out a mile in latitude.” The latitude for Ombongo was obtained just before this, on 16 April on his way back to the river (CJAP-3.3:165). On Baines' map this latitude is given as 17°37′53″S, but as can be seen, this is just one of Andersson's interim reckonings before the “True Latd 17°39′38″” is calculated (figure 1). Baines (1866:217) gave approximate coordinates for Okambombo in his text, writing “17°30′S. lat. and longitude about 19°E,” but the correct coordinates (17°46′08″S about 18°55′E) are plotted on the map. Both discrepancies are due to the fact that the published map was redrawn by Royal Geographical Society cartographers, who copied one that Andersson made in Cape Town at Galton's urging (GP-UCL 1866) based on “upwards of three thousand prismatic compass triangulations and astronomical observations of latitudes” (found in volume 10), while Baines copied from the letter Andersson wrote from Okambombo to Dr. Norton Shaw, secretary of the Society.