2. Historicizing Rumor and Gossip

This chapter is not concerned with how Africans might have believed that Europeans hired Africans to capture their fellows and take their blood; rather, it is about how historians might use rumor and gossip as primary sources in the writing of history. But rumor and gossip have very little in common. Lumping them together is as recent as journalism and communications studies; they were put together to create a category of unreliable oral information—rumor and gossip were not thought to be substantiated like newspaper accounts. But anthropologists, in an earlier, more functionalist era, had a less romantic view of the printed word: gossip and scandal were linked together as phenomena of speech and control, while rumor was news that one later learned was false. Social psychologists and sociologists, however, who had long claimed rumor as their own, argued that falsehood was not an absolute characteristic of rumor. What characterized rumors was the intensity with which they were spread. Indeed, the more widespread and widely told a rumor was, the more it had to conform to the laws of plausibility.^[1] The folklorists who struggled to disaggregate rumor from legend ended up struggling over the relative importance of the truth of stories compared to the importance of how or why they were told. Those who regarded legends as frozen rumors had not paid close enough attention to the full narrative style of legends, and those who saw rumor and legend as unrelated tended to focus on the truth of the stories, not how they were told. But folklorists understood that what made a rumor or a legend powerful was that people believed it.^[2]

But how the African peoples discussed in this book decided what was true and what was false, what was rumor and what was eyewitness account, and how much credibility to give to each is not a simple matter of how information was presented. In the case of Colonial Northern Rhodesia (Zambia), for example, the Bemba word for rumor, talk, and conversation is the same, ilyashi. It refers to how people exchanged information, not the credibility of that information.^[3] Indeed, how rumor is distinguished from fact by Bemba speakers is not at all clear. From the 1920s on, they heard tales of a twig that could strip a man of his will-power and of Congolese cannibals who kidnapped Copperbelt workers; 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.^[4] This does not mean that everyone believed each and every one of these tales, or that they believed or doubted them for very long, but it does suggest that why one such story was credible while another sounded ludicrous had to do with local people’s appreciation and apprehension of certain facts, not with whether a story was grandiose, frightful, and transmitted orally. But if Africans did not believe stories because they were written, nor did they doubt the rumors that were discredited in print. Published denunciations of rumor were often thought to prove its truth. The written word was as subject to debate and derision as any oral statement was. Newspapers did not underscore the truth of the printed word, but provided many contending versions of it. In the late 1940s, a Ugandan nationalist—and newspaper owner—proposed a radio station, with receivers in every chiefdom, to make sure that the correct version of events got around. Indeed, many African oral forms that were explicitly unreliable were known by names that played off official media—Waya Times in Kenya’s detention camps or Radio Katwe in Uganda.

Thus my concern is how to combine the insights of academics with those of Africans in ways that might historicize rumor and gossip. Gossip and scandal served to discipline people, both those who gossiped and those who were gossiped about; both asserted values and defined community standards. Scandal might best be historicized as accusation, a phrase that sums up the agency and the speech act of turning ordinary gossip into something on which action has to be taken. Rumor may simply be poised between an explanation and an assertion: it is not events misinterpreted and deformed, but rather events analyzed and commented upon. As a result, in this book, I use the term “rumor” with as much care and caution as my subject matter allows: it is a very poor term with which to discuss stories that the storytellers think of as true. Indeed, the important question may not be which phrases are gossip, which constitute rumor, and which are accusation. Figuring out how these labels can best be applied may not be particularly enlightening. It may be more useful, particularly for historians, to try to find out what these phrases meant to those who heard and repeated them, over time and over space. Not everyone hears or appreciates or understands gossip or rumors the same way—some gossip and some rumor may be unreliable to some people while sounding perfectly reasonable to others. Labels that foreclose this latitude of credibility may not be worthwhile.

For historians, rumor and gossip and accusation are forms of evidence we need to use with great care and caution. However much street talk in Paris in 1750 may have resembled street talk in Kampala in 1950, each must be examined in terms of the specificity of time and place. Understanding gossip requires understanding social rules, values, and conflicts; such understandings put scholars on the same ground as the gossipers.^[5] But the thorny question of how rumor and gossip make historiography depends on the history and the historian. Georges Lefebvre’s The Great Fear of 1789 reworked secondary sources.^[6] Arlette Farge and Jacques Revel’s The Vanishing Children of Paris is an account of the riots of 1750 when it was said that the king abducted children either to drain their blood for his baths or to send them as colonists to New France. The authors reconstruct rumor and riot from diarists’ accounts and memoirs; some sources repeat the talk of the day, others report parents’ worries years after the fact.^[7] In chapter 8, in discussing the trial of a “well-known stupefier” in Kampala in 1953, I have used gossip and scandal that was reported in newspapers as rumor as evidence. The high politics in which the trial took place was not reported in the press, but, I argue, it can be glimpsed through the newspaper accounts of accusations against various officials. In short, the terms “rumor,” “gossip,” and “accusation” are deployments, not separate and distinct categories.^[8]

Gossip and Respectability

Gossip, as Max Gluckman pointed out years ago, creates ties of intimacy between those gossiping. The subject of the gossip, personal and political, is secondary to the process of creating bonds and boundaries. Scandal serves to keep an individual in line when gossip no longer does the trick. Both gossip and scandal assert social values, not as static traditions but as learned and lived practices: “outsiders cannot join in gossip,” and “a most important part of gaining membership in any group is to learn its scandals.” ^[9]

Such insights make gossip more aural than oral; the fact that it is heard is more important than the fact that it is spoken. For Gluckman, a story is credible because it sounds likely—or interesting, or worth passing on—to those who hear it, not necessarily the skill of its telling or the reliability of who tells it. A contemporary critique of Gluckman emphasized the role of “the gossiper”—the one who could manipulate information for his or her own reasons—which placed the importance of gossip in speaking, rather than in listening and evaluating.^[10] Such a critique, however, extracts gossip from its social context: gossip “is not only a means for an individual to assemble basic information on his peers, but it is also a technique for summarizing public opinion.” ^[11]

How is gossip different from ordinary talking, from storytelling, or from just hanging out? Put simply, gossip is a matter of context and convention. It is talk about people when they are not present, but it is not just any kind of talk: it reports behavior; it rests on evaluating reputations. “One does not gossip about a prostitute who turns ‘tricks,’ but one does gossip about the respectable matron who is observed with men sneaking into her house day and night.” ^[12] Gossip reveals contradictions. Stating that colonial states captured Africans and took their blood revealed their true nature; quoting colonialists on this issue revealed their cynicism and control. “[S]ome of the District Commissioners were announcing that ‘if your goat or cow is lost—you find it, but if your relative is lost—you do not bother to find him because you will not be able to find him.’” ^[13]

In the Western world, the association of gossip with idle, malicious talk is relatively recent, perhaps dating from the early eighteenth century.^[14] Other cultures do not see gossip as a single form of speech, however. Historicizing gossip may require using academic categories in ways that peoples in the past might not have done. The men and women quoted in chapter 1 defined gossip as “loose talk” because its subject matter was grounded in speculation, not experience; gossip was not gossip because of who told it in which context. Gossip was gossip because it was a theory, a conjecture. It might be more rewarding to abandon analytical categories of “idle talk” and ask how we think about talk itself. I suggest an Africanist reading of Michel Foucault’s The History of Sexuality, not necessarily because of what it says about sexuality, but because that is the text where Foucault develops his ideas about speaking and the voice most clearly. These are concepts Africanists hold dear.

According to Foucault, however, speaking in modern societies is far more than how individuals enter the historical record, it is how people participate in the states and civil societies that manage them. The very act of talking about oneself, or others, disciplines; the very practices of sorting out the epistemologies that shock and scandalize creates and catalogues ideas about deviance and virtue, which are enforced with each telling. Modern subjects are not only studied, counted, and classified; they speak about these things for themselves. It is how they are managed. The “task of telling everything” allocated to subjects not only “enlarged the boundaries” of the subject matter on which they might speak but “installed an apparatus” capable of producing more and more speech that eventually policed itself. Thus, the “crude,” the “crass,” and the “vulgar” are not distinctive forms of speech, but speech outside of that management.^[15]

Does gossip police itself? Or is it too crude? Roger D. Abrahams’s 1970 article “A Performance-Centered Approach to Gossip” elaborated on Gluckman and his critics, suggesting that gossip is negotiated between gossiper and audience. On the Caribbean island of St. Vincent, gossip, like story and song, is judged according “to whether it is judiciously performed in the right setting and under the properly licensed conditions.” Gossip publicly condemned behavior that departed from community norms and could be used to build up the esteem of the gossiper; unsuccessful gossips were those who used the device badly, resulting in community disapprobation. Thus, unacceptable gossip is not the gossip that speaks ill of beloved persons; it is the gossip that is performed without skill or protocol.^[16] In Swahili, one way to condemn gossip is to call it takataka, rubbish.

But what about the accusations that fall on deaf ears, the complaints that backfire and undermine the position of the accuser? Is such gossip invariably against well-regarded individuals? Or does it mean that some individuals are beyond reproach or that there is an ahistorical category of reputation that cannot be breached by words? Neither, I think. Unsuccessful gossip proves that gossip and accusation are negotiated, that “unimpeachable reputation” is a specific historic construct that only the most skilled and the most appropriate gossip can impeach. “Bad gossip” invades a person’s privacy more than an audience will tolerate;^[17] in Foucauldian terms, “bad gossip” is crude. Failed gossip backfires and causes more problems for the gossiper than for the person the gossip is about. For example, when an African man was arrested for kidnapping a small child in Northern Rhodesia in 1944—an era of many custody disputes as men returned from the mines of the Copperbelt demanding bridewealth marriages—he stated that he was collecting blood for a white man who was banyama; after several days in jail, he changed his story, saying that he had wanted to rape the child.^[18]

If I add the concepts of bad gossip and failed gossip to that of successful gossip, it is obvious why rumor and gossip are such wonderful historical sources. They occupy the interstices of respectability, exactly following the contours of local and regional concerns. Rumor and gossip allocate responsibility; they contextualize extraction. In the Northern Province of Northern Rhodesia, the provincial commissioner was said to have given the Bemba paramount chief a “large bag of money” to allow banyama into his country.^[19] Rumors, more than gossip, move between ideas about the personal and the political, the local and the national. In Northern Rhodesia, it was said that African blood was made into medicine for Europeans; the long illness of King George V in 1929 was seen by many Africans as ample reason to be wary of strangers. Banyama snatched men and sold them to the Belgian Congo, where they were put to work in large secret camps not far from the border. Sometimes “they are sold to medical institutions for experiments and operations.” ^[20] Rumors explain; they naturalize the unnatural. In the 1920s, it was said that every town in East Africa had a Parsee whose house was equipped with pits and buckets to collect African blood.^[21] In Northern Rhodesia, Africans told a district officer during the Depression that they could not go to Tanganyika Territory to look for work “because the white ants had eaten all the money, a white man told them so.” ^[22]

Successful gossip and accusation tell us how penetrable a reputation may be, and when it is penetrable; gossip in particular discloses the boundaries of attack and subversion. In Uganda, my assistants and I asked men if they knew whether bazimamoto ever used prostitutes to help them capture men, as they were said to have done in Kenya. Many responded that they had never heard this, but that it sounded likely, since “they would do anything for money.” ^[23] Another thought the question important: “I don’t know anything about that but please, try to do research on that.” ^[24] Successful gossip and accusation must be keenly aware of the shifts in reception and credibility of certain issues. Accusation and compliment can coexist, but they are rarely spoken to the same audience, at least not at the same time. In Uganda in 1990, many people thought that “Dr. Duke” was the man in Entebbe “who received the blood.” Some people thought “he graded it and distributed it”; another thought “he was responsible for the blood transfusion at Entebbe.” Many people thought he was a doctor, but one man thought he was the governor. One man did not know which department took the blood but knew the man who did: Duke. But Duke was Lyndal Duke of the Tsetse Research Department, who retired in 1934. A man described the job: “to bribe people with some little money, take them to Entebbe tse-tse areas to be bitten by the flies, something which was intended for research purposes on tse-tse victims.” ^[25] But a few people also remembered Duke as the founder of what became the Entebbe Zoo: “Duke was a medical doctor and a fat man in size, he was working at Entebbe and he was a collector of different kinds of animals.…They were tamed and people could go there and see them…he had a pond for fish and people could go there and see how these fish were playing.” ^[26] “Duke had a farm for animals.…[H]is lions didn’t kill anybody, they were for the public interest, and when you went there you would find animals well fenced in a place made of iron bars, and you would see leopards, lions, crocodiles.” After such praise, gossip had to be presented with great circumspection and innuendo. “He was collecting blood, but there was a big hospital there. I don’t know if he was collecting blood for bazimamoto or not.” ^[27] A cautious accusation is far better—in terms of performance and reception—than a badly timed one, which can land the accuser in trouble. Thus, for historians at least, the power of gossip is more than a collective delight in the vices of friends and colleagues. Gossip is a reliable historical source because it traces the boundaries created by talking about someone. In that talking, a world of value and behavior is constituted: that’s what Foucault’s idea of discourse does; it does no more.^[28]

Between Gluckman and Foucault lies, structurally and historically, communication theory. These empirical studies of rumor attempted to prove, among other things, which qualities of transmission made oral information credible—was it overheard, made authoritative by liquor, or told with greater skill and attention to detail than a true story would necessarily merit? Such surveys and quantification of how belief is articulated may have made the ambiguities of hearing and thinking too concrete and clear-cut, but they also show the skill and the discretion with which oral information is evaluated, censored, amended, and passed on or withheld.^[29] They reveal how local, rather than how personal, the evaluation of gossip is.

Gossip and accusation are idioms of intimacy. How deeply do we care about the vices of people we don’t know? An interpretation of gossip based solely on Gluckman would imply that we might not care about the vices of public personalities. One that combines Gluckman and Foucault would argue that there is no difference between talking about strangers or talking about our neighbors—it is the very process of gossiping that creates the intimacy. In gossiping, a claim is made to knowledge and the right to speak it. How deeply we care about the vices of strangers depends in part on the meanings attached to information—how scarce, how important, and how specialized and hard to come by it is—with which the gossip gossips.^[30] Gossip about people we don’t know not only binds gossipers together in an imagined community of shared values, but binds gossipers to communities, states, and sanctions. Gossip about strangers may have meaning because of the very intimacy translated to daily life by the original usage of the term. But in the case of strangers, the epistemologies of our caring about the vices of strangers “percolates into formal agencies of social control” and out of them again.^[31] For historians at least, this is a crucially important and theoretically rewarding place to start, because if we can historicize gossip, we look at the boundaries and bonds of a community. Who says what about whom, to whom, articulates the alliances and affiliations of the conflicts of daily life. In 1940, in colonial Elisabethville, in the Belgian Congo,

A colonial official called a meeting with all the women…who made or sold beer or the local brew. The colonial official demanded of all these women the addresses of their homes, to tell the street, the number of the house and the name of the occupant. The women did this. Then, the colonial official asked that the women notify them each time a man was drunk at their houses. Then, the colonial official would send someone to arrest the drunken man. In fact, he was working for batumbula. And each time these agents arrested a drunken man they would give the woman 2,50 Fr for each person arrested.^[32]

The women in Nairobi described in chapter 5 aspired to own property that could be inherited by heirs they themselves chose, without reference to blood or filiation. These women told elaborate bloodsucking stories specific to the complications and contradictions of female property-ownership. Nevertheless, some property-owning women had earned the condemnation of others; they had worked for wazimamoto and “bought their houses with the blood of somebody.” Gossip was often about gossip; it criticized gossipers and their motives. A woman in Uganda said she had heard of bazimamoto, but “when anyone constructed a good house, he was suspected of being a bazimamoto, or of being involved in…capturing people and selling them to bazimamoto—because he had a good house!” ^[33] A form of speech that actively debates and establishes and reestablishes the criteria for success and failure, for prestige and scorn, is a tool for writing the history of communities, of neighborhoods and regions, in intimate detail.

Voices and Subjects

There are perhaps certain discomforting parallels between how modern regimes have required speaking subjects and how African historians have required them. The intellectual foundations of the oral history of living persons—the life history, or personal narrative—rest on some basic assumptions: that people are the most accurate chroniclers of their own lives, and that experience is evidence of the most reliable sort. While there have been piecemeal critiques,^[34] this view dominated African history by the 1980s. And why not? The twentieth century could be best explained by those who lived through it, especially when those lives were not always deemed important enough for the historical record: African voices could fill the gaps in official documentation and provide a version of events suppressed by colonial chroniclers.^[35] Voices were considered such a key tool to the reconstruction of African history that they were never problematized.

But recent work in history has queried some of the assumptions on which the oral evidence about twentieth-century Africa rests. The notion of an essential self, a persona that sees his or her life the same way over time, now seems rather quaint.^[36] The idea that experience alone can provide historical evidence, on the other hand, seems far too simplistic in and of itself. What counts as experience and what counts as fantasy? How are the two to be distinguished? Can accounts of the real ever fully purge themselves of the fantastic, especially when the fantastic contains debates about the real? And how would people report things that do not conform to their own norms of experience? The historical reconstruction of experience is no easier. How do historians interpret things we think our informants did not really experience?^[37]

All of this should make the practice of twentieth-century oral history problematic. The question of who is reporting what experiences in which way is crucial to our practice. Research into colonial subjectivities by historians is rare; when historians have written about African lives, it has been to show how those lives represent colonial experiences, or how they can be shown to illuminate and elucidate a history not discernible from the more conventional narratives of national histories.^[38] Stephan Miescher has studied men’s life histories as the history of ideas about self-presentation,^[39] but there has been no formal historical inquiry into what a Bemba or Ganda self, for example, was like in 1930 and how it saw the world. Such inquiries are, or should be, crucial to historians’ work of interpretation. When people talk about what they did in the past, are they talking about their present-day personas or about ones from an earlier era? How can we know what someone was like—what they felt and thought—in their twenties without rewriting their lives for them? In some parts of Africa, we have enough information to hint at some changing notion of self, but for most places, we have relied on vast oversimplification of personalities in order to use oral sources. What someone says in 1990 about himself or herself in 1935 is taken to be true because the same person is doing the talking. Historians rarely ask if the experience is described with the insights of 1990 or 1935, however. Similarly, what we know about African selves in the 1980s and 1990s is applied to recorded testimonies from 1913 or 1947.

Such ahistorical treatment of African selves has had historiographic consequences. The absence of historicized subjectivities in colonial Africa has given scholars African voices without selves, voices in which no embodiments, interests, and powers strive to be reinvented and reinterpreted as they speak.^[40] This means, in short, that the voice captured in an interview may be a risky source with which to know and understand the self of forty years before. Barbara Myerhoff has argued that much interviewing of the elderly involves their own self-conscious construction of a coherent self, whom they present “as a stable, continuous being through time, across continents and epochs.” A life is reworked by the informant for very specific and personal goals: “The discovery of personal unity between the flow and flux of ordinary life is the personal counterpart of myth-making.” ^[41] The evidence derived from people talking about their own pasts requires an understanding of who they were and how they saw themselves in their past worlds.^[42]

But what about people talking about others? What about gossip? I suggest that gossip is at least as reliable as people talking about themselves. If historians have failed to historicize African selves, let alone to interpret people’s words about their own lives, talk about others may be the only source left to them. Gossip, in practice, contains interests, embodiments, and local strands of power. It reveals precisely those passions, complaints, and revisions that are sometimes suppressed in the lives written about from oral interviews. In sharp contrast to the idea of Africans speaking for themselves in life histories,^[43] gossip reveals motivations and interests of the gossiper at a specific moment.

Kas Maine, a black South African sharecropper, for example, presented himself as hardworking and self-righteous when he recalled turning down the invitation of his landlord for a trip to town to watch boxing: “I refused to go and told him I could not stomach that shit—sitting in a tent all day watching others do their work while we left our own unattended.” ^[44] The recollection about another reveals a self. Indeed, in two separate interviews over a five-year period, with different interviewers, the normally laconic Maine described almost word for word his early years as an independent farmer in debt to a local shopkeeper. In each version, Maine’s attempts to settle the debt earned the admiration, praise, and favor of the trader, an Australian ex-soldier named William Hambly who had stayed on in South Africa after the Anglo-Boer War of 1899–1902.^[45] Do these two recollections, with their exact level of detail, reveal a terrain of memory or the strength of a performance? Such a distinction may not matter; the recollection may instead reveal the analytical possibilities of gossip, that in talking about Hambly, Maine presents himself best. Not only does Hambly become a vehicle with which Maine constructs memories of himself and his impact on others, but in recalling Hambly, Maine recalls Hambly talking about him. The line between speaking about oneself and speaking about others is hardly firm: that is precisely my point. A self is revealed in talking about others at least as much as it is revealed in introspection.

This raises another question altogether: is all talk about an absent party gossip? Most functionalists would probably say no, gossip is gossip when both parties know the absent one. Most Foucauldians would probably argue yes, but that it doesn’t matter, that Maine is the real subject of this recollection. But gossip is not such an autonomous, independent part of speech that it does not slip into something else. As Ulf Hannerz notes, “the same information may be gossip or non-gossip depending on who gives it to whom.” ^[46] A well-crafted memory may be recounted in a variety of contexts. The story of Hambly’s praise of Maine may have had more pointed meanings when told to those who remembered Hambly than when told to those who did not. The appearance of such a story in a series of interviews suggests that a way of talking about others—whether such talking was once or still is gossip in other venues—has become part of a repertoire of anecdotes, stories, and memories that the speaker uses to make points about his or her life.^[47]

Historicizing rumor and gossip means not only making them historical sources but utilizing the ways in which they are both historical and intensely personal. Reading gossip, rumor, or accusation for the clues suggested by Carlo Ginzburg allows historians to focus on the details with which these stories are told, rather than on the truth of stories. In his essay “Clues: Roots of an Evidential Paradigm,” Ginzburg argues that historical method was made scientific by a very specific reading of evidence that began in the second half of the nineteenth century. Details were to be interrogated by the trained professional; they became clues by which a broader framework was examined and defined. The details of the human ear, for example, reveal the authenticity of paintings and portraits or show familial relations; a cloven hoofprint reveals a herbivore. Ginzburg’s own historical practice went beyond body parts, however, and he read in the witches, werewolves, and sabbats of inquisition testimony a world of folk practices that had existed for centuries; they were not inquisitors’ fantasies.^[48] But the history derived solely from reading clues may be more linear than the personal inventions of gossip and the periodicity of rumor would require. How people talk about themselves and their experiences, with what words and imaginings, does not flow directly out of a folk past; the power of those images derives from their historical and cultural meanings, of course, but also from individuals’ ability to use them to describe their lives, their conflicts, and their fears. The power of any particular piece of gossip lies in the importance of the contradictions it reveals; the power of a rumor lies in the contradictions it brings together and explains. What the pits in houses or the rubber sucking tubes or cars without lights in vampire stories bring to personal narratives is not their fixed place in African understandings of colonial medicine or colonial technology but rather a broad genealogy of relationships of power, skill, and specialization. The tracks read to reveal the animal’s habits and history, so central to Ginzburg’s analysis, might get a looser reading somewhere else: the cars without lights and rubber sucking tubes in vampire stories reveal motives with as much clarity as the cloven hoofprint reveals a jawbone in natural history.

The clues and details of African vampire stories are not a special kind of thinking but a special kind of talking. They are the images and ideas and clichés that speakers reinterpret, interrogate, and problematize as they talk about themselves and others; this is the thick description provided by the use of hearsay. The power of African vampire stories lies in part in locating the similarities between a wink and a twitch in the motivations behind them; in vampire stories, there is the “thin description” of noting that twitches or winks, or rubber sucking tubes or cars without lights, are present; the thick description lies in the detailed analysis of the motivation and intent behind their presence in the African night. The cars and the tubes are not clues to a folk past or even to a recent past, but a set of meaningful images that are produced, perceived, interpreted, and parodied by the speakers themselves.^[49] Indeed, Lyndal Roper has argued that such clues should be read as personal statements, and that their deployment, particularly during interrogation, reveals individual conflicts and contestations, often over what the details and clues with which they speak really mean and who controls the meaning during interrogation.^[50]

If informants speak about the past and about themselves with the past, how can scholars get at African subjectivities? In most of this book, I use vampire stories as social rather than personal constructions. How might I read these stories to reveal individual histories? What if I were to read each and every vampire story as a personal statement, as evidence about a self hardly revealed in other ways? What if I were to look at the specific embellishments and embroideries in each story? What if I took the way it was performed as seriously as the content? These are the things African historians are trained to weed out.^[51] But looking at the ways people fashion well-known stories into their own experiences or performances may be a way to historicize an individual’s own ideas about his or her self.

Zaina Kachui told me and my assistant a wazimamoto story about prostitutes who trapped men in pits in their rooms. It was a story common in Nairobi the 1920s and 1930s, but her version had a level of detail and commentary other stories did not have:

A long time ago the wazimamoto was staying in Mashimoni, even those people who were staying in Mashimoni, they bought plots with the blood of somebody. I heard that in those days they used to dig the floors very deep in the house and they covered the floor with a carpet. Where it was deepest, in the center of the floor they’d put a chair and the victim would fall and be killed. Most of the women living there were prostitutes and this is how they made extra money…when a man came for sex, the woman would say karibu [welcome], and the man would go to the chair, and then he would fall into the hole in the floor, and at night the wazimamoto would come and take that man away.…It was easy for these women to find blood for the wazimamoto because there were so many men coming to Mashimoni for sex.^[52]

In chapter 5, this story informs an analysis of property ownership and inheritance in Nairobi; but the question I want to ask now is why did Kachui tell this particular story? No other woman told it. All the other versions I heard were matter-of-fact, without this level of detail and cunning, told without so much energy, enthusiasm, and wordplay. I suggest that by contextualizing this version in terms of the life of Zaina Kachui, I can write about her life without adding emotions she herself never expressed to me in many hours of interviews and conversation. There is no possible interpretation of this material as a cautionary tale about Pumwani neighborhoods. Mashimoni—Swahili for “many in the pits”—had lost its allure as a place for prostitutes several years before Kachui came to Nairobi; besides, she made no attempt to tell this story as a contemporary witness: she was crystal clear that this was a story she had heard. But the very fact that this is not “experience” makes it even more significant that she told this particular story more than forty years after she first heard it.

Zaina Kachui was probably born in Taveta around 1910; her father had immigrated from Kitui several years before. Her parents died when she was relatively young, leaving her and an older brother orphans in the early 1920s. Her brother encouraged her “to go with men” to support them both, but he died a year or two later, and she went to Kitui. She was not yet an adolescent: “I didn’t even have breasts yet.” In Kitui, she stayed with a relative of her father’s, who tried to marry her to a man so many years her senior that the district commissioner stopped it. “This is a daughter,” she recalled him saying, “not a wife.” The DC told her to return to her father’s home and, seemingly grateful for the direction, she did. But “that place was not good for me, I had to cook all the food and I hardly got any food to eat,” and so she went to Thika for several years and took up prostitution there. Sometime during her years in Kitui or Thika, she had a stillborn child. She eventually came to Pumwani between 1933 and 1935. She credited an older woman with insisting that she take money from men, rather than finding a boyfriend whose own impoverished state would drain her resources, advice Kachui never fully took to heart. During World War II, a man friend offered to keep her very considerable earnings in a safe place for her, with predictable results. Nevertheless, although she never acquired property, by the time I knew her, she was living in two rooms in the house she had been living in for thirty years, supporting a younger man, who was rumored to be her lover, and selling cooked food to supplement the savings she had managed to live off for almost twenty years. Kachui died in 1981 or 1982. She would have been offended at any characterization of her life as one of failure and misery: she was very proud of her accomplishments, of the fact that in her late sixties, she did not have to ask anyone for help. The self-confident Kachui I knew in 1976 and 1977 did not reflect on the intense vulnerability, pain, and confusion of the preteen prostitute in Taveta or the thirteen-year-old almost bride in Kitui. In fact, she told me about her arranged almost-marriage in the course of explaining colonialism to me and not as a description of the exploitation of adolescents in already overburdened extended families: “In those days the government went by age, if you were young, you got a young DC to make a decision about you.”

But if she did not describe her youth as one of pain and exploitation, how can I?^[53] Can I accurately represent Kachui with my own interpretation of what happened to her, or do I pay attention to her words, her use of language, her sense of metaphor, wordplay, and power? I suggest that her powerful fantasy of passive men, seeking only sex, falling into pits can be read to reveal her vulnerability. This is an interpretation, of course, but one that shows how gossip can be both a practical and intimate source with which to reconstruct the past. A woman passed from man to man during her adolescence might well delight in stories of men passing through property. A woman whose needs were ignored by every kinsman from whom she sought protection might well delight in a story of women’s agency and men’s powerlessness, a story that implied great male stupidity as well: “After a few years, men stopped coming to Mashimoni, because so many men had disappeared there.” A woman who had few choices about home, about family, about men, about everything, in fact, but the remuneration men gave her, might well embellish a well-known story with details about housing, and with enormous power and control. These details of pits and prostitutes are not necessarily unproblematic clues with which historians can see the past; rather, they are the technologies of speaking with which a woman described her world in intimate terms.

Gossip and Silence

But what about not speaking? According to Foucault, silence is an additional strategy. It is “the things one declines to say, or is forbidden to name, the discretion that is required between different speakers.” Silence “functions alongside things said.” Together speech and silence form discourse; speech or silence alone do not.^[54] Starting in the mid 1970s, women’s historians began to equate women’s silence with powerlessness.^[55] Within a few years, a new generation of scholars of colonialism, heavily influenced but not necessarily instructed by Subaltern Studies, began to look at Foucauldian silences as a point of opposition rather than a discursive strategy. The silence of omission—of colonial documents all about men, about elites, about colonized women described with all the malapropisms of white men—was read to reveal the gender and power of colonial agendas. Reading the silences of documents was a way to see who mattered, and how they mattered, under colonial rule.^[56] This insight quickly got out of hand, however: scholars of colonialism in general, and oral historians in particular, began to “listen to silences.” Anyone whose voice was not included had been silenced, and any number of interviews were interpreted for what was unsaid, rather than what was said. This gave interviewers much more power than they would admit wielding. Silence in an interview, a commission of inquiry, or a courtroom, was no longer strategic, it became another site of interpretation. Not speaking was not seen as resistance but as oppression. Listening to silences collapsed the differences between speech and silence; it turned silence into a sort of interpreted speech. Interpreting silences homogenized the different cultural meanings of specific silences.^[57]

I want to return here to an old-fashioned, but Foucauldian, interpretation of silences. They are neither spoken nor heard: that is their power. They evade explicit meanings. A clearly intentioned silence could elide complicity. For example, officials in colonial Northern Rhodesia regularly complained that Africans would never answer their questions about banyama. But when two men suspected of being banyama went on trial for murder in 1944, officials requested that they not be defended by the district commissioner, because that would fuel suspicions that the government was indeed behind banyama.^[58] Gossip is social; no silence stands alone. In two interviews my assistant and I conducted in western Kenya, one man commented on the other’s silences, not simply to comment on the other man’s interview material but to assess the nature of work, migrancy, and occupational loyalty. I had gone to Goma Village in Yimbo location in western Kenya in 1986 because I knew the grandson of Timotheo Omondo, a Luo man who had taught at Maseno from 1924 to 1947, when he helped found the Luo Thrift and Trading Corporation.^[59] Between his teaching—he taught English to perhaps three-quarters of western Kenya’s elites—and his politics, he knew almost all his neighbors and how their experiences of wazimamoto overlapped. Within a few hours of my and my research assistant’s arrival, Omondo told us to go to see Zebede Oyoyo, the man who had escaped the Nairobi fire brigade in 1923 and was still talking about it. We interviewed Oyoyo twice. In the first interview, he expressed something we were to hear again elsewhere in the district—that the policemen actively supported the fire brigade in capturing Africans. When Oyoyo was kidnapped, “policemen were right there but did absolutely nothing. In fact, they pretended not to notice anything.…I concluded they too were part of the kidnapping.” ^[60] After our interview with Oyoyo, Omondo suggested we go speak to Noah Asingo Olunga, who had been a Nairobi policeman when Oyoyo was captured. After we interviewed Olunga, we went home to Omondo’s house and did a formal interview with him.

We had asked Olunga if it was true that policemen and wazimamoto were the same. He was definite: “No. Policemen and wazimamoto were quite different. All I know was that wazimamoto were putting on black clothes and black caps, while policemen were wearing quite different uniforms.” He never saw any bloodsucking, although he knew people talked about it.^[61] When we interviewed Omondo, however, he did not talk about his years in Nairobi, but about Olunga’s interview, instructing us on how to interpret the silences and omission:

Whoever worked for the police force cannot tell you much because they were the ones who were very much involved in these activities. In fact, policemen were the ones doing this work as wazimamoto agents.…Once one was a policeman he remains so even after leaving his job. Policemen are always careful about what they leave out. Retired policemen cannot tell you exactly what they were doing during their working years.…I think Olungu was just fearing to tell you what exactly they did as policemen.^[62]

The silence here is not an additional, repressed version of the spoken, but a kind of socially constituted understanding of memory, loyalty, and accountability; the silence described here is what the gossip is really about. The meaning of gossip is as social as it is personal. It is pieced together by many people exchanging information over a short period of time or by one person over a lifetime—lived experience again. When several people exchange gossip because it is exciting, what is really going on is a debate, as people argue over the details and reliability of the information, about the issues involved. Silence carries hints, allusions, references, and opinions that are not contained in the other information, but it remains silence, powerful because it is not spoken, and cannot be pulled—or decoded—into speech. Silences do not necessarily lend themselves to the same straightforward interpretation that spoken words do; they require slower or looser analyses. Indeed, a silence from one time can be disclosed at another. In the 1940s, Northern Rhodesian Africans said that it was easy to find out about banyama, because if one of their employees quit, he could talk about them.^[63]

My argument here is that silences are not sites of repression but eloquent assumptions about local knowledge. They are not spoken of, not because they are unspeakable, but because they isolate fragments of powerful stories; they do not carry weight unless the gossip, to use Edgar Morin’s troubling phrase, metastasizes into rumors and accusations. This brings about a “transition from the singular to the generic” in which the isolated fragments, whether barely remembered or discussed daily, were shaped into a specific kind of accusation according to specific conventions.^[64] Without those conventions, without metastasis, such fragments remain the stuff of conversation, gossip, and demonstrations of common sense. “Why isn’t this written about?” my assistant asked Omondo. “Here is something that happened a lot and there is no record of it in any book and in any public history. Nobody talks freely about it, why?” He replied: “Those things were mostly happening during the night and besides, I think they were taking blood to the hospitals. Nowadays we hear that such and such a blood group is needed, but in the olden days nobody was willing to donate his blood.” ^[65]

The silences in gossip and the silences between gossip—even the silences in the neighborhood between Omondo and Olunga—allow gossip to be taken up anew, with new villains and new situations. The silences in gossip allow for gossip to have continued meanings to describe work and history and how old men talk; they allow old men to show off their knowledge of science and medicine in ways that public denunciations of neighbors and old friends would never do.

The full import of these silences is not always completely understood by those who speak and hear silences, let alone by a clever researcher. Like Omondo, several of the men quoted in this book claimed that, over the years, they had learned that stories about vampires were true or false. A man in Uganda never believed that people could disappear until the violence of the Amin years, but a man in Kenya said that over time, he came to realize that his friends and neighbors were actually talking about “blood donation rooms” when they told wazimamoto stories.^[66]

Who is Gossip About?

If gossip is transmitted over time, how is it different from historical memory? Is lore about local heroes gossip or myth-making?^[67] This raises another question: is all gossip useful for historians, or are there specific individuals about whom gossip is particularly useful for historians? Much depends on how much gossip there is, of course: the few African policemen named in official documents as vampires are not otherwise identified in written documentation; they simply became examples of an occupational category that had often been accused. The same is true of the men and women who entered the written record as victims of the vampires. But historical sources are produced in specific contexts: some gossip is foregrounded to underscore other historical concerns and constructions. In western Kenya, for example, “Oleao” was a dreaded kachinja years before he became a guard at Sakwa Prison during the Mau Mau rebellion. He had all the credentials a kachinja needed: he owned a mining company—Black Cat Prospecting—into which he conscripted prisoners; he spoke the local language well; and he drove a Landrover, but he entered the written record only as part of another man’s political memoirs. Indeed, local people only knew his last name and called him “Mr. Robinson.” ^[68] But gossip is often deployed in conversation as a way to discover whether someone is worth gossiping with. The gossip about ordinary people—people who left no paper trail—that seems to be too fragmentary to be made into history may be part of a larger framework by which informants interrogated their interlocutors. In Nairobi, for example, whenever I asked who the first women to build houses in Nairobi were, people scoffed at my question; it was too hard to answer, there were too many: no one came up with names. But when I asked about wazimamoto, I heard about Mama Amida, “the first woman to build here in Pumwani,” who “sold her sister’s daughter to wazimamoto.” ^[69] She was not alone: “There was a fat woman called Halima, and she sold her sister.” ^[70] These remarks were made almost in passing, as examples of what I was asking about. I did not realize it at the time, but these fragments were not presented as gossip but as invitations to gossip: they tested my local knowledge, to see if I knew the landscape—or at least the names and norms—well enough to join in.^[71] Gossip offered to, or written to bypass, the ill-informed remains fragmentary to those who do not know enough to participate;^[72] it is a strategy by which people ascertain how much someone knows and how much to reveal.

But what about gossip about extraordinary folk, famous people about whom various stories circulate and were written down? Is gossip a reliable historical source because it corrects the distortions of the written record or because it provides another dimension to official praise, or at least the writings about people clever enough not to commit their worst excesses to paper?^[73] Such questions imply that written words are better and more trustworthy than gossip. Much written material is gossip, however. The paper trail surrounding a Scandinavian settler farming in colonial Northern Rhodesia and accused by an African in 1944 of being banyama is all fragmentary gossip. The unpublished memoirs of a district officer say that the settler was frequently involved in labor disputes on his farm, which he lost; a trader reported a well-known story that the settler had insulted the Bemba paramount chief in a retail transaction. These fragments suggest why the settler may have been known as banyama. Other fragments reveal other things, but they are nonetheless gossip. The history of the agricultural society of Zambia, for example, notes the number of leopards shot near the Scandinavian’s farm but nothing else. Yengwe, Arthur Davison, another Northern Rhodesian, figured in vampire accusations for over twenty years. Davison was a labor recruiter of considerable violence. He lived in Ndola, and his private life is barely alluded to in published accounts of banyama. But when the author of one of the published accounts turned to write a history of the “characters” among Northern Rhodesia’s pioneers, Davison merited his own chapter. His violence is mentioned—he had once killed eighty Africans in an attack on a stockaded village in retaliation for robbery—but most of the material is the white community’s gossip: Davison had never married but had not “gone native” either and “as far as we know, left no half-caste off-spring”; he had a huge house planned in Ndola that was never finished: extensions were planned but never built, and rubble and foundation pillars dotted the landscape.^[74]

But where there is more than fragmentary evidence, most of the Europeans called vampires were not those who were uniformly despised by Africans. After all, Dr. Duke was described both as a man who took blood from Africans and as a man whose animals pleased Africans. Neither the man nor his job were suspect; how he performed his job was.^[75] Time and time again the white men said to be behind vampire activities were those whose activities were all but ambiguous, the men whose deeds were a hair’s breadth between the use and misuse of their authority and power, men whose actions required thick description to explain what they were doing, the twitch and the wink again. But gossip worth passing on is the gossip that reveals contradictions. Oleao, for example, was said to have spoken Luo fluently.^[76] Dom Grégoire had been the subject of batumbula rumors long before he was transferred to a mission on the Luapula. When he was accused of bringing African captives to the butcheries of Elisabethville, he was actively trying to encourage the Belgian mining company, and largest single employer in all of Katanga, Union Minière du Haut Katanga to buy dried fish from the women of his mission.^[77] When H. F. Cartmel-Robinson, a district commissioner in Northern Rhodesia’s Western Province ordered a smallpox vaccination campaign in 1934, he was accused of collecting blood for banyama.^[78] C. F. M. Swynnerton, the tsetse researcher who struggled against all odds to use African methods of tsetse control in two colonies, was known as the head banyama both in the Tanganyika Territory and Northern Rhodesia.

What is Rumor About?

Many scholars of rumor have argued that rumor is the product of ambiguous situations: rumors resolve contradictions; they explain not only misfortune but good fortune. Rumors in Africa, Karin Barber observes, could explain how someone grew rich without working hard.^[79] Tomatsu Shibutani has called rumor “collective problem-solving” in which “men caught” in ambiguous situations attempt to “construe a meaningful interpretation…by pooling their intellectual resources.” ^[80] Historians have been perhaps less than eager to see rumors as explanations and collective efforts; Alain Corbin has argued that even the most contradictory rumors “revealed collective psychoses, dreams and anxieties” of a period, as when French villagers tortured and killed a harmless nobleman in 1870.^[81] But suggestions of collective problem-solving and collective psychoses both make rumors the speech of unified and homogeneous populations who have no fractures in their vision of the world; such interpretations obscure the contradictory fragments of gossip that make up any rumor. Shibutani reports, for example, that during the American occupation of Japan there was a widespread rumor that General Douglas MacArthur had a Japanese grandparent. He argues that this rumor was a way for Japanese to reconcile their postwar experience of the general’s reforms with their wartime belief that Americans in general and MacArthur in particular hated them.^[82] But such a reading of the rumor ignores all the ways in which this particular story hints at Japanese anti-war and anti-military sentiments stifled during the war. Indeed, Farge and Revel have noted the ways in which rumor reveals wider terrains of belief and theory, of alternative visions of cause and effect, “the power of rumor meant that the whole hotchpotch of culture was in circulation in Paris, made up of snatches of knowledge, truths and half-truths, including a whole mixture of allusions which were called upon according to the needs of the moment.” ^[83] Circulating stories are not constructed on a moment-to-moment basis; they are drawn from a store of historical allusions that have been kept alive and given new and renewed meanings by the gossip and arguments of diverse social groups.^[84]

Rumors about colonial bureaucracies, corporations, events, and diseases thus are not really “about” those things at all; rather, they are narratives, explanations, and theories in which colonial bureaucracies, corporations, events, and diseases are subjects. It would be difficult to argue that rumors about clothing, food, and either deliberate or inadvertent additives—the snake in the coat, the rat in the fried chicken, the urine in Mexican beer—are about the corporations named in the stories; it would be as easy to say that these stories are about the bodily fluids contained, contaminated, or injured in these stories.^[85] Arguing that these stories are about the corporations named in the stories, or about the foodstuffs contaminated, suggests that there is only one audience for a rumor, and only one possible hearing of the rumor by that audience. Are the names of companies, countries, and corporations the site of the rumor, the level of detail that makes it a better, more credible rumor, or the subject of the rumor? Do the detail and specificity of brand names make a story any more compelling or important than does the presence of any other detail? When Africans say the headquarters of bazimamoto was in the Yellow Fever Department at Entebbe, or that batumbula was headquartered at the Hotel Biano near Jadotville, are they saying that bazimamoto and batumbula stories are about these institutions or that the specificity of these institutions locates the stories in a specific region and time? The regional variations within a rumor, however fragmentary and elusive, suggest genealogies of local concerns and historical fixations that would not otherwise be apparent. During the Great Fear of 1789 in France, some regions were said to fear a British invasion; others worried that Croatian troops were massing on their borders; and still others feared Poles or Moors. These were not hysterical accusations but concerns and interests grounded in local historical experience. They do not “explain” the rumor, but they explain how it was locally credible.^[86] The last five chapters of this book argue precisely that, and show how rumors can be a source for local history that reveals the passionate contradictions and anxieties of specific places with specific histories.

Asking, let alone deciphering, what a rumor is about makes a rumor about one thing. It makes rich texts of half truths and local knowledges linear and simplified. Several stories from East and Central Africa, each involving sugar, may make this point. In batumbula stories from the colonial Congo, “The captives of batumbula did not eat maize meal. They drank sugar water or they ate sugarcane. The captives who were favored this way became fat and hairy and were taken to the Hotel Biano, where they were killed and eaten. When there were enough captives, one group was transported to Belgium and another to America.” ^[87] A student at Makerere University College in Uganda wrote an essay about peoples’ anxieties about the 1948 census: “[R]umours are being spread by ignorant people that the government wanted to know the density of the population so they could check the increase of population by giving people medicine indirectly—say mixed with sugar—for indeed when brown sugar was introduced into my country people refused to buy it because of the rumours.” ^[88] In 1952, an anti–Central African Federation pamphlet circulating in Northern Rhodesia reported that “on 28th October the ‘House of Laws’ in London had decided to put poisoned sugar on sale for Africans, commencing on February 8…1953.” The poisoned sugar would cause stillbirths in women and would make men impotent. “The sugar would be recognized by the letters LPS on the packets.” ^[89]

In these stories, the level of detail and specificity about dates and locations are as great as anxieties about sugar, colonial power, and political processes. It is their ability to contain diverse elements that makes rumors powerful, or at least worth telling and retelling. Rumors contain “raw facts,” Jean-Noel Kapferer insists. “Rumors do not take off from the truth but rather seek out the truth.” ^[90] They are open to many interpretations and speak to different factions within the most homogeneous audiences. It is in their exchange and evaluation that they take on sophisticated analysis. To pull these sugar stories apart to explain the failure of brown sugar sales in one place or African conceptions of global commodity circuits in another would strip them of the rich ambivalence of the well-fed captives and the codes by which poisoned sugar was to be identified. It is the allusions and loose ends of the story that give it widespread currency and credibility. Rumors do not seek truth by themselves; the people who tell and the people who interpret rumors do. Indeed, the poisoned sugar accusations occurred shortly before Hortense Powdermaker began her fieldwork in Northern Rhodesia. She reported the stories without reference to the dates involved or the writing on the packets, and she interpreted them in terms of witchcraft poisonings. For the anthropologist, this was not a rumor, but what happened when “the rational fear of Federation moved into the realm of the supernatural.” ^[91]

There is no single correct interpretation of any single rumor; there are interpretations and contextualizations instead. If gossip reveals contradictions, rumor contains contradictions like a fishnet. Rumors rarely lose their specificity or get covered up, but once they are captured in oral or written texts, their diverse and contradictory elements become bundled together, so that teasing out a single meaning, or single hierarchy of meanings, is virtually impossible. Indeed, giving a rumor a single meaning turns rumor into something it is not, something much less rich and complex. As chapter 6 argues, the whole complicated story of a rumor is what makes it told again and again; its diverse elements are its vocabulary. These vocabularies—the details and the clues—are not unproblematic and constant; they change, are reinterpreted, and take on new and powerful meanings in large part because of the conduct and history of the rumor itself. In this way, however, pits first became associated with vampire stories in East Africa—there are published references to pits starting in 1923—and by the 1930s, surveyors, road crews, and prospectors were being suspected of being vampires because they dig pits, and by the late 1950s, white miners and geologists were being accused and sometimes attacked. Similarly, wazimamoto was synonymous with “vampire” in many East African cities long before they had fire brigades or any fire-fighting equipment at all; by the time there were actual, physical fire stations in Tanganyika and Uganda, Africans were said to fear Europeans driving red cars, and officials, equally attuned to the multiple meanings of rumor, used fire engines to patrol towns to discourage crime.

Rumor, Gossip, and Historians

Historicizing gossip may allow historians to access a more intimate terrain of personal experience and of thinking than other historical sources can do. The intimate anger and judgmental scorn of gossip map the changing fortunes, values, and standards of communities that other sources identify only broadly. Disembedding gossip, however, should not entail such radical surgery that gossip is interpreted as a separate and distinct form of speech, to be segregated from the other material presented in oral interviews. Instead, historicizing gossip involves thinking about gossip as a way of talking in which people express their interests more intimately, and more personally, than they might if they were talking only about themselves. Historicizing rumor, on the other hand, may reveal little about the individual life or experiences of the speaker, but contextualized with other rumors by other speakers, it may reveal an intellectual world of fears and fantasies, ideas and claims that have not been studied before. The contradictory elements of rumors can be read to reveal the complications of everyday concerns. Chapters 3 and 4 argue that a broad interpretation of specific rumors can reveal the changing meaning of biomedical intervention or the fissures within a differentiated labor force.

Notes