7. Blood, Bugs, and Archives

Debates over Sleeping-Sickness Control in Colonial Northern Rhodesia, 1931–1939

This chapter is about the interaction of African ideas and imperial science. It argues that the very specific vampire accusations that emerged in the Northern Province of Northern Rhodesia in the 1930s involved local and colonial ideas about the relationship between wild animals, tsetse flies, authority, and citemene, the form of shifting cultivation specific to the poor soils of the Congo-Zambezi watershed. Neither African nor European ideas on these subjects were fixed, nor was one untouched by the other. Although I shall present European ideas and African ideas in sequence, I do not see them as separate and distinct. African experience with tsetse flies shaped European ideas about control of such insects, which were necessarily tailored to fit African realities.^[1] Indeed, it may be more useful to think of what follows as a presentation of scientific evidence (i.e., the kind of evidence we expect to find in essays about shifting cultivation) followed by a presentation of evidence of a very different sort (i.e., the kind usually considered inappropriate for historiography).

This chapter also suggests, with some trepidation, that the African ecological nightmare, whether disease or overpopulation, is in part a trope.^[2] Over the past 100 years, “science” and “medicine” have become ways of talking about Africa that embody ideas about disaster and renewal. The ecological history of Africa needs to incorporate data that will move scholars away from this paradigm. Sleeping sickness is a real and virulent disease, and my task here is not merely to identify a discourse but to describe and elaborate other visions in which sleeping sickness was seen as a manageable disease in Northern Rhodesia. The other visions are not expressed in the language of germs, parasites, or apocalyptic epidemics; they are expressed in the language of colonial departments, officials and assistants, and blood. There is a body of thought in cultural studies that claims that people not only debate the changes taking place around them, they debate the terms in which those changes are described.^[3] But to label either of these constructions “African” or “European” would be a mistake, I think, and I suggest that readers think of both of these as colonial constructions, in which the project and the materials are the same, but the position of the narrator is different.

But where are these narrators found? The sources I use here are from European archives: all the descriptions of vampires here have been mediated through the writings of colonial officials, colonial doctors, and the like. Over the past two decades, African historians have regarded such mediations with grave suspicions; the historian’s task was to find ways to hear the African voices submerged in such archives and to unveil the processes of inscription and recoding that constituted each mediation. In this chapter I look directly at the mediations, at the African ideas distorted in the sources, so that I can relocate African voices—and the vampires they talk about—in those archives. Archives, James Hevia and Gayatri Spivak remind us, do not merely report colonial activities, they report the elaborate colonial attempts to recode local space, local property, and local ideas into imperial terms. The making of a colony out of a variety of African landscapes and disorderly states required that it be reterritorialized, made into a new unit, with new maps and rules to fit the British empire.^[4] The official recoding of African spaces and ideas was as partial as it was elaborate. No land, no population, and no institution was ever made fully identical to the imperial categories in which it was placed, and no imperial codes were ever completely adopted by farmers, chiefs, and porters. The inability of the colonial state to fully recode and reterritorialize is revealed in colonial archives. Reading colonial archives to tease out African voices suggests layers of domination that can be stripped away to reveal a colonial subject buried beneath the imperial project. Another kind of reading might reveal intense struggles over domination in each archival reference to an incomplete recoding, or to another bungled reterritorialization. In such a reading, archives report the struggles, imperial and local, over the vocabularies and tools of domination. Evidence such as vampire rumors opens a space in which historians can accurately see the failures of recoding and the incomplete reterritorialization that was the practice of colonial rule.

But what kind of evidence is rumor, and how can I use it to move from the politics of representation to the politics of tsetse control? In its most positivist form, rumor is the officials’ term for information they have not engendered, shaped, or controlled. It is a category that simultaneously reveals popular conceptions about the actions and ideas of those in authority and declares the weakness of official channels of information and education. But what happens when I read rumors alongside naturalists’ studies and colonial biomedicine? I want to suggest that for academics at the end of the twentieth century, the differences between rumor and research reports are great; they are recounted in different media and they have completely different levels of credibility. But for the subjects of the research and of colonial biomedicine, rumor and our own notions of fact may not have been all that different. The Bemba language does not have separate words for rumor, talk, and conversation. Indeed, how rumor was distinguished from fact in the 1930s is not at all clear.^[5] Both covered the same ground, both contained the same actors and issues, but the rumor—at least as it was told and retold in colonial Northern Rhodesia in the 1930s—was often presented as a personal narrative. Rumors were not thought to be less believable if they were not first-person accounts, however: no one thought something untrue because it was said to have happened to a friend of a friend. Scientific knowledge, however, could be and frequently was disseminated in fragments, without the very frameworks that made it make sense.^[6] The vampire accusations of the early 1930s, for example, referred to activities in Tanganyika Territory that were said to be about to shift to Northern Rhodesia’s Northern Province, but were no less apparent—or frightening—to officials because of that.^[7] Official arguments about citemene, cassava, and deforestation, on the other hand, were often made without reference to African ideas about tsetse flies, ecology, and wildlife. Moreover, officials frequently anticipated that their arguments would be ignored.

The world in which black Northern Rhodesians lived seems to have had more varied forms of information than that of their British counterparts. Starting at least in the 1920s, Bemba-speaking peoples heard tales of a twig that could strip a man of his willpower and of “Kasai cannibals” who kidnapped African mineworkers. They heard that Catholic priests ate people. During the early years of the Depression, they heard that the king of England was in jail and that black Americans would come to replace the British. They heard that Europeans hired Africans to capture other Africans and take their blood; they heard the dangers of citemene.^[8] This is not to say everyone believed everything they heard. People believed stories—even if belief was not a constant state—because of how they appreciated and apprehended certain facts, not because a story was grandiose, frightful, or told orally. From the vantage point of a Bemba village, belief in “rumors” and “facts” appeared to be equally tentative. Audrey Richards dismissed muchape, the transnational witch-finding movement of the early 1930s, as precisely the kind of novelty the Bemba took up and quickly abandoned.^[9]

But colonial science was not the mirror image of an African intellectual faddishness. Colonial science was anything but a monolith; officials continually argued with the state and one another about forests, wild animals, and African agriculture. “Scientific research” had a credibility in colonial circles that the eyewitness accounts of naturalists did not have. But rather than evaluate various trends in colonial thinking, I want to find a way to interpret them all, as representing different visions of the world and ways to understand it that changed over time. Recent trends in literary criticism have argued that it is worthwhile to read scientific texts the way we read novels, as cultural products that reveal the concerns and anxieties of a specific milieu.^[10] In this chapter I suggest that it is possible to read the fictive as the same kind of historical source as scientific texts.

This chapter is about the mosaic of colonial beliefs, African and European, the supposedly superstitious and the supposedly scientific, about sleeping-sickness control. It argues that these beliefs, like so many tiles, can be placed alongside one another so that an observer can discern the different narratives of science, land use, and medicine and see how no single vision of fact and consequence ever fully dominates another. What follows are two discrete histories, one of pathogens, the other of vampires. Without oral evidence, this may be the only way I can proceed. The questions of who is saying what, when, and of who repeats which rumor with intense belief and who argues against it with equal passion—the very evidence that makes rumors form a debate rather than a monolith—is not discernible from archival sources. Where individual African viewpoints appear, they do so at the behest of colonial authorities, so that I am hesitant to read the words of an African writing for a government newspaper or a district clerk’s words as anything more than those of a man doing his job. What follows, I hasten to point out, is a very conservative interpretation, in which I have stayed very close to my documents. This exegesis is based on a reading that could best be called vampire-driven: the questions I have asked and the files in which I have sought answer to them have all been determined by my reading of the Northern Rhodesian vampire accusations of 1931.

Bug Stories

Tsetse flies carry the protozoa, called trypanosomes, that cause sleeping sickness (trypanosomiasis) in humans and domestic livestock. There are two kinds of trypanosome and two kinds of sleeping sickness, the origins and nature of which are by no means agreed upon: some think these are different environmental responses, others that the structure of the trypanosomes differs. Thus, the two kinds are either called by the names of the protozoa—Trypanosoma gambiense and Trypanosoma rhodesiense—or by the environments in which they occur, riverain and savannah. The terminologies of both types of sleeping sickness involve hosts (sometimes called the reservoir), vectors, and ecologies. The vector is the only method of disease transmission, as the trypanosome transforms in the fly’s body over several days to become infectious. In T. gambiense, infected flies live in the shade on riverbanks and feed off humans, or occasionally reptiles, and infect them; because the disease can be transmitted from human to human, it can be spread by relatively small numbers of flies. Humans are the hosts, flies the vector. In T. rhodesiense, tsetses live in wooded areas—the bush—and feed off wild animals, which do not become infected, but they can also feed off humans or domestic ungulates when they are available: wild animals are the hosts, and the flies are the vector. Entomologists—amateur and professional—have tended to ignore the protozoon for the fly and studied the behavior of various species of tsetse in order to show how different varieties of trypanosomiasis are spread and how different ecosystems encourage that spread. Sleeping-sickness control organizations in British Africa invariably included entomologists.^[11] Protozoologists, who seem to have been more influential in francophone Africa, regarded the differences between the trypanosomes as crucial and saw T. gambiense as an entirely different disease from T. rhodesiense.^[12]

The “discovery” of sleeping sickness was truly a colonial phenomenon. While the disease had been known in West Africa for centuries, its spread in the havoc of colonial conquest to previously uninfected regions—the Congo River basin and Busoga in Uganda are perhaps the most dramatic examples—created epidemics of apocalyptic proportions. The other discovery, of the cause and etiology of the disease, is one of the great stories of tropical medicine, combining all that was exotic about epidemics in Africa with all that was memorable about scientists’ and explorers’ egos.^[13] It was a discovery that would not have been possible without the scientific advances of the late nineteenth century, particularly germ theory. Germ theory made the debilitating diseases of the tropics avoidable; they were not caused by the gaseous matter of climate and decaying organisms (miasma), as had been previously thought, but by protozoa and bacteria, which could be conquered as they had been conquered in Europe.^[14] But as Maryinez Lyons has argued, germ theory had its drawbacks. If the miasma theory had related tropical diseases to their geographical location, the bacteriology and protozoology of tropical medicine alienated disease from the landscape.^[15]

But the sleeping sickness of this grand tradition was T. gambiense; the discovery—or invention, depending on whether one stands with the protozoologists or the entomologists—of T. rhodesiense was pursued with far less excitement and even some trepidation, as researchers concerned themselves with identifying an etiology and relating its cause to the trypanosomiasis of domestic stock, called nagana. T. rhodesiense was difficult to identify in part because local doctors expected humans to develop T. gambiense and in part because victims sickened and died so rapidly that Africans only identified the last stages of the disease, and then only for adults; presumably children succumbed so rapidly that sleeping sickness was confused with other afflictions. It was only in 1912 that the Luangwa Sleeping Sickness Commission, headed by investigators from the Liverpool School of Hygiene, demonstrated that the trypanosome carried by Glossina morsitans could feed off wild animals and humans alike.^[16] Research in Nyasaland and South Africa in 1913 showed that T. rhodesiense was identical to T. brucei, discovered by David Bruce in Natal in 1894, the cause of nagana.^[17] Not everyone accepted the idea that T. rhodesiense was caused by the trypanosome of wild animals and domestic livestock, but the fact shaped sleeping-sickness and tsetse-control policies in the 1930s.

In areas where T. gambiense was prevalent, attempts to control sleeping sickness became attempts to control populations—either by restricting their movements, by isolating the sick, or by removing whole villages.^[18] But areas where T. rhodesiense was prevalent were, according to the thinking of the times, areas where cattle keeping was impossible, so that attempts to control sleeping sickness became attempts to control land use and relations between humans and wild animals. There were never as many cases of T. rhodesiense in East Africa as there were of T. gambiense on the riverbanks and lake shores of East and Central Africa. T. rhodesiense was more virulent, but since it was carried from animal to human, rather than from human to human, it was far less contagious. For that and for economic reasons, there was far more concern about nagana in East and Central Africa than there was about human sleeping sickness. Studies of T. rhodesiense tended to be centered on cattle rather than people. As late as the 1950s, when livestock losses from trypanosomiasis were less than those from rinderpest, tsetse flies and the fear thereof prevented profitable land use.^[19]

In the case of sleeping sickness, the politics of land use was mediated through the new discipline of tropical medicine. In his history of yellow fever, François Delaporte charts the origins of the field. It mapped the interactions of living things to arrive at pathologies and in doing so, imbued insect vectors not only with the power of life and death, but the power of science: they could be controlled by knowledge about them. Tropical medicine drained one ancient symbol of its meaning and replaced it with another: “[D]eath came not now in the form of a man with a scythe but of a biting insect.” ^[20] But if germ theory simply swept the miasma-ists away, along with their intimate sense of peopled locations, parasitologists swept the bacteriologists out of British tropical medicine. “Non-tropical” bacteriological diseases were ignored, and tropical medicine concentrated on worms, insects, and protozoa. The link between parasitology and tropical health convinced experts that these diseases could be prevented without studying how local populations became ill. Much colonial health policy focused on protozoa and vectors. Insect vectors and animal hosts were where protozoa spent part of their lifecycles, and killing the insect or animal could kill the protozoa.^[21]

In the case of T. rhodesiense and G. morsitans, this pitted tsetse control against a vocal hunting lobby and one faction of imperial science. When in 1913, for example, David Bruce was convinced that T. rhodesiense and the trypanosome that caused nagana were identical, he became an even stronger advocate of the extermination of wild animals than he had been previously. His reasoning had to do with ideas about what formed an infectious reservoir and how best to control it. Questioned before a 1913 Colonial Office Sleeping Sickness Committee that had many hunters on it, Bruce was asked why wild animals were the host for infectious T. rhodesiense, rather than birds, immune herds of cattle, or even people. The birds in G. morsitans country were too small and too mobile to be a good source of food for tsetse flies, there were no herds of cattle in G. morsitans country, and humans made poor hosts for T. rhodesiense, because only a few were infected and those were too sick to travel about and spread the disease, he responded.^[22] He strenuously opposed the preservation of big game in “fly country”: “It would be as reasonable to allow mad dogs to live and be protected by law in our English towns and villages.” ^[23]

Although some suggested localized experiments in game eradication,^[24] few of Bruce’s contemporaries agreed with him about the relationship of T. rhodesiense to wild animals. German scholars disputed his findings; Alward May, Northern Rhodesia’s medical officer, disregarded the findings of the Luangwa Sleeping Sickness Commission and claimed that man was the principle reservoir for T. rhodesiense; E. E. Austen of London’s Natural History Museum argued that tsetse were specific to certain habitats: these could be emptied of people and left to game, inasmuch as tsetse flies did not follow game.^[25]

Such debates about vectors and hosts, about flies and buffalo, were debates about how to classify and categorize animals. Such classifications were and are as much a part of scientific research about animals as they are artifacts of “traditional” society.^[26] In the 1920s and 1930s, what was known about T. rhodesiense was the supposedly contagious relationship between reservoir, vector, and victim. Attempts to study the specific relationships—human to landscape, human to animal—that might cause or limit the disease gave way to the study of a vector abstracted into “the fly.” While it is tempting to suspect that this was the result of pressure by the hunting lobby, it seems more likely that it was part of the intense focus on vectors and pathogens that characterized early research in tropical medicine.^[27] Thus, the very people studying fly-human or fly-animal interactions anthropomorphized tsetses—saying, “The Tsetse fly loathes the presence of man,” ^[28] for example—and the fly became as important in research as was the disease. By 1935, there was a Parliamentary Tsetse Fly Committee.

C. F. M. Swynnerton, the most important tsetse researcher in this story, understood that fly behavior was based on human observations. A naturalist of extraordinary capability, Swynnerton had come to Africa as a nineteen-year-old farm manager and first attracted attention with his study of a mixed fly belt in North Mossurise, on the Southern Rhodesia–Mozambique border, in 1921, in which he noted, among other things, that male flies sometimes traveled on humans, causing some observers to think they were attacking.^[29]

Swynnerton’s studies of the tsetse fly’s ecological niches, including his 580-page monograph The Tsetse Flies of East Africa (1936), which catalogued the various species’ eating, breeding, and resting habits, put the fly in the foreground. Swynnerton read the landscape to show how tsetse could be limited without the wholesale slaughtering of game. Knowledge of the fly, Swynnerton argued, would allow science to combat the vector without significantly disrupting the reservoir or destroying the hosts—a method of disease control that David Bruce had characterized twenty years before as “a nice pious wish.” ^[30]

Land, Flies, and Science

Different varieties of tsetse fly live in different places. In East and Central Africa, the colonial concern was about G. morsitans, the fly that carried the trypanosome fatal to domestic livestock, and, to a lesser extent, humans. While the riverbank-dwelling T. gambiense could be transmitted from human to human, the trypanosome carried by T. morsitans required an animal host, so that methods of describing and of preventing one kind of sleeping sickness came to be about the relations between humans and animals. If sleeping sickness of the savannah was carried by wild animals, then the goal of biomedical policies was to separate humans and wild animals, big and small.

From these policies came studies of “the fly.” By the 1930s, most scientific knowledge of G. morsitans was based largely on Swynnerton’s research, which demonstrated the viability of African methods of tsetse control. Nevertheless, the major impact of Swynnerton’s work both on his own career and on the shape of tsetse research was to suggest an either/or paradigm in which centralized settlements and tsetse flies were inexorably opposed.

Swynnerton showed how knowledge of various tsetse flies’ behavior could be used to control their numbers and habitats. G. morsitans, for example, breeds on barren ground toward the end of the rainy season; G. brevipalis lives in wooded undergrowth that remains in leaf throughout the year; all tsetse flies need shade. Well-timed grass burning could therefore limit the habitats of two species of Glossina. Ngoni in North Mossurise had burned grass late in the dry season, when leaves had fallen and the grass was at its driest, so that the fires would be intense enough to draw a wind. Such a fire would not only destroy the grass but much of the young growth and some high shade; with sufficient rain, however, the grass would rapidly grow back. Swynnerton became a proponent of late burning, a method of tsetse control that he believed white settlement had greatly disrupted: “Under the white man everyone burns as he pleases.” White farmers’ uncoordinated grass burning failed to check tsetse populations, and different species of fly flourished.^[31]

The behavior of wild animals, according to Swynnerton, was shaped by human intervention as well. Under Ngoni domination, large parts of the Central African countryside had centralized states with concentrated populations. Densely populated areas and mile after mile of cultivated fields surrounded by deforested areas allowed Africans to live and keep cattle in health.^[32] When the population decreased, or when an area was raided and the population scattered, the land reverted to bush and game, and tsetse became widespread—a medicalized version of tribal warfare. Swynnerton had been very impressed by Ngoni accounts of methods of tsetse control. When Umzila conquered North Mossurise in the 1860s, the somewhat scattered population lived near belts infested by G. morsitans; cattle had to be sent to highlands or they died. But Umzila ordered his population to draw near the king, moving villages and settlements to the lower altitudes of the territory. “Every one of my informants has described most graphically the result of this concentration,” wrote Swynnerton. “The bush simply disappeared and the country became bare, except for the numberless native villages…and gardens.” All that was left of the woodlands was an uncleared and uninhabited “Oblong,” virtually a game reserve for Ngoni hunting parties. Outside the Oblong, hunters tracked wild pigs and buffalo herds whenever they appeared.^[33]

It was on the strength of this research that Swynnerton was appointed the game warden of Tanganyika Territory in 1921; his goal was to control tsetse flies without the wholesale slaughter of game, the policy already in sporadic operation in Southern Rhodesia. In 1923, he chose Shinyanga for the site of his research, where he was to experiment with ideas about competing ecosystems, bush and wild animals, and cattle and cultivation—mainly through bush clearing and centralized settlements.^[34] Shinyanga had been the site of a major epidemic in 1923, but Swynnerton had selected it because of its particular cycle of retreat and advance of tsetse flies: cleared land was free of tsetse, but when the population declined or moved on, the tree roots sent up shoots on which tsetse flies from adjacent infested bush alighted and then traveled to human settlements on passersby.^[35] In practice, however, Swynnerton’s subtle analysis of the local landscape was overwhelmed by the sheer scope of widespread bush clearing, which by the mid 1920s required a levy of almost 8,000 men.^[36] Swynnerton has been contested academic terrain in recent years. Although John Ford praises his attention to seasonal details, John Iliffe sees him as a harbinger of soil erosion, and John MacKenzie as an agent of the hunting lobby.^[37] A close reading of Swynnerton’s work reveals the complexities of daily life in sleeping-sickness areas, however, rather than an ignorance of rainfall patterns or apologies for bushpig and buffalo. During the Mwanza epidemic of 1922, for example, he suggested that T. rhodesiense could be transmitted by man-to-man contact, based on his observations of the division of labor between sick and well Sukuma in their households. The absence of animal vectors, however specific, was taken up by the hunting lobby, but not because Swynnerton was their mouthpiece. He was a keen hunter and very close to the hunting lobby, but, as we shall see, from 1923 until his death in 1938, he proposed a variety of methods of tsetse control, including the seasonal and the agricultural. Time after time, however, the proposals that were implemented were those that conformed most closely to official agricultural policies.

The importance of the landscape to the location of tsetse flies made epidemiology and land use overlap. A new dimension, population density per square mile, one of the most important markers of scientific discourse in this story, became the cause and cure of sleeping sickness. There might be tsetse flies in areas with population densities as low as one person per square mile, but that was not sufficient to sustain an epidemic of sleeping sickness. Areas with population densities of five per square mile would not have enough cleared land to prevent tsetse flies, and they would have epidemics. Where there were twenty-five people per square mile, there would be enough cultivation and tillage to prevent tsetse advance. Best of all would be a hundred people per square mile, a population density that could crop and clear a fly-free area in which Africans and cattle could live in health.^[38] In 1930—a year after Swynnerton had resigned to start the Department of Tsetse Research—Tanganyika Territory attempted to establish large, compact settlements, as if population density, in and of itself, could combat tsetse flies and sleeping sickness.^[39] Concentrated villages, like tracts of cleared bush, encouraged soil erosion, however; Swynnerton’s biomedical theories of the 1920s became the environmental terrors of the 1930s and 1940s.^[40]

In tsetse research, the science enshrined in population ratios was translated to tsetse populations, which were measured according to the numbers Swynnerton’s African assistants could catch. In the early 1920s, the “fly boys” of Tanganyika Territory stood still in the bush and recorded the number of flies each caught per hour. The resulting figure—flies per boy/hour—was later rejected by two of Swynnerton’s entomologists, W. H. Potts and T. A. M. Nash, as unscientific.^[41] To time discipline, they added the discipline of distance. Potts divided the Shinyanga bush into sections according to vegetation; each fly boy was assigned a section and would then walk along a path, stopping to collect flies every twenty or hundred yards, to establish the density of flies per boy 100 yards or, where flies were densest, flies per boy/yard. These fly rounds became increasingly complex, precisely laid out in grids or octagonal spirals to compensate for seasonal variations in flies’ whereabouts or the number of flies scared away by these activities. Even so, in 1930 a zoologist, C. H. N. Jackson, published a critique of the flies per boy/mile measurements, writing that they did not systematically accommodate flies’ eating habits.^[42] By the early 1930s, if not before, the very extent of fly rounds—in some areas there were thirty miles of paths—was thought to have disturbed mammalian hosts, so that tsetse and animal populations moved elsewhere.^[43] In Northern Rhodesia, which had no official tsetse-control organization comparable to that of Tanganyika Territory, fly rounds were considered a viable research method by district officers well into the 1930s: “In order to survey the density of the fly an African (immunized by injection, of course) walked along the bush path with a white cloth pinned to the back of his shirt. A man with a notebook walked behind him and counted the flies which settled on the cloth. The result was later recorded in a graph as ‘Density of fly per boy mile.’” ^[44]

Also in Northern Rhodesia, human population densities had more than a biomedical meaning. They offered ways both to understand and to problematize citemene, the very productive and reliable system of shifting cultivation taken up by hoe cultivators when they entered the area. The Bemba, who engaged in widespread raiding until the end of the nineteenth century, had lived in large, stockaded villages. Once they too took up citemene, a very specific pattern of settlement and land use developed. Citemene cultivators scattered to mitanda, the garden huts families lived in during the growing season: by 1904, for example, no villages could be seen in Mpika.^[45] In large-circle citemene, the system of the Mambwe and the Bemba, trees are lopped, not felled, and pollarded trunks are left at chest height. Branches are carried for miles to form a large circle, often of about an acre, in a clearing and left to dry in the sun. Burning takes place late in the dry season: too early and the ash would scatter in the wind; too late and the wood stacks would be wet and the burn incomplete. After the branches are burned, the large ash circle is planted with a sequence of crops, starting with finger millet the first year. Burned-over land is sometimes cultivated for as long as five years.^[46] Citemene depleted the woodland—burned areas made up from 4 to 10 percent of deforested areas—and the intense heat of the late burning destroyed the forest canopy and retarded the growth of new trees.^[47] Nevertheless, it was burning that transformed cut branches into a garden. Burning was the only occasion Richards observed in which the Bemba acted as “one economic unit.” ^[48] Citemene was also a system of great productivity: experiments at the research station in Lunzuwa compared yields from Mambwe citemene fields and Mambwe hoed mounds: from 1935 to 1940, citemene gardens produced at least three times more finger millet per acre than hoed gardens.^[49] The official construction of citimene in the 1930s, of primitive cultivation, performed with an ax, gradually gave way to an official and scientific understanding that citimene was combined with permanent gardens and was both adaptable and productive.^[50]

However productive it might have been, citemene was considered an administrative and agricultural nightmare. Whatever it did to forests and ecosystems, it made ordinary Bemba hard to rule and harder to tax. In large-circle citemene, gardens were scattered and temporary, and for much of the year families lived in mitanda. From the early years of colonial rule, mitanda dwellers guarded their independence fiercely: they were “malcontents who renounced the authority of the chiefs and the Boma.” ^[51] Mitanda became sites of new community and social relations for Bemba.^[52] In 1906, the British South Africa Company, which then governed Northeastern Rhodesia, banned citemene. Widespread hunger followed. By 1908, the Chitimukulu, the Bemba paramount, complained of his people’s starvation and poverty. Administrators complained of a new, menacing attitude of opposition among the Bemba, and expressed some anxieties about the newfound unity of chiefs and commoners—against the boma. Citemene was restored in 1909, and in 1910 administrative units were made smaller to accommodate mitanda.^[53]

Objections to citemene did not cease, but they ceased to be described in administrative terms. The problems with citemene became agricultural, not political: environmental degradation replaced social disintegration. By the late 1930s, a new official line had emerged, claiming that while fly-infested areas had long existed, they had also been uninhabited, but as the security of Pax Britannica went on, villages broke up and smaller villagers formed, and Africans “drifted” into fly areas, just as the bigger villages ceased to function as fly barriers.^[54]

The scientific study of African land use and population densities gave expression to colonial anxieties about deforestation, late burning, and Africans’ relationship to authority. But colonial anxieties, even those sanctioned by science, were not uniform. The biomedical view was that the greater the population density, the fewer renewable resources, whereas entomologists believed that the greater the population density, the fewer the tsetse flies. Population density statistics were used in Northern Rhodesia in the 1930s to show the damage done by citemene and, sometimes, land alienation. Officials bandied grim figures for population per square mile to proclaim ecological doom, to be sure, but also to participate in scientific discourse. None of the figures so pronounced took male migrancy or environmental variations in the landscape of the plateau into account.^[55]

Animals, Flies, and Officials

Throughout the 1920s and 1930s, many officials thought that the low population densities of citemene encouraged garden raiding by elephants, which was said to cause “serious starvation” in Lundazi, for example, starting in 1919, when elephants had first come into the area. In 1922, villages in Luwingu, Lundazi, and Abercorn were said to have no food.^[56] Africans were too poorly armed to fight off garden raiders themselves, and after much debate, the district commissioner in Abercorn adopted the policy already in place in southwestern Tanganyika and gave a white hunter free license to shoot elephants in inhabited areas.^[57] Soon after, many white hunters wrote to the boma requesting employment on Tanganyika Territory terms. Throughout the 1920s, the basic structure of authorized elephant hunting remained the same. A hunter was appointed elephant control officer and was allowed to kill any elephant tracked from garden raiding. To avoid the authorization of ivory harvesting, it was specified that elephants had to be caught in the act or shortly thereafter, but according to Norman Carr’s description of elephant control in Nyasaland in the 1920s, this rarely happened. A hunter “went out, as I did, and shot the first elephant he saw with reasonable tusks and called it a garden raider.” ^[58]

The hunter’s fee for shooting an elephant was, with variations according to weight, to keep one tusk. The other tusk went to the boma and the meat went to the villagers, distributed in theory according to local hierarchies, but in practice at the discretion of the hunter.^[59] The importance of this meat is clear. Audrey Richards provides a compelling description of the meaning of meat to Bemba villagers in 1931. They claimed the meat from an antelope she had shot gave them energy, “not only before the food was digested, but before it was cooked!…The next day they went to work early, declaring that their arms were strong.” ^[60] Indeed, 1920s elephant control almost at once became a local source of meat, the sale of which proved almost as profitable for hunters as the sale of ivory. One hunter worked out complex equations by which he might profit most: he would not shoot young elephants; he would keep all tusks under thirty pounds or 35 percent of tusks under fifty pounds, or he would be allowed to sell the meat without a butcher’s license.^[61] Indeed, false reports of raiding, designed to bring a well-armed hunter into an area, were common, and attempts to withdraw hunters or verify complaints were unpopular: chiefs tended to encourage false complaints whenever continual hunger eroded their authority.^[62] By the late 1930s, for example, hunters readily acknowledged the pressures on African game rangers: they were subject to “a good deal of temptation” if posted to their own areas and were frequently used by their chiefs as suppliers of meat.^[63] Even elephant culling became a food issue, which then seeped into issues of race and propriety. Once a rudimentary Game Department was established, it was loathe to hire African game scouts, because “the villagers with whom they stay are often naturally keener on getting meat than on genuine crop protection,” and lobbied for an increase in white rather than African personnel.^[64] African game scouts may have seen the situation somewhat differently. According to Norman Carr, “A ‘fundi’…with tons of meat at his disposal became the most popular man in his community.” ^[65] For these reasons, most conservationists were uneasy about elephant control.

The official emphasis on population densities and deforestation meant that citemene was thought of solely as a system of shifting cultivation. It did not take into account the ways in which citemene engendered relationships with government hunters that could provide valuable sources of meat and prestige for chiefs. Nor did it take into account the late burning that citemene required, and the decrease in G. morsitans populations that late burning caused year after year. Not only did late burning destroy the shade tsetses needed, but grass became dense in the burned areas during the rains, destroying G. morsitans’s breeding grounds. Moreover, the barren areas produced by citemene created barriers to tsetse flies.^[66] Despite some hints in tour reports, citemene was on the whole not considered a tsetse-control measure. In all probability, this was because sleeping sickness, particularly the variety caused by T. rhodesiense, was seen as a cattle disease, the absence of which was best demonstrated when people kept cattle. As a cattle disease, T. rhodesiense implied a specific landscape, free of bush and big game, occupied by cultivators and their herds, a paradigm so powerful that scholars and scientists were disinclined to see T. rhodesiense in the same terms in which non-cattle-keeping Africans might have seen it.

Game preservation had been under attack for years for causing the spread of sleeping sickness. As early as 1911, a game park bordering Nyasaland and Northeastern Rhodesia—named, appropriately enough, Elephant Marsh—was closed because of missionary complaints about the increase in tsetse flies and sleeping sickness.^[67] By the 1930s, game parks created another layer in the either/or paradigms of tsetse control, adding the issue of people versus animals and tsetse to a landscape already mapped by official thinking of cultivation versus tsetse. The landscape that emerged from debates about game preservation and game parks was mandated to house animals for European sensibilities and hunting, rather than those of Africans. Animals in game parks, as in some tsetse-control schemes, became an undifferentiated category of prey in which hunting was racially and technologically specific, although assisted by Africans.^[68]

Where did these biomedical ideas that proposed either the destruction of fauna or the destruction of flora come from? The line between African ideas and imperial science was never a sharp divide. Swynnerton’s bush-clearing campaigns had originated in Ngoni practices of land use, relations between domestic and wild animals, and authority: in Shinyanga, he tried to make them scientific and experimentally sound.^[69] The game preservation of tsetse research and reclamation of the 1930s had European as well as African antecedents, however, and these blurred the lines between hunters, conservationists, and scientists. The creation of game parks and game preserves, for example, came out of two contradictory turn-of-the-century motivations: to safeguard the natural world, in manageable proportions, so that the modern world would not lose touch with it, and a desire to subdue that same natural world, regardless of whether or not such actions had any social context. White hunting in Africa had a very different meaning in 1885, for example, than it did in 1935. Early game parks were created to be “unpoliced spots on a map,” to be protected from African exploitation. Early colonial game parks were places where African hunting was illegal. They preserved hunting land for foreign hunters—many of whom were museum collectors and scientists as well.^[70] In the early years of game preservation, wild animals were not romanticized, white hunting was.^[71] Within the next thirty years, game parks became sites in which decreasingly important imperial interests could assert their power against growing biomedical and social lobbies.^[72]

Blood Stories

In June 1930, Major Hingston of the Society for the Preservation of the Fauna of the Empire—the ideological wing of the hunting lobby—visited Northern Rhodesia as part of his African tour. Northern Rhodesia was, however, a special concern because it was the only British colony in Africa without a game warden. Without a game warden or national parks, there was nothing to prevent the excessive slaughter of game. Officials agreed, but noted: “All agriculturalists, whether European or native, are equally anxious to prevent game damaging their crops.” Hingston proposed that a game department and game parks be established; he dismissed the idea that African nutrition would suffer if African hunting was limited by citing evidence from Uganda and Kenya, and noted that in Tanganyika Territory, Africans were not subject to game laws unless they indulged “in barbarous practices or wholesale slaughter.” By March 1931, many officials concurred with the recommendations, which had proposed game parks in areas considered unsuitable for European settlement.^[73] Also during the first six months of 1931, and between April and August 1932, C. R. S. Pitman, then game warden of Uganda—which had East Africa’s model game department—visited Northern Rhodesia to prepare a report on wildlife there. He traveled 8,000 miles and corresponded with white hunters in twenty-seven districts. He proposed game regulations and licensing quotas, based on his game census, and recommended the creation of an official elephant-control department and game parks from which a “meat-eating armed population” had been removed.^[74]

Northern Rhodesia may have been the only British colony in Africa without a game warden, as Hingston noted, but it was also the only British colony where Africans understood “game ranger” to mean vampire. In March 1931—during Pitman’s first visit—the provincial commissioner in Kasama District first wrote about banyama, a word that combined Swahili and Bemba to mean “people of the game, or meat.” That banyama (singular, munyama) was never elided into its literal Bemba form reveals how both the term and the concept were maintained as a neologism, something foreign and new.

Science, Flies, and Land

Rumors about banyama seemed “to have arisen from a perverted notion in the Native mind as to the function of the Tanganyika Territory Game Department,” which had been so fully recoded and reterritorialized that the banyama were said to consist of “large bands” of Africans under the charge of European officials. These officials wore khaki uniforms and helmets with a badge with a small antelope’s horns on it; what the African staff wore was not known. These banyama were said to make their camps in the bush where Africans, “expert in the art of tracking and hiding…are sent out to murder any native, male or female, found alone,” an official report noted. After killing their victim, the banyama drained his or her blood and, “by making an incision behind the ear, extract a certain portion of the brain. The body is left in the bush, and the blood and brain forwarded to the Medical Department to be used as medicines in hospitals and dispensaries.” ^[75]

The rumor had been circulating in the district for about three years and became particularly intense early in 1931, when Africans returning from Tanganyika’s sisal plantations claimed that “the natives of Tanganyika Territory had found effective methods of frustrating the efforts of banyama who were accordingly being sent to Nyasaland and Northern Rhodesia in search of victims, the employers remaining in Tanganyika Territory.” ^[76] It was said that the Tanganyika Territory Game Department aroused suspicion because the game wardens “‘do not walk along the paths like normal men, but wander through the bush like outlaws.’” Moreover, the insignia with the buck’s head may have reminded many Bemba in Northern Province of witchcraft, as the skull of the buck is often associated with those practices.^[77]

As we have already seen, much of this was true. The men of the Tanganyika Territory Game Department did wander through the bush on foot or bicycle.^[78] Their purpose was not to hunt Africans but to count tsetse flies in units unique to imperial systems of measurement and discipline. Tanganyika Territory’s fifteen African fly collectors wore dark blue puttees; Swynnerton’s chief fly boy, Saidi Abdullah, wore a brass buffalo head on his pillbox cap. According to T. A. M. Nash, Abdullah was a superb naturalist who knew trees and plants and could track animals. He discovered the species of tsetse fly that was named for Swynnerton. In the Mwanza sleeping-sickness epidemic of 1922, Abdullah ferreted out victims who had been hidden by their relatives; it was said that he took blood smears from corpses.^[79] He and many other tsetse researchers, African and European, diagnosed sleeping sickness by taking a sample from lymph glands, hence the “incision behind the ear.”

But what did fly boys do to earn this reputation? There is no documentation regarding their actual activities during tsetse control: entomologists and officials tended to write about what they intended to do, not what transpired. In the absence of such data, however, we should not assume that each and every fly collector was overzealous in his performance of his duties. Such an assumption would presume a uniform identification with a job and a regime for which there is no evidence; it would assume that Africans misunderstood, rather than recoded, imperial practices. Indeed, the question of what these men did on the job may not be the best way to approach the complicated and often contradictory origins of vampire rumors. There was probably not a simple correlation between an action and its fantastic representation as banyama; it seems more likely that banyama represented the often puzzling meanings of activities in which counting flies and protecting animals or prohibiting agricultural practices fused.

The association of Tanganyika’s Game Department and vampires was specific to the Northern Rhodesia. In August 1931, officials in Morogoro, Tanganyika Territory, identified surveyors as those thought by Africans to drink their blood, a group referred to by the older Swahili terms cinjacinja or mumiani.^[80] It was not until the mid 1930s that whites in Tanganyika began to hear of “Bwana Nyama,” the veterinary or game officer who went alone into the bush to look for blood.^[81] Although individuals may have been called that during the 1930s, no form of bwana nyama took hold as a collective term for vampire in colonial Tanganyika, which remains mumiani or chinja-chinja to this day.^[82]

The Northern Rhodesian rumors continued well into 1932, but there were no other official alarms about their import until 1936. Then it was the tsetse-fly pickets, men stationed on paths to physically remove flies from travelers leaving infested areas. How fly pickets came to be considered a viable method of tsetse control discloses many of the concerns and blind spots of colonial thinking, both about tsetse flies and about African agriculture, in the mid 1930s. The language of tsetse control and the language of citemene control both described a landscape that was either usable or had to be abandoned—the simplified either/or paradigm of Tanganyika’s Game Preservation Department, headed by Swynnerton, in the mid 1920s. Where the boundaries between infested and uninfested landscape were violated, or ignored, cattle and people sickened and died. In Northern Rhodesia, some cattle deaths were the result of ambitious attempts at cattle keeping in known tsetse areas, so that when 2,000 head of cattle died in the Luangwa Valley in 1930–31, no one paid much attention. When cattle became infected in the formerly fly-free area of Isoka in 1932, “notoriously bad” African husbandmen were blamed for bringing the disease into the area, and the cattle were left to die.^[83] But for European coffee growers in Abercorn—the district of greatest labor migration to Tanganyika—already reeling from a drop in coffee prices, cattle keeping was a necessity. Their industry depended on successful mulching and manuring, and their farms bordered on fly-infested areas, so they needed either to keep their cattle free of sleeping sickness or to import fertilizers.^[84] When in 1932 one settler had lost half of his cattle, he blamed the increase in fly numbers on motor transport from fly-infested areas. His complaints to the governor met with the stern response that in impoverished times, Abercorn farmers had to accept the risks of having plots in fly areas.^[85] After two years of discussion the Abercorn Coffee Growers Association invited Swynnerton, then of the Tanganyika Territory Department of Tsetse Research, Shinyanga, to come and investigate Abercorn—one of the areas suggested by Pitman as an elephant reserve—on his way to a hunting holiday in Nyasaland.^[86]

The main work of the survey was done by Swynnerton’s two head fly boys, including Saidi Abdullah, who preceded him to Abercorn. Their report was finished at the end of October 1935. It showed that the planters had good reason for alarm: the coffee area was to the west of a “cultivation steppe”—plains and glades that had been grazed bare and hills cut for agriculture, the result of the close settlement of African reserves—that would be an ideal habitat for G. morsitans, should it ever reach there. Before Africans were removed into their present reserves, their cultivation had provided a natural barrier to the fly. Because of years of slash-and-burn agriculture, the low trees and dense shrubs were among “the best” example of miombo woodlands Swynnerton had ever seen. The regular burnings of grass had maintained semi-open patches, vipya, on the ridges and hillsides. The creation of African reserves in 1930 had not only destroyed that barrier, it had provided a good source of food for G. morsitans.

It was the relationship of land use to fly feeding and breeding that Swynnerton and his team examined. G. morsitans is associated with miombo but cannot survive in “large stretches of homogenous wooding, even if miombo.” Instead, it requires miombo-like wooding “in which to lie up when fed and in which to breed” and “glades and dambos interspersing it” that provide good visibility and animal life on which to feed “every two or three days in the late dry season and every dozen days in the wet.” Any semi-open country is thus hospitable to G. morsitans when interspersed with miombo, but only if it does not cover large stretches of the countryside. The gallery forest of Abercorn covered too large an area to be an ideal habitat for G. morsitans, but because buffalo lived there, tsetses could feed at its edges.^[87] Given that the landscape was “naturally” appealing to G. morsitans, why was sleeping sickness only becoming a serious problem in the early 1930s?

In the “natural” scheme of things, Swynnerton argued, sufficient African population and their burned gardens “would constitute an excellent barrier to tsetse and a measure for tsetse control,” however dangerous they were to forests: “[T]setse do not like country that is generally open and do not for some years find very favorable the regrowth of felled miombo.” ^[88] When native reserves were formed in 1930, they created unprecedented population densities, as high as sixteen or twenty per square mile. Although citemene-cultivating Mambwe responded by reducing their citemene and expanding their hoed gardens, the degeneration of the surrounding woodlands was rapid and spectacular.^[89] Administrators extolled the virtues of hoed garden crops, but had “no reason to believe,” the provincial commissioner wrote in 1933, “that hoeing will become general until the last tree in the Reserve is cut and burnt.” ^[90] But according to Swynnerton, the removal of citemene cultivators into “reserves far from the tsetse” was harmful to the ecology of the region. It “appears to be responsible for such danger as attaches to the present position.” The land to the east of Abercorn had been protected by Africans’ citemene, but with the Africans gone, “fly advance will be facilitated.”

Swynnerton dismissed any move to eradicate wild animals, because new animals would simply move in to to replace them; miombo was a good animal habitat as well. Instead, he suggested that tsetse be controlled by African land use, especially citemene. In a powerful argument for citemene as a method of fly control, contrary to all official thinking in the mid 1930s, he claimed that leaving areas fallow and unburned for years at a time would actually cause a retreat in G. morsitans populations.^[91] Tsetses need bare ground to settle on; G. morsitans in particular need visibility and dry ground to breed. When grass is not burned for years at a time, the conditions for fly breeding are destroyed: a grass mat forms and the undergrowth becomes dense and inhospitable. There is no place for G. morsitans to settle or breed. “ I have seen no area which is so suited to this measure as in Abercorn, ” Swynnerton wrote (emphasis in original). Moreover, burning the grass “very early” in the dry season—a method Swynnerton had already used in Southern Rhodesia—“made grass so dense that it proved unfavorable to morsitans.” Since Africans in Abercorn did not burn their woodpiles until late in the dry season, he thought that they would support bans on burning at other times during the dry season.^[92]

Late burning—burning woodpiles late in the dry season—had been problematized at least as much as citemene itself, as we have seen. Even when officials tolerated citemene, they emphatically opposed late burning. Slash-and-burn systems depend on fire: when it is time to burn the stacked branches, the flames run riot, sweeping through the leaf litter and vegetation nearby. This vegetation would be particularly dense if the ground had been burned early in the dry season. The later in the dry season burning takes places, the dryer the surrounding vegetation, and the hotter and more intense the fire. According to Richards, “The piled-up branches, dry and brittle from three or four months exposure to the sun, crackled and flamed in a moment.…Next morning nothing was left but a circular bed of ashes around the blackened trunks of mutilated trees.” ^[93] But late burning, year after year, was held to retard the regrowth of trees, and destroy the forest canopy. The “scientific proof” of burning experiments in the 1930s was offered to show that citemene had only survived because there had not been repeated late burnings. Early burnings—early in the dry season—which were less intense and cooler, allowed trees to regenerate more rapidly.^[94] But the intensity of late burning at the end of the dry season produced overgrown ground, its shade cover retarded by fire, early in the rainy season that was inhospitable to G. morsitans; the long-term effects of late burning destroyed areas of shade that were essential to G. morsitans’s survival. While Swynnerton argued that the overall patterns of citemene provided natural barriers to tsetse flies,^[95] he underestimated the extent to which late burning also prevented tsetse breeding.

Swynnerton’s recommendations were sent to the deputy director of agriculture, William Allen. His response was swift and severe: these plans were not really worthwhile and were too expensive “to protect a few Mtungu Road coffee farms.” ^[96] But Abercorn’s white farmers persisted, offering their own version of reterritorialization; they proposed early burning over large areas and argued with officials who said that the value of the district’s cattle did not warrant drastic ecological or hunting measures. Within a few weeks, the provincial commissioner requested that Lusaka authorize Swynnerton’s other recommendations: the resettlement of Africans in areas they had formerly occupied and the use of fly pickets to remove flies from travelers.^[97] Pickets immediately became a charged issue. Various planters’ association wanted pickets installed at once, as did the medical officer at Abercorn, who wanted pickets on selected paths, to which Africans would be confined.^[98]

A month later, the survey botanist of the Tanganyika Territory Department of Tsetse Research, B. D. Burtt, explained how pickets worked and what they were expected to do. Fly pickets were to consist of two Africans in khaki uniforms—“proved to be the most suitable garb” for catching flies—and provided with a hut; their hours should be 7:00 a.m.–5:30 p.m.; “each picket would examine each passing native coming up from Isoka or Mewilo areas” or passing up and down the Abercorn-Mpulungu road; flies caught should be placed in a tube and shown to inspecting officers, who should visit frequently to make sure that the work was being done efficiently; and a record of “the fly take” at each picket should be kept and inspected.^[99] Pickets were assigned to the posts nearest their villages in order to simplify rationing procedures, and Africans traveling on picketed paths were to submit to examinations by the pickets and could not “proceed until this has been done to their satisfaction.” ^[100] Such a system may have allowed pickets great latitude in determining who to stop or what “their satisfaction” might require. Nevertheless, fly pickets did not stay on the job very long, and gave notice often.^[101]

Pickets began operation in June 1936 and were soon in use throughout the district.^[102] They were also soon recoded:

Curious rumours have become extant concerning the existence of human vampires. These were alleged to prey upon solitary persons come upon in the bush whose blood is conveyed to a white master for the manufacture of pernicious medicines. In 1936 the anti-tsetse operations and the natives armed with nets to take fly off travellers at fly posts furnished fuel for these stories and at one time hindered the efficacy of control measures. Propaganda caused these rumors to die out but they have unexpectedly broken out again in other parts of the district and it is difficult to say what, or to whom, their origin is to be ascribed. Variants of the rumor suggest that Government is itself interested in the taking of blood.^[103]

The fly pickets were said to “stupefy their victims, murder them and extract their blood which is an essential ingredient in medicine concocted by the European.” ^[104] It was said that the butterfly nets of the fly pickets could expand to capture a grown man.^[105]

Officials, Animals, and Flies

What was government doing at this time, to earn such a reputation? Also in June 1936, provincial commissioners met in Lusaka to discuss the implementation of Pitman’s recommendations for game parks, a game department, and elephant control by licensed hunters. The PCs liked the idea of a game department, and all suggested ways in which it could be established inexpensively: they could second someone already in service to head it, for example, and employ the African game scouts “already being effectively trained by elephant control officers.” Settlers were so keen on the idea of a game department and game preservation that they “had gone so far as to guarantee part of the cost of the establishment of such a Department.” PCs neither wanted any Africans moved to create a new reserve, as Pitman recommended, nor wanted to distinguish between amateur and professional hunters’ licenses the way other colonies did.^[106]

In 1938, a white hunter was accused of being banyama, although the evidence about the accusation is vague. In July 1938, G. Kennedy Jenkins, a cadet, toured part of Mpika District with two goals in mind. First, he planned to preach the control of grass fires and the merits of early burning—a scheme, he wrote, that “must have struck some as revolutionary” but that would protect young trees—and second, he hoped to investigate a “disagreeable encounter” between a white hunter, identified only as Captain Henderson, and local villagers, who thought he was a munyama. Jenkins wanted to find out what had happened and “to dispel the fanciful rumours which made such a incident possible.…I made it clear at the outset that my intentions were simply to find out why these stories received credence and to allay the fears they aroused.” Yet almost no one in the village would talk to him; villagers seemed to believe that “any confession of complicity in the Captain Henderson incident or even belief in the banyama stories would at once be rewarded with punishment.” The only person who spoke of the incident was one of Henderson’s porters. Traveling with Henderson had brought him within a few miles of his home, but when he went to find his mother working in her garden, “To his surprise she greeted him with an outburst of maledictions and harangued him for allowing himself to be engaged in so dishonorable a project.” She ran to her village, where her “grief…was calculated to unsettle the minds of her fellow villagers.” The villagers did not threaten Henderson, but they did believe he was a munyama and feared “he might conduct his business on their doorsteps.”

Despite his belief in Africans’ “ingrained appetite for such fancies,” Jenkins quickly realized the limits of his investigations and his inability to recode those fantasies into imperial words. “It was impossible to arrive at any very clear conclusions or to discover in what manner the minds of these people had arrived at any such alarming ideas.…I doubt whether people with such a curious turn of mind are at all likely to be won over by rational argument,” especially as villagers denied any knowledge of banyama beliefs. His solution to such silences was to warn locals that “if harm came to any strangers punishment would not be mediated because the victim was an alleged munyama,” in his hope that this might guarantee an end to such incidents.^[107]

What did Henderson do? It is impossible to tell from these records; all we can really discern is that he was a sport hunter with an African staff—although in Northern Rhodesia amateur hunting had the same status, and the same appearance, as elephant control in the late 1930s. But just as banyama accusations did not emerge from the specific deeds of “fly boys” they did not emerge from official or sportsmen’s acts. The extractive power of Europeans—whether taxation, migrancy, poorly paid fly pickets, or the creation of new reserves—was in the air: banyama accusations were leveled at figures in a specific locale because of local issues and concerns.

Banyama rumors related to hunting and tsetse control seem to have died out after 1939. There were few banyama scares during World War II, most of which were anticipated by officials.^[108] Wartime economies reduced administrative staff and curtailed activities not directly related to the war effort; W. V. Brelsford, DO, Chinsali, noticed a decline in banyama accusations when the Game and Tsetse Fly Control Departments were not working. According to Brelsford, Unwin Moffatt, the agricultural officer at Abercorn, told how in 1939,

accompanied by one or two Tsetse Fly Catchers, he went to speak to a group of African men resting by a stream. The whole group rushed off into the bush abandoning all their loads and possessions. Unwin Moffatt, a descendant of David Livingstone, whose father was a famous missionary and whose two brothers were administrative officers, all well known to the Africans of the area, had been stationed in Abercorn for many years and he was universally liked by Africans. But even the appearance of a familiar and trusted man did nothing to banish the panic raised by the sight of those little white Banyama nets.^[109]

I have combined Jenkins’s self-conscious account and Moffatt’s unreflexive one to make a text. As such, it raises another question: were officials feared because they were considered banyama or were they considered banyama because of their individual association with anti-citemene policies? But that question is read off my text; it is not necessarily one that Africans in Northern Rhodesia in the late 1930s would have worried about: the very question imposes a distinction between tsetse control, early burning, and banyama that individuals in the region may not have made. Indeed, such ambiguity may be a hidden strength of documentary evidence, however. I think it is unlikely that oral interviews conducted in Mpika today, for example, could uncover whether or not Africans believed in the late 1930s that white hunting was emblematic of official disdain for citemene as a form of tsetse control. Even if such a viewpoint could be extracted from personal memories and narratives, such evidence might obscure the ways in which banyama rumors blended and disputed the issues of African and European ideas about disease and agriculture. The very layers of meaning that make some oral accounts so rich might, in this particular instance, decontextualize the ideas and concerns these rumors contained. Just as villagers refused to speak about banyama to Jenkins and many other officials, individuals do not speak to me; I can identify neither speakers, believers, nor nonbelievers. However, without individual African voices, I may nonetheless be able to listen to general anxieties and concerns. Using only archival sources, I lose a great deal of meaning, of African specificity, of who believed what and possibly even why, but I gain the complexities and confusions of and about agriculture and science, late burning, and blood in a colonial situation. The archive reveals not only its own confusions and contradictions but the inability of colonialists to locate their practices completely in imperial rather than African terms.

Notes