Part Two: How the Kaonde Became a “Tribe”

The tribe is at once the only bulwark we have against anarchy and the only foundation on which to build progress in local government.

One's conception of the world is a response to certain specific problems posed by reality.

The name Chizela was taken from the local senior Kaonde chief and the district was in the heart of the region most closely associated with the Kaonde. But what kind of an entity are, or were, “the Kaonde”? In the 1980s, as I discussed in the introduction, “the Kaonde” certainly existed in the sense that people would identify themselves as Kaonde and point to a particular way of life as being “how we Kaondes live.” It is doubtful, however, whether prior to the colonial period “the Kaonde” existed as any kind of distinct and self-conscious entity either politically or “culturally,” even if by the end of the colonial period such an entity did indeed exist. This kind of “creation of tribalism” (to borrow the title of Leroy Vail's 1989 volume) has been common throughout Africa, and has been well documented in recent years.^[4] There are many parallels to the story of the category “Kaonde.” ^[5] Nonetheless, it is worth looking in some detail at the specifics of how the Kaonde, like other peoples in the region, became a tribe, with all the colonial baggage that implies.^[6]

There was a hegemonic, taken-for-granted assumption within the British colonial world that the basic social unit within which rural Africans lived was the “tribe”; it was as “tribes” that African communities were named and theorized. For colonial officials, for example, the answer to the question, What kind of entity are the Kaonde? was simple: the Kaonde were a tribe with an origin lost in the mists of time. In fact, however, what this category “tribe” came to mean in British colonial Africa was the outcome of a long, drawn-out dialogue between colonizer and colonized. The particular meanings it acquired in the case of particular “tribes” in particular places depended both on the overarching colonial discourse on the African “tribe” and the specifics of particular localities and the nature of their economic and political realities. A crucial party to this dialogue were the anthropologists, who played a major role in the development of the notion of the “tribe” as a purportedly “scientific” category. In telling the story of the emergence of “the Kaonde” as a “tribe” I have focused on three groups: firstly, the colonial anthropologists of Northern Rhodesia's Rhodes-Livingstone Institute; secondly, the colonial officials stationed in North-Western Province; and thirdly, the Africans whom the name “Kaonde” attempted to locate. While these three groups can be seen as engaged in a complex three-way dialogue, it was a dialogue in which the different groups were far from equal in their power to “name” in an authoritative way the social realities of the colonial state. The data I am using to recover this dialogue are taken primarily from the archives or from secondary sources, in both of which the little corner of Chizela is a rather shadowy presence. Inevitably, therefore, there are few actual Chizela voices; this is a more general story of the mapping of the conceptual terrain that those living in places like Kibala and Bukama were told they inhabited, and also how in a particular colonial context the name “Kaonde” acquired a particular set of meanings.

I want to begin the story of how the Kaonde became a “tribe,” and what the implications of this “tribal” identity were, with the more general concept of the “tribe,” and how it shaped, and was shaped by, the untidy and dynamic realities it was supposed to explain. To do this I have chosen to focus on the nature of the “tribe” as a theoretical concept in the work of one of the most prominent anthropologists to work in colonial Zambia (Northern Rhodesia), Max Gluckman. Gluckman was the director of the Rhodes-Livingstone Institute before leaving Northern Rhodesia to take up the newly created chair in social anthropology at Manchester University in Britain.

Max Gluckman and the Rhodes-Livingstone Institute

The Rhodes-Livingstone Institute (RLI) was founded in 1937 with a dual mission.^[7] On the one hand, it was to be an independent institution, free from direct control by the colonial state, that would use the procedures of the relatively newly professionalized discipline of social anthropology to generate “scientific” knowledge about the subjects of British colonial rule in Central Africa. On the other hand, it would also provide the colonial authorities with useful information that could be used to facilitate the smooth and humane operation of colonial rule. Such information, it was argued, “is necessary if we are to base on any sure foundation the improvement we are seeking to make in African conditions” (quoted in Brown 1973:184). Colonial officials and Northern Rhodesia's white settlers may, as Brown shows, have been somewhat skeptical as to the validity of the latter argument, but it was an important strand in the RLI's own understanding of the role of anthropology. Godfrey Wilson, the institute's first director, was an especially strong believer in the important role that anthropology (and the social sciences generally) could, and should, play in solving social problems.^[8] Like Wilson, whom he succeeded in the early 1940s, Gluckman was convinced of the importance of anthropology beyond the academy. The range of the RLI's aims under Gluckman, which he explained in his “Seven-Year Research Plan” (Gluckman 1945), was reflected in its different publications. There was the scholarly Journal of the Rhodes-Livingstone Institute, produced for an academic audience; but there was also “provision for presenting the results of scientific research to laymen” (1945:28) in the journal Human Problems in British Central Africa, while Communications from the Rhodes-Livingstone Institute provided the “detailed data beyond what the sociologist customarily publishes…[which] “Government often requires” (1945:28).”

The colonial authorities themselves remained distinctly ambivalent about the value of anthropological research, and as the RLI developed the emphasis became increasingly scholarly and academic. The relationship between the RLI and the colonial state was always, however, a complicated one. It would certainly be a gross oversimplification to see anthropologists like Wilson, as certain critics of colonial anthropology have done, as “the handmaidens of colonialism.” What I am interested in here, however, is the rather different question of the extent to which a theoretical concept such as that of the “tribe” and its associated problematic embody in themselves, quite independently of the explicit aims of those who use them, specific ways of naming the world that have implicit in them their own political claims.

A key feature of the colonial state in British Central Africa from the late 1920s was its commitment to the principle of Indirect Rule. This combined a cheap form of administration that used locals to police its lower tiers, with a comforting illusion of local autonomy. For Indirect Rule to work, both ideologically and practically, it was essential that everybody involved—or at least everybody whose voice was likely to be heard—believed that rural African society did indeed in some essential sense retain its old precolonial structures of authority and forms of social organization. In fact, however, the imposition of colonial rule represented a moment of fundamental rupture. Even if a genuine de facto colonial presence was slow to establish itself in many areas, analytically, I would argue, colonialism has to be seen as introducing a real discontinuity. One of the great silences underpinning the British colonial state's enthusiasm for Indirect Rule in sub-Saharan Africa was the unspoken denial of the basic fact that the establishment of pax Britannica, and of the colonial power as the ultimate authority over land, law, and so on, necessarily eroded radically the power base on which precolonial political authority rested. It is against this silence that the project of understanding and controlling the colonized society takes shape; and it is within this project that the work of the RLI anthropologists was located.

Central to this work was the concept of the “tribe.” When fledgling anthropologists (such as Victor Turner, Elizabeth Colson, Clyde Mitchell, A. L. Epstein, William Watson, and Norman Long, all of whom did fieldwork in Northern Rhodesia) arrived at the RLI to carry out their first fieldwork, they were sent off on a preliminary trip to a designated “tribe.” Clutching a bundle of RLI index cards, they had the task of plotting out the basic structures of this “tribe.” A characteristic piece that resulted from one of these first forays was an article by William Watson (who went on to write Tribal Cohesion in a Money Economy: A Study of the Mambwe People of Zambia) on “the Kaonde” entitled “The Kaonde Village” (Watson 1954). After this initiation the neophyte anthropologist would then return to the RLI and set out for his or her main fieldwork in some other area. Whatever the ultimate topic of their research, the starting point was always a particular “tribe”; for instance, for Victor Turner the Ndembu, for Watson the Mambwe, for Elizabeth Colson the Tonga.^[9] An unquestioned assumption was that the basic community within which rural Africans lived was the “tribe.” To trace out something of what lay behind this name of “tribe” for the RLI anthropologists, I have chosen to focus on a single text by Gluckman, Politics, Law, and Ritual in Tribal Society (hereafter Politics, Law, and Ritual).

There are several reasons why I chose this particular text. Firstly, there is Gluckman's dominating role within the RLI and in the anthropology of Central Africa. Secondly, while it may not represent Gluckman at his best, because it was a book written as a teaching text for undergraduate anthropology students in Britain, in it Gluckman is very much concerned with laying out what he saw as some of the fundamental concepts in the discipline. Finally, Politics, Law, and Ritual (published in 1965) was written in 1964 at the precise moment when Northern Rhodesia was gaining independence as Zambia, and it can be seen as representing a summation of one variant of colonial anthropology. Indeed, by this point functionalist anthropology of this kind was already being attacked as ahistorical, as, for instance, in Leach's The Political Systems of Highland Burma (1954). Consequently Politics, Law, and Ritual is very much a defense by Gluckman of his approach, and he takes pains to explain very precisely just what that approach is.

Gluckman describes Politics, Law, and Ritual as “a statement of how one social anthropologist, working in the full tradition of the subject, sees the general problem of rule and disorder in social life ” (Gluckman 1965:xxiv, my emphasis). This a formulation that surely echoes a central concern of the colonial state. The particular area Politics, Law, and Ritual deals with is that “of political struggle and order, of law and social control, and of stability and change in tribal societies” (1965:xxi). Gluckman explains his choice of the term tribal as follows:

By “tribal society” I mean the kind of community which was once described by the term “primitive society”, a term now rightly rejected. Others call this type of community “pre-literate” or “pre-industrial”. These are appropriate terms, but I prefer “tribal”, since “tribe” was used to describe most of the communities of Europe, virtually up to feudal times. And forms of social organization akin to those communities, are what I am dealing with.

Gluckman goes on to define the characteristics of tribal society:

Basic to a tribal society is the egalitarian economy, with relatively simple tools to produce and primary goods to consume. The powerful and wealthy use their might and goods to support dependants; for they are unable to raise their own standard of living with the materials available.

He explicitly distinguishes “tribal” societies from peasant societies.

On the whole I judged that the study of peasants was another field. The study of tribal society has stimulated, and been stimulated by, the study of peasants. It would have produced a far more superficial book had I tried to draw on the wealth in this somewhat distinctive field, even though many of the social processes with which I am concerned are represented there.

We are not told, however, what it is that constitutes this “distinctiveness”; significantly there is no entry in the index under “peasant.”

For Gluckman the category “tribe” is essentially descriptive and unproblematic, referring to a straightforward “fact” of colonial life. Taking the passages I have quoted, we can summarize the basic characteristics of Gluckman's category “tribe” as follows. Firstly, “tribes” represent a distinct type of social organization, and one that was characteristic of an earlier, prefeudal stage of European history. Secondly, this social organization is based on an “egalitarian” economy. Since this egalitarian economy can also have hierarchies of wealth and power, its egalitarianism would seem to refer to the fact that the wealthy and powerful “are unable to raise their own standard of living with the materials available” and instead “use their might and goods to support dependants.” Thirdly, its technology is that of “relatively simple tools” producing “primary goods.” What I want to argue is that much of the meaning of this model of the tribe is to be found not so much in its explicit features, but in its silences—the questions with which it is not concerned. It is on some of these I want to focus, and on their significance within a colonial context.

Although “tribal” social organization is located as a particular stage within the development of Europe, the question of how “tribal” societies become transformed into nontribal societies is not addressed. The question of history, in the sense of nonreversible change, is for Gluckman, as for other functionalist anthropologists, not the business of anthropology. “Anthropologists analyze a society as if it were in a state of equilibrium”—equilibrium here being “the tendency of a system after disturbance to return to its previous state” (1965:279). In this context Gluckman goes on to explain, “While we are concerned with tribal societies it is easier to make this kind of analysis [i.e., assuming a tendency to return to equilibrium] because they were restricted in their external relations and their economies were stationary” (281). “Tribal” economies therefore are not only outside history, they are self-contained entities that can be analyzed in isolation from the wider colonial economy. By the time Gluckman was writing Politics, Law, and Ritual he was sensitive to the growing criticism of the functionalist paradigm, criticism which he saw as essentially misguided, insisting that “Every study of a particular tribe that I have cited in the course of this book, after analysing the tribal equilibrium, considers the tribe's position since it came under European domination” (285; my emphasis). This formulation merely emphasizes that the key entity to be analyzed is the “tribe” and that, although colonization may have brought changes, this basic entity, the “tribe,” persists, a particularly problematic assumption given that Gluckman's definition of tribal social organization defines it in terms of particular economic structures. Even if we accept “simple” technology and the lack of the possibility of direct economic accumulation as features of precolonial African economies—and this is questionable in itself—these are characteristics that are likely to be profoundly affected by incorporation into a wider colonial economy; we certainly cannot simply assume that they will persist.

Just as Gluckman's model of the “tribe” is silent as to how a “tribe” might cease to be a “tribe,” so too is it silent as to the history of how this form of social organization developed. The very nature of the model represses questions about how the structures of the rural areas in British Central Africa have come to have the specific form they have at this particular historical moment. However much the turbulent history of nineteenth-century or even eighteenth-century Africa may be acknowledged, there is an implicit, unspoken assumption that all this history happened to some basic entity, the “tribe.” Individual “tribes” may have disappeared and others come into being, and all may have been subject to that favorite process of functionalist anthropologists, “fusion and fission,” but in some essentialist sense the “tribe” as a form of social organization persisted.

It was this implicit assumption that enabled the RLI anthropologists to assume so confidently that it was possible to uncover the “tribal equilibrium” of peoples whose structures of political life had undergone a radical transformation little more than a generation previously. A crucial dimension of the Yao people's economy, for instance, prior to the imposition of pax Britannica, was slave raiding; similarly the precolonial Lozi state depended to an important extent on tribute labor. Yet Clyde Mitchell, who wrote about the Yao (1956), and Gluckman himself, who wrote about the Lozi,^[10] nonetheless treated these “tribes” as if such radical changes could be ignored and the continuing Yao and Lozi “tribal” structures, surviving since precolonial times, could be discovered through research on contemporary communities living under colonial rule. The point here is not that such a claim is ipso facto false, but that neither Mitchell nor Gluckman—nor the other RLI anthropologists—felt it necessary to demonstrate why this continuity could be assumed.

As I have stressed, Gluckman's brief definition of “tribal society” is not concerned with distinguishing this form of social organization from other forms, those characteristic of peasant society, for instance, or with explaining why “tribal society” can be treated as an autonomous entity despite its embeddedness within a colonial state. Essentially, the model is a descriptive rather than a theoretical one. I would argue that why Gluckman can, as it were, get away with presenting this vague and incoherent definition of the name “tribe” as a “scientific” and rigorous model is because standing behind it, and providing it with its real substance, is a powerful “commonsense” or popular notion of the “tribe.” A commonsense “fact” within colonial discourse—a “fact” that survives remarkably unscathed in current popular discourse about Africa—is that the basic set of social relations within which Africans live is the “tribe”; it is the nature of “tribal customs” that explain Africans' identities and behavior. Ultimately, what the definition of “tribe” boils down to in practice is simply “the unit within which Africans live.” What Politics, Law, and Ritual is concerned with is the exploration of aspects of this empirically given entity, the “tribe”; and the empirical reality within which the category “tribe” assumed its particular African meaning was that of colonialism and the problems of colonial “rule and disorder.” The reality was that there were colonized peoples among whom order had to be maintained and disorder avoided. Part of what the name “tribe” does is to locate the colonized not in terms of a power relation between them and their colonial overlords, but in terms of a separate world that they, but not their colonizers, inhabit. This is a way of seeing colonial society that, regardless of the intentions of individual theorists, helps divert attention away from its real power relations.

But if “scientific” categories gain resonance and power through their unspoken links with lay or popular categories, so these in turn gain authority and credibility through their links with “scientific” discourse. Indeed, many popular notions represent odd fragments of earlier scientific or philosophical discourses that have over time gradually drifted down into the great sedimented stratum of “common sense.” As Gramsci put it:

Philosophy and modern science are also constantly contributing new elements to “modern folklore” in that certain opinions and scientific notions, removed from their context and more or less distorted, constantly fall within the popular domain and are “inserted” into the mosaic of tradition.

The substantive meaning the term tribe acquired in British colonial Africa is a good example of the complex way in which “scientific” and commonsense categories reciprocally inform and shape one another.

In the next section of this chapter I want to shift the focus from the place of the name “tribe” within the work of Gluckman and the RLI anthropologists to the way “tribe” was used by locally based colonial officials. And this is where our story returns to North-Western Province and to the local dialogue between the colonial state and its inhabitants as to the identity of “the Kaonde.” Given the hegemony of the name “tribe,” for colonial officials this dialogue could not but center on the question, What are the characteristics of “the Kaonde tribe,” and what are the differences and the similarities between “the Kaonde” and other “tribes”? That “the Kaonde” were a “tribe” was simply taken for granted. Part of what hegemony means is precisely this establishment of particular names as no more than a way of indicating obvious and banal “facts,” the existence of which nobody but an overzealous pedant could question. Colonial officials seeking to understand the nature of their colonial subjects therefore could not but use the notion of the “tribe.” Using the reports and other writings of colonial officials stationed in what is now North-Western Province, I look first at the place of the general notion of the “tribe” in the thinking of these officials whose concern was not academic research but the practical problems of administration. I then go on to look at how such officials used the notion of the “tribe” to explain to themselves the nature of “the Kaonde.” Having described something of what could be termed the theoretical landscape in which the colonial state located “the Kaonde,” we can finally return to the question of how those so named, “the Kaonde” themselves, understood “being Kaonde.” Until that point, in order to keep, as it were, a question mark over the nature of this name I have usually used quotation marks when writing “Kaonde.”

Administering “Tribes”

The situation of British colonial officials stationed in rural districts, and especially those in the more remote regions such as the northwest, had a very particular character. Such an official was one of a tiny band of Europeans made up of a handful of other officials—all male—together with some wives and families, set down in the midst of a sea of Africans, often with a couple of hundred miles or more of dirt road between them and the nearest other Europeans. The autobiography of one such official stationed in Kasempa in the 1950s, gives a sense of the intoxication of power this could lead to—and of the heavy baggage of British literary references that these men brought with them.

Kasempa was real, the outer world but a shadow, and letters from it, from ghosts. It made many Europeans terribly assertive. One acted one's part upon an enormous empty stage: the “compleat man” was untrammelled if unseen. Grey of Falloden feeding pigeons, Gladstone hewing at trees, even Dr. Watson with a brisk manner and a bottle of iodine. One could play out such fantasies in real life with real people.

The job of the colonial official involved being both a clear and unambiguous symbol of the might and authority of an empire beyond challenge; it also involved managing the practical day-to-day realities of colonial law and order, collecting taxes, punishing criminals, and so on. In addition there was always a concern on the part of the colonial administration as a whole that the costs of administration, especially in areas like the northwest, seen primarily as labor reserves, should be kept to a minimum. This concern made the idea of Indirect Rule a very attractive one. Co-opting local authority figures as a cheap lower tier of colonial administration was one way of reducing the number of vastly more expensive British colonial officials.

The mental and physical boundaries between colonial officials and the colonized world over which they were set were always fraught and dangerous. On the one hand, the maintenance of their position as awe inspiring symbols of the empire—and indeed their security as isolated individuals usually far from much tangible imperial might—demanded that they maintain a proper distance between themselves and the colonized. This was also important for their own psychological well-being, particularly since the social conventions of colonial society were severe on those who strayed across this all-important boundary, and in such small isolated communities ostracism was a powerful sanction. The danger of “going native” was always one of the structuring fears of European colonialism. This is indeed one of the collective nightmares that underpins Joseph Conrad's Heart of Darkness, and gives it much of its power. On the other hand, the need to be effective administrators demanded that colonial officials should “penetrate” the mysteries of the colonized society; their “effectiveness” depended on their understanding of local realities. Then too there was the psychological need to impose some kind of conceptual order, and to be able to explain at the very least to oneself, why so much of what one was trying to do, such as implement the various colonial schemes for “development,” was not working. The colonial official therefore, had simultaneously to “get to know,” and remain aloof from the colonized. One result of this was a knowledge structured around essential, unalterable difference, a difference that explains why “we” (the colonizing power) have a “natural” right to rule “them” (the colonized).

For British colonial administrators throughout sub-Saharan Africa the central name in which this notion of difference was gathered together was that of the “tribe.” Within colonial discourse on Africa it was hegemonic that the basic unit of rural society was the “tribe,” and that “tribes” were the primary actors in rural life. The people colonial officials were charged with administering were seen as “tribal” peoples living according to “tribal” norms. The “tribe” was both a comfortingly familiar and yet “scientific” category. It should be remembered that almost all these colonial administrators were the product of an education system in which the classical authors of Greece and Rome loomed large; accounts of barbarians, Germanic tribes, and ancient Britons formed a basic part of their mental furniture. This was how Europeans themselves had lived in some distant past. The familiarity of the term meant that its meaning did not have to be spelled out; what tribal meant could be simply taken as self-evident, which removed any need to look too closely at just what it was that differentiated the colonized from the colonizer. It could function as a marker of difference, but a marker that also apparently explained that difference.

One colonial official was F. H. Melland, who spent eleven of his more than twenty years of colonial service in northwestern Zambia in a Kaonde-speaking area and wrote a book about “the Kaonde” entitled In Witch-Bound Africa: An Account of the Primitive Kaonde Tribe and Their Beliefs. A “historical digression” from this book illustrates how parallels from the Ancient World could be seen as offering lessons on the nature of local “tribal society.”

One of the most close parallels to our rule in tropical Africa is furnished by the Romans in Britain. For some time the Romans thought of nothing but law, order and discipline. Boadicea's rebellion taught Rome a lesson, and under a more enlightened policy…the Britons were taught to build houses instead of huts, to cultivate, to start industries, develop mines, export their produce, and so on: result, progress and peace.…That Roman policy laid the foundations of a progress that made those despised savages advance further than any advance dreamt of by the Romans.

The concept of the “tribe” was used both to explain Africans in general—all Africans were “tribal” and shared certain characteristics that marked them out from Europeans—and to explain differences between Africans. As Melland stated in the preface to his book, he hoped that although he had focused on the Kaonde, the book would be useful for

all who are going to live and work among similar Bantu peoples, even if far from the BaKaonde; for, while their customs, habits, and beliefs differ there is, still, a great similarity in these matters.…To acquire an insight into one tribe helps one to understand others.

And it was the “tribe” that was the basic unit in the system of Indirect Rule, even if the uncomfortable reality that some “tribes” did not seem to conform to the stereotype would insist on complicating things from time to time. The lack of fit between the “tribal” model colonial officials took for granted and what they saw happening around them could induce extreme irritation at the apparent failure of some rural Africans, such as “the Kaonde,” to behave as a “tribe” ought. The following two extracts are from reports written by the same colonial administrator in 1930 and 1931, the period when Indirect Rule was being introduced.

The difficulty in the Kasempa district is that the Kaonde have little or no tribal organisation.…the idea of sitting together in a Court, is extremely repugnant to them. As one elderly petty chief.…put it to the writer, “I am a lone elephant bull, I wish to walk alone,” and this is the attitude taken up by the majority of petty chiefs which is not conducive to the success of native courts as at present constituted.^[11]

In the Kasempa Province we have the Bakaonde, the Alunda, the Andembo and other lesser tribes, all of a low standard of intelligence, who…now have very little tribal sense.

The Bakaonde tribe in particular, now have little, if any, tribal organisation or cohesion.…If the people were intelligent, and not so scattered, it would be easier to revive a tribal sense, and with it there could be a better application of indirect rule.^[12]

The basic category “tribe” included a wide range of different “tribes” with different “tribal customs,” and within colonial discourse there developed an extensive repertoire of “tribal” stereotypes. The annual reports and district notebooks that colonial officials were required to keep reveal the major role played by these stereotypes and the way the different characteristics of different “tribes” were evoked to explain what was happening in particular areas. Each official would tend to have his own favored “tribe,” or “tribes,” and his own bêtes noires. In the northwestern region two groups of people who were continually contrasted were “the Kaonde,” and “Luvale” and “Chokwe” people. One basic difference between the two groups was their association with agricultural systems organized around different staples: sorghum in the case of “the Kaonde” and cassava in that of “the Luvale” and “the Chokwe.” The tribal stereotypes used in the district notebooks and other official documents, however, go far beyond any simple difference in cultivation patterns, as the following quotations illustrate.

In 1940 we find one official writing,

The Kaonde are naturally of fine physique being often six feet or more in height and well proportioned. As these people have no cattle largely owing to tsetse fly their physical condition must it is presumed be largely due to the high dietic qualities of the kaffir corn.^[13]

While in 1948, another writes,

The two elements in the population present a contrast in health and fitness which one cannot but ascribe to diet to some considerable extent. The immigrant element in the population [i.e., Luvale and Chokwe] are well fed, and cheerful, whilst the Kaonde are under nourished morose, and diseased. Not one was free from some deformity and I can only say that they represented about the lowest ebb of humanity that I have yet seen in Northern Rhodesia. They had no nuts, cassava, or goats at their villages, and were short of the kaffir corn which is their staple diet. “We are Kaonde: we don't grow groundnuts or cassava or keep goats, we leave that to the Chokwe and Lwena [Luvale]” which was the invariable reply heard over and over again on this tour, epitomises the inertia of the Kaonde who would prefer to remain under nourished and diseased rather than expend a little more energy in cultivation and adopt new practices.^[14]

Ten years previously, another official, more enthusiastic about “Kaonde potential” had noted, “As a whole the Kaonde build good villages with well thatched huts and kitchens attached and should form a fairly fertile soil in which to start a movement for village improvement.” ^[15] This view was echoed in 1948: “I was disappointed in the living conditions amongst the immigrants [i.e., Chokwe, Luvale, Luchazi, Mumbunda]. Their villages were, on the whole, extremely dirty with poor houses. The Kaonde villages had large, well built houses and were usually much cleaner.” ^[16] This view was not shared, however, by another official in 1957: “Village housing is, generally, good amongst the Chokwe, who take a pride in their houses…and poor among the Kaonde, who prefer to spend their money upon bicycles rather than sound houses.” ^[17] About the same time, yet another official, who clearly did not have much faith in “Kaonde potential,” wrote, “The Kaonde as a whole are not a tribe which believe in a vast amount of exertion and it is, therefore, here that one finds the element of self-help less well developed than in some other parts of the Province.” ^[18]

These comments on physique and housing are particularly illuminating about the role of “tribal” stereotypes and their power to shape what colonial officials saw when they looked at the rural world around them, in that ostensibly what was at issue were straightforward, observable “facts.” Those who saw either Kaonde of “fine physique,” “often more than six feet tall,” or “under nourished morose, and diseased” Kaonde, all of whom had “some deformity” and who “represented about the lowest ebb of humanity…in Northern Rhodesia,” were observing the same population at more or less the same time. Just as those who saw “good” or “poor” Kaonde housing had traveled through the same rural areas.

There are also more obviously subjective comments.

We visited one Mbundu village and one Lovale village. Apart from the fact that they have grown groundnuts and the Kaonde do not, I could see nothing of worth in these two villages; they have few children, are very dirty, and their housing standards are poor. They were obviously more primitive and less manly than the Kaonde who treat them with considerable contempt.^[19]

Quite often, as here, the concept of the “tribe” shades into that of the “race.” The notion that humankind is divided into distinct “races,” each with its own specific characteristics passing from generation to generation, had become increasingly hegemonic in nineteenth-century discourses not only around colonialism, but around history in general. As with the “tribe,” the notion of “race” was used both in academic and “scientific” discourse,^[20] and in that of popular common sense. What “race” named in the interwar period is captured in the first edition of the enormously influential Cambridge Ancient History, published in 1924.

Ancient peoples come upon the stage of history…in a certain order…each with a make-up congruous with the part they will play…history presupposes the formation of that character…in the greenroom of the remoter past: and the sketch which follows…is intended…to describe how men came by these qualities of build and temperament.

As with “tribe,” much of the explanatory power and persuasiveness of “race” as a category was, and indeed still is, its combination of being apparently “scientific,” describing what seem to be undeniable “facts” (such as that there are inherited physical differences between people), and at the same time including a shifting range of highly subjective psychological traits and propensities. The “scientific” authority here provides credibility to the purported mental characteristics. The power of these porous grab bags of mental and physical qualities, which makes them so impervious to specific challenge, depends on them being both ordinary “commonsense” terms whose meaning is shared within a particular culture, and inherently vague. Take a quality like “manliness,” used in the earlier quotation, for instance. We can assume that within the British public school culture of colonial officials, everybody “knew” what “manliness” meant but at the same time would probably have been hard put to explain in any precise way just how you could measure a “tribe” on the scale of manliness “scientifically.” As with the predictions of astrologers, terms that seem to say something but which can be so stretched as to fit almost anything, can be profoundly reassuring. In the cloudy soup of tribe or race, as Raymond Williams puts it, “Physical, cultural and socio-economic differences are taken up, projected and generalized, and so confused that different kinds of variation are made to stand for or imply each other” (Williams 1983:250).^[21]

But how did those the colonial officials so confidently named “Kaonde” identify themselves? And what part did they play in the dialogue about who and what “the Kaonde” were?

The Administered

Did the people living in the northwest during the colonial period describe themselves as belonging to “tribes”? Clearly at one level they did. As in the quotation on page 66, people would indeed refer to themselves as “Kaonde,” “Luvale,” or some other “tribal” label. Local interpreters used the term tribe when writing up court cases in English for the records. and there were ready answers when the district officer, or a visiting anthropologist, asked about “tribal” law, or “tribal” customs. But the meaning of this term tribe, and the resonances it had, were not necessarily the same for those who belonged to “tribes” as they were for the colonial officials or the RLI anthropologists.

There are four main linguistic groups associated with the northwestern region: the Kaonde, Lunda, Luvale, and Luchazi. Oral traditions suggest that the Kaonde-speakers came from Luba country to the north, partly in response to the emergence of the Luba and Lunda states, in a long series of migrations from the sixteenth to the nineteenth centuries. These migrations, as Roberts argues, “were probably a prolonged and mostly rather haphazard process of small kinship groups gradually moving to the south and east in search of new land” (Roberts 1966:106; see also Jaegar 1981:46–68). The Lunda, Luvale, and Luchazi came from the north and west, and were still pushing east in the nineteenth century (Roberts 1966). In the course of these migrations existing populations tended over time to be either displaced or incorporated. The Mbwela was one group of people to whom this happened in the area now associated with Kaonde.

It was during the long period of their migrations that the Kaonde language emerged as a distinct language, and it is the Zairian languages, Sanga, Luba, and Hemba, to which it is most closely related (Wright 1977:109). When I was collecting oral traditions in the late 1970s, people would often begin their account of Kaonde history with the formula, “Atweba twi baSanga” (We are Sanga people). Indeed, prior to the imposition of colonial rule, it is unclear to what extent the Kaonde constituted any kind of political entity. In the previous half century or so, certain Kaonde headmen^[22] had managed, with the help of guns, ivory, and slaves, to raise themselves and their lineages above those of their fellows, and it seems as if some kind of a centralized power structure was beginning to emerge; but things were still very fluid when the imposition of colonial rule froze a particular moment in a continuing and turbulent power struggle. The largest political unit in the period prior to colonialism appears to have been a groups of clans. It was as leaders of senior lineages within particular clans that chiefs seem to have emerged and began to amass power, gradually gaining dominance over a whole clan. Over time certain clans might manage to establish themselves as royal or chiefly clans—the emergence of a royal clan meaning that other clans were now commoner clans.

The Kaonde term for clan is mukoka, and interestingly the only English-Kaonde dictionary in print (Wright 1985)^[23] gives mukoka (pl. mikoka) as one of the translations of tribe. The other translation of tribe it gives is mutundu (pl. mitundu), a very general term that can be glossed as “kind” or “type,” and can be applied to people, animals, and plants as well as inanimate objects. The point here is that there is no term in Kaonde which refers to a particular kind of political organization, the “tribe.” There were, and are, only mikoka and actual empirical entities, the Kaonde people, the Lunda people, the Europeans, all of which may constitute a distinct entity in many different ways. These apparently “ethnic” categories are in fact highly protean.

At the time of my fieldwork, people would use labels such as “Kaonde” or “Luvale” to refer to a whole variety of different things, not only particular settlements or whole regions, but also agricultural systems, ways of life, particular beliefs, and so on. Such “ethnic” terms were used in a taken-for-granted, unselfconscious way, and, because such categories were so unquestioned and undefined, people had no problem in using them in different contexts to refer to a bewildering variety of realities. Given the lack of precise linguistic terms in the language, it seems reasonable to assume that this fluidity was also characteristic of the colonial period. The comment quoted on page 66 illustrates one usage of the category “Kaonde”: “We are Kaonde: we don't grow groundnuts or cassava or keep goats, we leave that to the Chokwe and Lwena [Luvale].” This statement was made to silence a colonial official bent on “developing” the Kaonde, not given as a definition of some primordial Kaonde identity. Indeed, there may have been individuals who would have described themselves as Kaonde, possibly living in the next village, growing groundnuts and cassava and keeping goats. Probably all that is common to these ethnic labels is that these were groups who perceived themselves—or were perceived by others—as groups sharing a language and into which people were in general born; and even then, an individual's ethnic identification might change over the course of her or his lifetime.

In the 1980s the Europeans connected with the local, and primarily German funded, Integrated Rural Development Project tended to be seen locally as part of a common entity, magermani. Not all the Europeans working for the project were German, and local people were sometimes puzzled to discover that not all magermani spoke the same language. Magermani could indeed be seen as sharing a common culture, the culture of the “development expert” and foreign aid worker. Although this is by no means a homogeneous and uniform culture, from the perspective of local people those who belonged to magermani constituted a distinct “tribe” just as much as did “the Kaonde” or “the Luvale.” But this local understanding of the name “tribe” as signifying no more than some empirically existing group, whose shared characteristics could take many very different forms denied the basic premise of the colonial discourse on the “tribe” that “tribe” named a particular kind of social organization associated with a particular level of social development. However much it might have represented local perceptions, any definition of “tribe” that went against this colonial hegemony had little chance of becoming a part of the dialogue between the RLI anthropologists, the colonial state, and the people of the northwest.

Whether or not the Kaonde constituted a distinct political entity at the beginning of the colonial period, during the colonial period they certainly learned that in the world of colonial administration Africans were located within “tribes” and it was primarily in the name of a “tribe” that claims could be made on the colonial state. This was the only voice with which Africans were allowed to speak. And this is why by 1951 we find one Kaonde chief saying in a meeting of Kaonde chiefs organized by the colonial government,

We are not here to discuss matters of precedent or to discuss totems. We have heard about the Barotse and the Bemba tribes; we must throw away the totem barriers [i.e., think in terms of “tribes” rather than clans] and be one people. We must be proud to be called Kaonde.^[24]

It had become clear that the only language in which the colonial state was prepared to listen to claims by Africans was a language of “tribal” law and “tribal” customs, a language that assumed that these laws and customs derived from an ancient and unchanging past—in other words, that they were “traditional.” Since the model of the “African tribe” whether that of anthropologist or colonial official, presupposed homogeneity within the “tribe,” the question of just which interests within the “tribe” were reflected in “tribal” law and custom did not arise within the colonial discourse.^[25] But to what extent and in what ways did the name “tribe,” as it was used in colonial Zambia, help more generally to shape the realities it was supposed to explain? Answering this question provides a good example of what hegemony looks like in practice.

Of “Tribes” and Hegemony

For the anthropologists associated with the RLI and the colonial officials, the concept of the tribe helped map the reality of the colonized world. It gave a particular shape to what they saw as the problems of “law and disorder” in British Central Africa—as when struggles around working and living conditions on the mines were seen as a problem of rural “tribesmen” becoming “detribalized” rather than, say, struggles between labor and capital. This naming also had a number of powerful effects for those categorized as “tribal.” The hegemonic naming of Africans as “tribal” within colonial discourse contained them in a particular way. It froze them into supposedly homogeneous communities, gemeinschaften, remote both from the workings of the world capitalist economy and from history, ruled by “tribal law and custom” handed down from an ancient past. In line with the principle of Indirect Rule, questions of the role and nature of the state and its relation to individual colonial subjects were deflected onto the role and nature of the “tribe,” which for the African, as the colonial eye saw it, was the state. Similarly, inasmuch as major realities such as the development of a huge copper-mining industry in Northern Rhodesia were pushed to the sidelines of analysis while the searchlight shone on the “tribe,” the notion of the “tribe” did alter the landscape that was seen by colonial officials, the RLI anthropologists, and indeed by the colonial state's African subjects—just as the colonial administrators' stereotypes of “manly” or “diseased” Kaonde affected how they saw those they met on their tours.

Africans were inescapably enmeshed in the reality of institutions based on the category of the “tribe,” such as the Native Authorities of Indirect Rule. In addition, they were also enmeshed within an overarching colonial hegemony with at its heart a notion of irreducible difference—and not only difference but inequality. What this difference consisted of in Africa was above all a difference between the modern, the civilized, the developed, and the tradition-bound “tribe,” which was condemned by its very “tribal” nature to slumber in a sleep of superstition and indolence until awakened by the kiss of the colonial prince. Those Africans who had the temerity to awaken on their own to any of the new opportunities offered by the new colonial world were likely to be dismissed as dangerously “detribalized.”

The miscalled “mission boy”, the worker on the mines and on farms, the house boy (all equally anathema to the chiefs and headmen), pick up bits of knowledge, lose their old tribal and religious checks and become a disintegrating, destructive element.

The name “tribe” helped provide those engaged in the colonial enterprise with a morally justifiable gloss explaining why colonial rule was necessary. Africans were “tribal” and therefore simply could not handle the economic or political institutions of “modern” society. At the same time, the name “tribe” helped those engaged in running the colonial state to feel that they understood the world of the colonized, that they knew where the roots of colonial “law and disorder” lay, and that their little world of busy administration had some purpose and function. The fear that the whole colonial enterprise might in the end be absurd and meaningless, and ultimately doomed, is another of the basic fears that underlies many novels of colonial life, from Conrad to Orwell and Forster.

What the “tribe” as a name came to mean in colonial Zambia is in part a story of what “hegemony,” or one dimension of it, can mean as a lived reality. The relationship between names such as “tribe” and the realities they point to is like that between maps and the physical topography they represent. Just as particular maps take particular forms depending on what they are to be used for, and are judged ultimately on how useful they are, so too in the case of the names used to map social reality. There is always a dialogue between our representations and that which they represent, as we attempt, concepts and categories in hand, to chart our way through the ever-shifting sands of social life.

One measure of hegemony is the degree to which the conceptual maps of dominant groups, and their naming of reality, manage to maintain their unassailable authority for subordinated groups even in the face of a lack of fit between those subordinated groups' experience and how this experience is named to them. After all, the authority of a map depends to a large extent on its provenance and the professionalism with which it was produced; only some in society have the necessary resources to produce authoritative maps. The problem for the subordinated is, How do you move from a sense that you cannot trust the maps and, even though you may pack them in your rucksack, it is best not to use them, to a confidence that not only can you make your own maps, but that they are superior to the existing maps? And this shift is not only about confidence; hegemony is also about the power to demarcate arenas of struggle—in other words, the degree to which a ruling group is able, through its practical control of institutional structures, to confine the challenges of subordinated groups within a terrain mapped out by the dominant group. However creative Africans might have been in “imagining” their social relations, and however perceptive as to the real relations underpinning colonial society, they were confronted with a colonial order that in very concrete ways insisted that demands or claims put forward by Africans use the discourse of “tribes.”

The story of the “tribe” in colonial Zambia provides a good introduction to my attempt to map out the various landscapes of power inhabited by the people of Kibala and Bukama. Firstly, because it illustrates something of what hegemony means in practice, and the important role played by the power to name. Secondly, it shows us the complex interaction there is between the supposedly scientific categories produced within the academy and those of nonscientific “common sense.” Finally, the story of the “tribe” reveals with a particular clarity the embeddedness of the academic pursuit in existing hegemonic relations. Social theorists are necessarily embedded in their historical moment; all accounts of social reality, whatever their claims to scientific rigor, are always produced in a particular political context that necessarily shapes how that social reality is approached. This book too, of course, is imprisoned in its time just as much as were those of earlier anthropologists. My story of the “tribe” will, I hope, not be read as an exercise in the “condescension of posterity,” but rather as a reminder of the complex threads that bind theorists to their times. And bearing that salutary lesson in mind, we are now ready to begin mapping out the political and economic landscapes of Kibala and Bukama.