Peter R. Schmidt

5.1 FOREGROUNDING THE ISSUES

Sitting in a smoke-filled house in northwest Tanzania, an aged oral historian once turned the tables on me during one of our many interviews by asking, ‘Why do you Whites call our history prehistory?’ Taken aback, I mumbled a pathetic ‘What do you mean?’ in response, but he remained silent, knowing that I had understood his meaning. My command of the language was insufficient at that point in my fieldwork to explain the peculiarities of our divisions of history and prehistory. By the standards of the day, this gentleman had received a good education, sufficient to be exposed to colonial ways of representing local history as prehistoric. His question mirrors one of this book’s central themes: the rendering of the oral histories of Others as prehistory, outside of history.

A necessary beginning point for exploring Western attitudes and philosophical dispositions underlying the separation of prehistory from history is to examine some of the writings that amplify alterity between the two domains rather than their interrelationships. Some of the most penetrating insights into how this division arises come from Michel de Certeau (1988, 2000). A French historian, de Certeau captured the history of this intellectual tradition better than most Western scholars.

In this chapter, I use a postcolonial critique to uncover essentialized historical representations in Africa (Schmidt 2009a, 2009b), particularly those that conflate myth with prehistory and then add a dose of fiction. I explore a chain of reasoning that reveals an elaborate entanglement of present and deep past within colonial Uganda to demonstrate an

Extensive manipulation of prehistory, resulting in one of the most visible interpretations in the historiography of ancient African history. By unveiling this historical fabrication, it is possible to illustrate how colonialist treatment of Bunyoro legendary histories, widely called myth, in Uganda considered it prehistoric, that is, outside historical experience and thus open to manipulation because it was perceived within the colonial setting as without factual basis. The now-widespread popularity of this colonial-era fabrication—sometimes referenced in speeches by the Ugandan head of state—is ensured by its unquestioned place in contemporary history instruction, a pedagogy that valorizes a prehistory that arose out of colonial needs and imaginations. Postcolonial analysis comes into play when we examine power relations and how history is made within the political and cultural settings of the colonial world.

My final goal is to examine structural analysis as it has been applied to oral traditions, some of great antiquity. Structural analysis has been deeply stigmatized within anthropology as an analytical methodology that imposes timelessness on oral texts. I intend to show how the universal application of this critique has marginalized structural approaches in historical studies.28 Because structuralism has much to offer in deep time studies, it is important to understand how it has been diminished and made subaltern in the repertoire of historical analyses by those who see structuralism as strictly synchronic. This assessment cuts to the heart of the prehistory/history divide as reified in our everyday language in archaeology. The stereotyping of structural analysis as a synchronic imposition that transforms history into a timeless past has not come under critical scrutiny, nor have we explored the potential of such analyses for unwrapping interconnections between deep time history and more recent history. I will locate my study in the literature of African history and archaeology, where a number of innovative attempts at marking periods of change and isolating the mixing of historical genres have yielded positive results, yet results that have been submerged by conventionally authoritative voices.

It is also important to understand that there is an alternative scholastic legacy that successfully examines folk texts, written and oral, to find forms of elaboration at both paradigmatic and syntagmatic levels, elaborations that mark significant cultural and institutional changes. These insights lead to a realization that structural analyses do not ipso facto make pasts into timeless myth and, moreover, that structural analyses have the potential to help sort through historical manipulations and entanglements of the deep past (and its fragmentary inclusions) with more recent histories.

5.2 TROUBLING OTHERNESS

Michel de Certeau asserts that one of the historian’s primary responsibilities is to introduce troubling differences so that we can come to understand that ‘no society is totally homogenous and unified, that there will be new irruptions of a troubling otherness’ (Giard 2000: 19). This perspective captures our goal in this book: to introduce troubling alternative thinking into historical archaeology and to rile the waters of complacency in a discipline that has drawn hard definitional boundaries to ensure homogeneity in its treatment of the past. De Certeau’s thinking on history is also pertinent to our theme because he addresses the writing of history—what we desire in our practice of historical archaeology. His discourse on writing compels reflection, especially that inspired by Foucault’s treatment of power, viz.:

Writing produces history. On the one hand, it accumulates, it keeps an inventory of the secrets from the West, it loses nothing, it preserves them in

An intact state____On the other hand, it declares, it goes to the end of the

World, toward those destined to receive it according to the objectives that it desires—and ‘without budging an inch’, without having the center of its

Action being moved____With writing the Westerner has a sword in his hand

Which will extend its gesture but never modify its subject. (de Certeau 2000: 132; emphasis added)

With this language, de Certeau prefigures postcolonial analyses,29 as he clearly sees that the power of writing has profound implications when practised either in colonial or postcolonial settings. Issues embedded in the power of writing engage de Certeau in an examination of those without writing, dependent on speech to pass on the past. His perspective on those without writing resonates with scholars who document the historicity of unwritten, orally transmitted texts. He poignantly captures the Western mindset when it comes to history vis-a-vis those ‘outside of time’, that is, those who use oral forms of history-making: ‘Speech is now located in an entirely different position. It does not preserve. . . . Speech, to the contrary, has much to do with custom, which in turn “turns truth into falsehood”’ (de Certeau 2000: 133). Historical accounts held in oral form, he reasons, are seen in the West as fable, and fable ‘is a drifting away—adjunction, deviation, diversion, heresy, and poetry of the present’ (de Certeau 2000: 133).

Articulating the processes by which we erect intellectual barricades— the drawing of boundaries that relate to power over history-making—is perhaps de Certeau’s most important contribution to our discussion, a phenomenon seen within the practice of historical archaeology with the inscription of boundaries that create alterity vis-a-vis what has come before literary times. De Certeau’s exegesis parallels the themes of this book, to wit: ‘historiography separates its present time from a past. . . . Thus its chronology is composed of “periods” (for example, the Middle Ages, modern history, contemporary history) between which, in every instance, is traced the decision to become different.’ De Certeau goes on to suggest that, ‘In their respective turns, each “new” time provides the place for a discourse considering whatever preceded it to be “dead”, but welcoming a “past” that had already been specified by former ruptures’ (1988: 4). In the history of North American historical archaeology, the decision to become different arose when Deetz (1967, 1991), Fontana (1965), and others (e. g., Schuyler 1978) (see Lane, this volume) circumscribed historical archaeology to make it strictly the practice of archaeology about the colonial and immediate postcolonial experience in North America. These boundaries were later modified with the inscription of comparative colonialism (Leone 1999; Leone and Potter 1999) and more generalized approaches concerned with capitalism and its impacts—a distinctive Western fixation that, no matter how important, is peripheral to the full spectrum of history-making among myriad peoples around the globe. Archaeological practice and thought occur within these rigidly prescribed boundaries, excluding how people construct history in cultures outside this specific Western experience. But more is involved than rigid boundaries. Embedded within such prescriptions are deeper ideas about the Other, those who lived outside our historical time—those in a state of prehistory when intersected by the Westerner who entered their worlds wielding the sword of history.

The differences that de Certeau discusses are manifest in the isolation of historical archaeology from prehistory, constructed as non-history that requires, as Deagan (1988) and others have argued, a different set of techniques and theories in its study—characteristics that additionally amplify its alterity (see Lightfoot 1995, this volume). The rupture between written history and deep time history, of course, was created centuries ago by writing and later enlarged by printing. With this distinctive break—fuelled in part by printing technology—arose new understandings of the past that privilege writing. De Certeau warns that such ruptures do not produce a clean break with the immediate past, but that they invariably carry with them remnants of that past, and that ‘whatever this new understanding of the past holds to be irrelevant’ (e. g., prehistory), it will be bedevilled by the appearance of remnants of a deeper past that arise ‘on the edges of discourse of its rifts and crannies’ to ‘discreetly perturb the petty order of “progress” or a system of interpretation’ (de Certeau 1984: 4; compare Freccero 2001).

De Certeau’s insights resonate with a growing consciousness about the provincialism of historical archaeology that compels us to further perturb the order. Yet, de Certeau also cautions that to interrogate the new order ‘has become unthinkable... [if the] new identity [is] to become thinkable (de Certeau 1988: 4; emphasis added). To keep a highly circumscribed historical archaeology safely protected from the ‘unthinkable’ or ‘above questioning’ is a major maintenance project in North American historical archaeology today. Trouillot’s (1995) discourse on the unthinkable as beyond the pale helps identify how power plays out to obscure the unthinkable (historical archaeology as it is practised in Asia, Africa, and parts of Europe) while validating the thinkable. Those who do not subscribe to the new (thinkable) identity are subalternized—often thrust into a liminal state where their unthinkable work remains in a kind of scholastic purgatory (see Mignolo 2000 for an exegesis on the processes by which subaltern conditions are created). Those who engage the unthinkable and work ‘on the boundaries’ to dismantle subaltern conditions and neutralize forces that create them also imagine and seek a better anthropology practised within historical archaeology.

This is territory known well among Africanists who work with deep time histories. On the boundaries, Africanist subalterns stir up prehistory and archaeology to draw out entanglements, show linkages, and explain change. When knowledge of entanglements between the deep past and present does emerge in debates about history, then discussion about those entanglements are categorized in North America as unthinkable and ignored. De Certeau’s (1990) conceptualization of citation is also important here, for his focus on the use of quotations in historiography as ‘absolute weapon[s] of make believe’ (de Certeau 1990: 315) sets the scene for understanding that recourse to documents and their quotation—of whatever variety—produces reliability. This valorization of knowledge is central to what he calls a ‘reality effect’ (Carrard 2001: 472), which imbues writing with legitimacy. The opposite side of this phenomenon is the absence of quotation or citation, casting subaltern accounts deeper into the abyss of the unthinkable.

This mapping or inscription of new identities by those who make history is a uniquely Western and colonial trait. De Certeau illustrates this by referencing newer forms of history in India that never displace the older representations. Using Dumont’s observations of Indian historiography, we learn how a ‘process of coexistence and reabsorption is a principal fact of Indian history’ (de Certeau 1988: 4; Dumont 1964). Dumont also draws from the Merina of Madagascar and the Fo of Dahomey—two important African illustrations of how deep time is wrapped into more recent historical representations, examples of pre-sencing of the past.

African history-making provides numerous additional illustrations about how other cultures weave the deep past into their representations of the recent past, seeing no rupture or need to separate the two. My research among the Haya of northwest Tanzania illustrates how historical sensibilities—revolving around issues of political legitimacy—digest and incorporate deep time histories, weaving them into the fabric of recent history (Schmidt 2006). Reference to deep time histories places political authority in a more secure position vis-a-vis local knowledge systems linked to sacred places on the landscape (Schmidt 1978, 2006; Schmidt and Walz 2007a). Such creations of new historical syntax eschew rigid definitions that underwrite separation. They depend on synthesis of different pasts, without prehistory and history as impeding concepts.

5.3 ENTER THE POSTCOLONIAL CRITIQUE

A central tenet of the postcolonial critique holds that the colonizing world erases the histories of the colonized as a way to gain more effective economic and political advantage over subject populations (see Liebmann 2008a; Liebmann and Rizvi 2008; Lydon and Rizvi 2010; Schmidt and Patterson 1995b). A colonized population bereft of historical identity is more easily manipulated, and a colonized population stripped of historical identities encompassing technological accomplishment and scientific advancement, for example, is a population more easily compelled to labour compliantly without challenging the conqueror’s superiority. The colonial project that we have seen around much of the globe in the form of erasure of local histories and their replacement with essentialized metanarratives continues today under the guise of North American historical archaeology (Schmidt and Walz 2007a, 2007b). Such hegemonic processes, arising from the boundarymaking just discussed, depend on non-reflexive practice (the unthinkable); those engaged in such practice remain unaware of how they are implicated in actively promulgating colonial legacies. We may unveil complicity in such engagement by showing how historical archaeology controls production of knowledge and peripheralizes other societies’ knowledge-making.

Historical archaeology in North America is a fit subject for postcolonial analysis. The initial focus of North American historical archaeology on the colonial period solidified its identity as practice valorizing the colonial agent and rarely, if ever, the agency of the colonized subject— something relegated to contact studies or prehistory rather than historical archaeology. An exclusive focus on the European experience in the New World and the marginalization of Native Americans as primary agents in the formation of the colonial experience is closely linked to the construction of Native Americans as survivors of prehistoric life. Prehistoric life in the Americas was replaced (the rupture) by literate, historic-minded Europeans who were capable of recording their own histories and occasionally making notes about the prehistoric populations with whom they came into contact and eventually colonized (see Kehoe 1998, this volume).

5.4 CREATING THE SUBALTERN IN MYTH AND PREHISTORY 30

Is treated as beyond question. I use Mignolo’s concept of subalterniza-tion, accepting the active practice of ‘making the subaltern’ to understand how fragmentation and oppression of alternative knowledges occur. This requires that we also be vigilant in recognizing ‘subaltern reason striving to bring to the foreground the force and creativity of knowledges subalternized during a long process of colonization of the planet’ (Mignolo 2000: 13).

Representations of oral traditions as myth, especially by missionaries who were often the first interlocutors in African cultures, facilitated erasure of local belief systems in possible competition with Christian doctrine. The first European representation of Ugandan oral traditions was published by Ruth Fisher (1970 [1911]), one of the first Victorian female missionaries posted to Uganda (Posnansky 1970). Her publication focused on the lifeways and oral traditions of the Banyoro of northwestern Uganda, though it was titled Twilight Tales of the Black Baganda to appeal to a broader audience in Britain.

There is no denying the importance of Fisher’s contribution to Uganda history, particularly as the oral traditions she published were collected using embandwa or spirit mediums as informants. Embandwa (or ‘witchdoctors’ in Fisher’s missionary lexicon) were the most informed keepers of ancient oral traditions. Yet, it is important to keep in mind that these tales have been truncated and manipulated to fit missionary sensibilities. Fabrication of these texts occurred without apology and with an authority that stuns us a century later, to wit: ‘Heaps of non-essential details have had to be cleared away, and in many cases modifications been [sic] made, or passages entirely discarded, to purify the story and render it suitable reading to the general public’ (Fisher 1970 [1911]: xlii; emphasis added).

Assumptions of historical ignorance among local people are stated with poignant intensity: ‘The people generally are strangely ignorant of their past, and evince very little curiosity with regard to it’ (Fisher 1970 [1911]: 64). Whatever value Fisher personally recognizes has to do with the possibility of local recognition of a ‘Creator’ underlying the overwhelming detritus of myth and fiction: ‘Under the legends and fiction of these central African people—under the thick refuse of it all, there underlie fragments of a primeval revelation that have not entirely been lost even after all the ages’ (Fisher 1970 [1911]: 64; emphasis added). A glance at her contemporaries’ quest for primordial religion remnants (e. g., Frazer’s Golden Bough) shows how Fisher considered herself an archaeologist mining her African site for the much-sought Urgott.

It is the Western fiction that myth is untrue that permitted Bunyoro oral traditions to become legitimate fodder for sifting and excising portions of prehistory seen as useless, offensive, or insubstantial—a process that simultaneously created subaltern histories by keeping them outside the corpus of published Bunyoro history. Moreover, oral traditions were seen as tainted by their inclusion in the provenance of witchdoctors, those who opposed the growth of Christianity. Not surprisingly, at the same moment that Fisher assumed licence to manipulate oral traditions in Bunyoro, parallel beliefs prevailed among colonial officers in their treatment of ancient history as they searched for ways to enhance the political goals of the colonial regime. Such manipulations unfold during Uganda colonialism when the history of ancient Bunyoro was diminished after a British alliance with Buganda kingdom led to key Bunyoro counties being ceded to Buganda in the 1900 Uganda agreement. These territories included Mawagola, where the famous and very extensive Bigo earthworks are located (Fig. 5.1). Bunyoro myth, seen by British Christians as untrue creations of ‘fetish worshippers’, was transformed by the colonial pen as circumstances required. The particular oral traditions within our focus here were Bacwezi tales commonly related by embandwa spirit mediums in Uganda and northwestern Tanzania. Colonial manipulation unfolds in 1909 with the publication of an article titled ‘Ancient Forts’ in the Uganda Official Gazette. We do not learn until more than two decades later (Wayland 1934) that the author was D. L. Baines, a district officer (Schmidt 1990, 2006).

The Baines account must be examined from the context in which he wrote it—as administrator of a territory that was under the recent control of Buganda kingdom. The presence of Bigo had elicited an early interest by British residents, who sought to tie it to ancient Bacwezi history, apparently as a means to add historical depth and to legitimize their indirect rule of these territories. Thus, it is hardly a revelation to learn that the first of three oral traditions related by Baines is from Buganda, whose hegemony over the region was but nine years old: ‘The most generally accepted tradition amongst the Baganda attributes the construction of this place to the “stranger” (Mugenyi) a personage who is supposed to have entered Uganda from the north soon after Kintu’s time, roughly 600 years ago’ (Baines 1909: 138; emphasis added).

This characterization of an enormous anthropogenic landscape feature, under the control of the Banyoro for centuries, suggests an outright fabrication that suited the needs of both the British and Baganda colonial interests. The second oral tradition is supposedly a Banyoro tradition that says a Bunyoro chief erected the earthworks as fortification against attempts to recover the royal cattle he had supposedly stolen, a tale whose source is unattributed. Baines references a third oral tradition that he says was challenged by the Ganda, namely, that ‘a chief of

Fig. 5.1. Map showing key sites in Tanzania and Uganda, including the Bigo earthworks

Mawogola by name Mugenyi’ constructed Bigo as protection for raids by the Ankole people. Baines’s commentary reveals that he used selfinterested Ganda informants to vet his narrative. Prominently missing in all of these treatments is any mention of Bacwezi association with the site, now the most renowned archaeological site in East Africa. A pastiche of different oral traditions from different groups, the Baines compilation lacks integrity and credibility. The cavalier fashion in which oral traditions were collected and then represented is testimony to a colonial mentality, captured in Mignolo’s (2000) concept of ‘coloniality of power’ that creates a timeless past and diminishes historical accountability.

The Baines account is the first link in a chain of reification that is illustrative of how other more specific and localized accounts have been made subaltern and over time erased because of their subalterneity. The next link in the chain is Pierre Gorju (1920), a Catholic priest whose seminal work on interlacustrine societies accepts the Baines account about Bigo—thus demonstrating the power of quotation that de Certeau (2000) unveils—and presents Mugenyi of Bigo as the minor Bacwezi figure of that name (Schmidt 2006). Like others to come, Gorju underwrites Baines’s treatment by affirming that there are other (unmentioned) oral traditions about Bigo—additionally subalternizing any alternative accounts. Wayland’s republication of the Baines account in 1934 without critical commentary—further concretizing its legitimacy— was immediately followed by ‘The Riddle of Biggo’ published by Sir John Gray (1935) in the Uganda Journal. De Certeau’s discourse on quotation again comes into play here, for the complete reprinting is an extensive quotation that reaffirms the reality effect of the Baines account, lending it additional legitimacy through reiteration.

Gray (1935: 229) further elaborated the Bigo-Bacwezi fiction when he wrote, ‘one of the many traditions regarding the earthworks is that their constructor, Mugenyi, was buried there’. This unattributed affirmation by an otherwise respected colonial historian is not supported by any other oral traditions. Gray’s references to Ganda oral traditions pertaining to the region not only play into the colonial political game by affirming Buganda political hegemony but has subsequently been dismissed by Kiwanuka (1971), a Ganda historian. Perhaps most fascinating at this juncture is Gray’s citation of Peter Bikunya (1927), one of the earliest indigenous historians writing in Kinyoro (language of Bunyoro). He affirms that Bigo was called Biggo bya Mugenyi, though again without acknowledgement that it meant ‘the forts of the stranger’. By 1935, the fabricated myth of the Bacwezi at Bigo had five links in its chain—Baines, Gorju, Wayland, Bikunya, and Gray—all of whom were colonial functionaries or products, civilian or missionary. Gray’s sanguine conclusion (Schmidt 2006), against all the evidence, that ‘the Bacwezi, of whom Mugenyi was one, seems to fit in with the evidence afforded by the earthworks themselves’ (Gray 1935: 233) sets the scene for additional fabrications.

This politically constructed history sealed in place the biases introduced by myth being considered untruthful and timeless prehistory. It led to the masking of vast corpuses of local histories, particularly alternative histories expunged from testimonies of spirit mediums, clan histories, and histories in opposition to the British-recognized royals, all considered outside of the purview of discourse and analysis. This story does not end in 1935, however. A half-century after the Baines account, Roland Oliver (1953), the doyen of African history and professor of African history at the School of Oriental and African Studies, added his authoritative voice by citing a manuscript in press written by two indigenous writers, Katate and Kamugungunu (1955); he accepted—contrary to the overwhelming evidence available—that someone called Ruyanga (Ruhanga) was a Cwezi who ruled from Kashozi, opposite Bigo (Oliver 1953: 136). This narrative confused both history and geography (see Schmidt 2006). It also added two more links to the chain of reification (Baines, Gorju, Wayland, Biku-nya, Gray, Katate and Kamugungunu, Oliver). Oliver (1959) followed with another unsubstantiated affirmation, asserting that the Bacwezi had indeed constructed Bigo. No one stepped forward to question what oral traditions sustained such a construct. The chain of reification continued with Shinnie’s (1960) excavations at Bigo in 1957, which also did not entail an examination of the oral traditions that drove the enquiry. This research gave further credence to the constructs of earlier years, along with Merrick Posnansky’s excavations of 1960 (Posnansky 1966, 1969) that similarly tried to link archaeological evidence with oral traditions (Fig. 5.2).

Though Posnansky (1969) was to treat such associations with judicious scepticism (Schmidt 2006), Oliver did not. He rushed to judgement with a conclusion that declared the colonialist construct as historically sound, using preliminary archaeological evidence before Posnansky published his report: ‘According to tradition, the last capital centre of the last of the Chwezi kings was in the celebrated entrenched earthwork site at Bigo on the Katonga river’ and that ‘Recent archaeological investigations have tended to confirm the traditional evidence that Bigo was the seat of a Hima dynasty’ (Oliver 1963: 181; Fig. 5.3). This completed the chain of reification with ten links (Baines, Gorju, Wayland, Bikunya, Gray, Katate and Kamugungunu, Oliver, Shinnie, Posnansky (secondhand), and Oliver), tying down a colonialist process of making an essentialized history, a fictive account that conflated fabricated myth with what was perceived as prehistory and drove alternative ways of seeing the past into subalterneity. Such interpretations have now become enshrined in East African historiography, with alternative analyses that dismantle this chain of reification joining the indigenous accounts in their subaltern status (e. g., Schmidt 1990, 2006; for similar processes see Gould, this volume), being ignored and mostly un-cited within East African historiography. On the opposite side are the now popular accounts in Uganda that accept and valorize the colonialist accounts about the supposed capital site of the ancient Bacwezi Empire.

Fig. 5.2. Bigo earthworks in Uganda c. 1960 (photo: Merrick Posnansky)

The key issue here is that when it was in the colonial interest, myths or ‘prehistoric stories’ were accepted as historically accurate (without rigorous investigations), appropriated by the colonial state, changed and mixed together, transformed into the written word, and printed—a materialization that conferred historical status. Moreover, the association of these newly minted histories with a significant ancient site further materialized and underwrote these new histories. Any separation of prehistoric from historic in this instance would have defeated the state’s political goals and undermined the manipulation of local history on behalf of state interests. Licence to conflate and freely manipulate prehistory and myth was an integral part of how colonial renderings of history became elevated to historical canon. The widespread popularity of this fabrication—sometimes referenced in speeches by the Ugandan head of state—is reinforced by its unquestioned place in contemporary history instruction, a pedagogy that valorizes a prehistory that arose out

Of colonial needs and imaginations. The idea that contemporary political rhetoric in Uganda about deep time identities draws from these colonial constructs is a poignant lesson about how such entanglements play out in lives today.

As we explore historiographies like that of western Uganda for issues that do not fit with other evidence—be it material or textual— contradictions between texts (oral and written) will inevitably arise from such an examination. When we find and engage such contradictions, then we tread on a path that also diminishes distinctions between prehistory and history, entanglements in which deep time evidence (material and oral) and historic texts are brought into contact (see Hantman, Aguilar and Preucel, this volume). Such entanglements have occurred within colonial settings such as Uganda, but they may also unfold through indigenous manipulations (e. g., the Haya incorporation of myth into more recent genealogical history) or by archaeologists using ancient material evidence to challenge and alter representations in a historiography (Mrozowski, Walz, this volume). When any of these permutations enter our discourse, we have the potential to develop an historical archaeology that escapes the negative conundrum of distancing the Other and may build vital histories that draw on all forms of historical memory.

5.5 ANTHROPOLOGICAL SCEPTICISM ABOUT THE HISTORIES OF THE OTHER 31

Not alone in his assertion that structural analysis amplifies the synchronic, treating oral texts as timeless. Given that Levi-Strauss and his followers dealt mostly with myths and legends—the so-called timeless tales of the primitive—this issue naturally grew to become a major concern with those who wanted to consider deep time as possibly embedded in these oral genres.

Vansina’s (1965) initial optimism for the use of archaeology to verify oral traditions appears to have waned by the time he published Oral Tradition as History in 1985, in which he treats archaeology in a series of sidebar discussions to other topics. This means little examination of much of the experimental archaeological work that was unfolding in East Africa (e. g., Schmidt 1978; Skully 1979; Soper and Golden 1969; Van Noten 1972), southern Africa (e. g., Maggs 1976; van der Merwe and Skully 1971; Wright and Kus 1979), and West Africa (e. g., Connah 1975; Darling 1984; Garlake 1974, 1977; McIntosh and McIntosh 1980, 1986), particularly the important research of Roderick and Susan McIntosh into oral traditions and archaeology in Mali (e. g., 1980, 1986). The latter research showed the importance of bringing multiple sources of evidence to bear on widely accepted historical narratives and opened important insights into Mali’s ancient history. Research by the McIntoshes showed how European and Arab ethnohistoric accounts about Jenne-jeno’s origins, fixed to the thirteenth century ad, were contradicted by ancient indigenous oral traditions pointing to a much earlier genesis for this ancient community. The ancient material evidence excavated at Jenne-jeno affirmed that there had been a mid-first millennium rise of urbanism—a finding that escaped the cliched representations of the published literature while rewriting history bolstered by ancient material evidence—a wonderful example of the entanglement of history and prehistory that has escaped wider understanding and acknowledgement.

In a similar vein, in 1978 I published an exegesis of several oral traditions about the ancient Bacwezi and the more recent royal Hinda clan in Buhaya, northwestern Tanzania (see Fig. 5.1 for site location). It was an examination of how oral traditions may be used with archaeology, as inspired by Vansina. It is informative that Vansina makes no mention in his 1985 book about the analyses used to unravel the antiquity of Haya oral traditions but instead remarks (Vansina 1985: 187): ‘P. Schmidt dug on the spot where according to tradition the Tower of Babel had once stood and found an ironsmelting site from as far back as 500 b. c. The memory of the site was kept but had no relation with the cliche of oral tradition.’

Vansina’s recognition of the role of memory in preserving knowledge about the Rugomora Mahe site for over two millennia, though important, omits how and why this happened. The answer emerges elsewhere in his discussion—that association of an oral tradition with a concrete place or object enhances its continuity and vitality (Vansina 1985). Vansina’s (1985: 145) gloss of the Kaiija or iron tower oral traditions as a Tower of Babel myth (a cliche that signifies differences between people) portrays the tale as a timeless universal, a normative interpretation that usually pertains to linguistic or cultural differences.32 As such, it denies any additional examination of historical process within this Haya genre—a curiously incomplete and synchronic conclusion by one of the most eminent voices on oral traditions.

The iron tower of Katuruka in the Rugomora Mahe oral traditions, simply explained, does not relate to differences between people. Instead, it is a specific origin or etiological myth that uses the structural skeleton of the Tower of Babel cliche—communicating about the origins and memorializing ancient ironworking in the region. By imposing a universal interpretation on the Katuruka iron tower tale, we are also left without any sense of the myth’s place in the symbolic life of the Haya and other people of East and Central Africa, especially when such stories are explicitly associated with iron production and human reproduction (see Schmidt 1997, 2006, 2010a). As an etiological story that memorializes the role and antiquity of iron, the iron tower myth also captures the central symbolic reproductive role of iron among the Haya. To interpret the iron tower myth as a Tower of Babel cliche rather than as a specialized etiological myth occludes a deeper understanding of what the iron tower myth means for Haya sexual practices and human reproduction and how these related meanings are linked to technology and history.

5.6 STRUCTURALISM IN COMPREHENDING DEEP HISTORIES

When searching for ways to sort out multiple and varied tales about Rugomora Mahe’s iron tower, I recalled how Victor Turner had once

Taught in a Makerere University seminar that Levi-Strauss’s development of a structuralist approach opened ways to see and better understand structural change and similarity when applied to various periods, particularly when they were genealogically organized. He reasoned, and I resonated with the possibilities of his thinking, that it should be possible to recognize reordering of cult influence and practice if the oral traditions provided different structural readings over time when ordered by genealogies (taking into account that genealogies, too, can be manipulated). To affirm the usefulness of structural analysis in historical studies contradicts the thinking of many historians and anthropologists who have, mostly correctly I believe, maintained that the structural method is arbitrary and not replicable. Apart from the search for universal dialectical relationship, the chief objection has been that structuralism treats myth and other oral traditions as disconnected from historical process, as synchronic, timeless expressions (e. g., Barnes 1971; Fabian 1983; Hughes 1970; Leach 1966, 1970a, 1970b; Paz 1970).

Vansina (1983, 1985) also belongs to this persuasion; that is: ‘The analysis cannot be falsified because ethnographic validity is irrelevant, the discourse being unconscious. Whether members of the culture involved approve or not is irrelevant. All that can be verified is whether the images used actually occur in the text. What one makes of them is arbitrary and cannot be falsified’ (Vansina 1985: 165). Such characterizations became commonplace by the 1980s, but in the interim there had also developed a literature finding structural analysis useful to historical analysis, an occurrence presaged by Edmund Leach’s observation that ‘The intelligibility of diachronic transformations is no greater and no less than the intelligibility of synchronic transformations. By implication, the only way to make sense of history would be to apply to it the method of myth analysis that Levi-Strauss is exhibiting in his study of American mythology’ (1970b: 98). Leach’s proposal was precisely the direction that innovative analyses took in Africa, Latin America, and the Pacific.

The absence of a thorough and more exhaustive discussion within African history and archaeology about using structuralism for important historical insights is mystifying. Both Feierman (1974) and Atkinson (1975) in history and Schmidt (1978) in archaeology during the 1970s made profitable use of structural analysis with oral tradition texts. Feierman’s (1974) structural treatment of the Mbegha myth among the Shambaa about the kingdom’s founder (a hunter), for example, demonstrates the significant affinities between the Mbegha story and an earlier myth about Sheuta, also a hunter who became chief of the Shambaa. In reviewing this analysis, Vansina wants a better explanation about why this structural repetition occurs, but he fails to acknowledge or discuss that it is structural analysis that successfully unveils this important historical relationship—an omission that subalternizes innovative work that deconstructs the timeless, prehistoric other to find historical substance. Feierman’s analysis was a harbinger in African history in its use of structuralism to strip away mystifications and relate a tale to rites of passage and a much more ancient myth. That his method was not openly discussed in methodological debates in African history is informative.

The utility of structuralism was kept subaltern, not to be taken up in discourse about analysis of oral texts in Africa. We see a similar but more explicit treatment emerge with Atkinson’s (1975) analysis of early Buganda myth. Atkinson sorted out narrative patterns in early Buganda mythology through structural analysis, and even though these patterns were logical and persuasive, ‘[Atkinson]... documents the full extent of the structuring’, the analysis remains unacceptable because, ‘unfortunately, he interprets the contents away in structuralist fashion’ (Vansina 1985: 170). This was a turning point. The most eminent practitioner and critic of oral tradition analysis found value in such analyses but simultaneously denied their application to historical analysis. Such critiques discouraged other experimental attempts and created a subaltern genre of historical analysis—the use of structuralism.

More than a decade before Vansina repulsed these innovative attempts to make sense of possible change embedded in oral texts, anthropologists had been examining other ways to apply structural analysis to historical and folk texts. Eugene Hammel (1972) was one of the first, arguing that folk texts reflect the times—different institutions and social contests— during which they were told and recorded. Using diachronic control over the texts, Hammel reasoned that we should be able to observe important historical processes, struggles over institutions, changing values, and behaviour. Hammel’s research into The Three Bears,33 one of the early historical experiments with structural analysis, provides explicit guide-posts for textual analysis. His analysis demonstrates how dated texts displayed increasing elaboration through time, particularly in the addition of more elaborate binary relationships—the paradigmatic elements of tales distilled from content. As well, Hammel’s innovative treatments show how elaboration in the sequencing of tale element—the syntag-matic component of his structural analysis—could also be discerned. For Hammel, elaboration and increased perfection of a text through time marked critical periods in Western life—changing values, behaviour, and institutions. Given how the concept of increased perfection is linked to tightly dated published texts, we may question if this index has utility for oral traditions in Africa or other regions; yet rapid increases in binary elements clearly held promise as diagnostic markers for social change, particularly when different social groups hold similar evidence that is structured by genealogical principles of sequential ordering.

A key question to emerge in this alternative school of thought was posed by Edmund Leach (1966, 1970a), who asked about biblical texts: ‘How... does this essential diachrony of the traditional hermeneutic relate to the synchrony of a structural analysis?’ (1970a: 250). Leach’s goal in his analysis of biblical texts about the legitimacy of Solomon was to show that the chronological sequence in biblical history has a structural relevance and that the stories with all of their contradictions and incongruities must be considered as a unity, as the uniting structures depend on chronological arrangements. Leach reveals how structural analyses of Old Testament texts show a three-phase unit in which the same characters appear but in different costumes. Thus, he is able to demonstrate how myth-history is consistently structured over the chronology that the biblical texts set out, showing that variation and different themes such as endogamy/exogamy and Israelite/foreigner are integral to all phases of the stories. This perspective puts structural analysis to work in explaining sociological and political relations and how these reverberate in the way the texts are fabricated. Both Hammel and Leach developed ingenious ways to finesse the structuralist conundrum of synchronic dominance, creating instructive guideposts in understanding the potentials and possibilities of such analyses for African oral texts.

When anthropologists of African oral tradition tried to apply these insights to oral texts, they met the barricades erected by Vansina. These are best illustrated when Roy Willis used structural analyses in his Herskovits lecture on Fipa oral traditions to create insights into historical process among the Fipa of southwest Tanzania. He observed that ‘oral material can be made to yield meaningful qualitative information in terms of social-evolutionary sequences’ (Willis 1976: 2).34 Willis acknowledged Feierman’s work as influencing his perspective and he argued that his analyses reveal three structurally distinct levels of traditional lore, with myth expressing cosmological symbolism as the deepest stratum. Willis observed that the middle stratum of tales from three different ritual-political sources are randomized and lack the binary relationships found in the first. The third period, during which there were two centralized states, is marked by significant structural change manifest in texts by the absence of symbolic elements and a regular sequencing of events, what we might consider as increased arbitrariness—fitting Willis’s observations with those of Hammel and, later, Eva Hunt (1977).

Realizing that his construct lacked a firm chronological framework, Willis called for ‘archaeological research in Ufipa [that] will eventually provide such a framework... Such research would also constitute an important and possibly decisive test of the hypotheses on Fipa social history which I am proposing’ (1976: 17). Rather than receiving a positive response by archaeologists and historians, Willis’s ingenious proposal was dismissed: ‘Roy Willis... [was] not well-versed in Levi-Strauss, but a Leachian, at least when he published his Herskovits lecture’ (Vansina 1983: 308). While the substance of this critique remains opaque, it nonetheless conveys a tone and lack of receptivity that subalternizes such experiments simply because they applied structural analyses.

When I turned to analysis of oral traditions gathered among the Haya of northwestern Tanzania, Hammel’s short monograph had just been published. It opened new possibilities for structural analyses at the same time that Atkinson and Feierman were taking up related but as yet undisclosed analyses. I reasoned that since genealogical ordering is a characteristic that is highly valued by the Haya in their oral traditions about social groups (clans as well as larger political entities such as kingdoms), then structural analysis seeking to find rapid expansion of binary relationships should point to significant social change. Analysis first focused on various genealogical representations by different social groups about Kiziba royal history of far northern Buhaya. Different clan histories agreed that at about the seventh generation from the beginning of royal history there was a moment of significant social experimentation and institutional development.

Using structural analysis of these texts, I found that there was a proliferation of binary oppositions that pointed to and affirmed the creation of a new royal cult. Paradigmatic relations arising from the content focused on oppositions between expensive/cheap, white/black, dangerous/benign, Bacwezi/Bito royal, and so forth. The creation of a new royal cult to counter the influence of Bacwezi spirit mediums who steadfastly opposed the throne for generations accounts for a significant increase in these binary oppositions. This structural analysis stretched across many generations of oral traditions, highlighting patterns of opposition that otherwise would have gone unnoticed within the content of individual generations. Once aware of the proliferation of symbolic oppositions arising out of new cult creation, it was then easier to recognize parallel processes in the more southern kingdom of Kyamut-wara at a similar generational moment.

A second focus in structural analysis of Haya oral traditions takes inspiration from both Hammel and Leach and illustrates how we may engage the longue duree without stigmatizing it as prehistory. Working through a large corpus of tales from different social groups in Greater Kyamutwara kingdom, I noted similar sequencing in the structure of the oral traditions of King Rugomora Mahe and those associated with the ancient Bacwezi. A syntagmatic analysis of the sequencing of events in some of the stories about Rugomora Mahe showed a pattern that replicates some but not all of Bacwezi myth. These sequences were drawn from bits and pieces of Bacwezi myth and clearly testify to a significant complexity in the genesis of the Rugomora myth. The analysis also examined paradigmatic relationships between the various elements in the story, revealing paradigmatic relationships that often bore significant affinities between Bacwezi myth and the Rugomora legend within royal Hinda history. There was not a wholesale adoption of Bacwezi mythology into royal legend; rather, it was Bacwezi myth that structured part of the content and form of the Rugomora legend. The names of the actors were changed along with some other details, yet without structural analysis, I might have inappropriately concluded that the history of Rugomora Mahe was a bounded historical legend relatively free of other influences. Furthermore, without such analyses, most anthropologists would be inclined to relegate the Rugomora Mahe myth as a timeless adoption, as an anachronism rather than a commentary on negotiation and political appropriation of sacred space (for parallel processes see Hantman, this volume).

The reasons for the conflation of ancient myth with more recent historical legend relate to the hegemony of the Hinda royal clan being extended into territory and over sacred shrines that were previously under the control of indigenous groups. Myth sequences adopted directly from Bacwezi myth may be explained by royal appropriation of the oral genres of the most powerful practitioners of religion and ritual—the embandwa spirit mediums of the Bacwezi. Such appropriation yielded two advantages: it helped neutralize considerable embandwa political influence and it created the illusion that the royal usurpers had a continuous link to the ancient past, a key element in their legitimacy as the new ruling group. Decoding the Rugomora stories and their rootedness in the Bacwezi past has been closely linked to unveiling the deeper meanings associated with the iron tower myth and the geography of the tower. The iron tower myth, linked to Kaiija shrine and to reproductive iron symbolism, took a different kind of potency when archaeology

Revealed an ancient iron forge dating to 500 bc, directly tied to the construction of the tower—the earliest iron working evidence found in East or Central Africa. The Hinda royals, archaeology disclosed, took over this ancient shrine about 1675, a date congruent with the royal and other clan genealogies. And so, structural analysis does not ipso facto diminish history; it can lead us to consider more seriously indigenous entanglements of deep time narratives with those of the recent past—the use of oral texts for the presencing of the past—no less potent than material remains of the past (see Aguilar and Preucel, Lane, this volume) and hardly something to be banished as prehistory.

The structural anthropologist Eva Hunt presented additional innovative uses of structural analysis while also employing archaeology as a metaphor for how structuralism could be used together with historical texts to isolate key social change. Her explicit goal in Transformation of the Hummingbird (1977) is ‘to demonstrate the value of using a documented historical approach, in combination with structural analysis, to the interpretation of mythical symbolism’ (Hunt 1977: 25). For Hunt, ‘an anthropologist can become the archaeologist of his society’s culture. But for the cultural anthropologist, unlike the real archaeologist, few materials remain that can be dug from the ground. The stratigraphic layers are in the culture itself, in media that are sometimes ephemeral’ (1977: 32). By employing the archaeological metaphor, Hunt burrows into Zinacan-tan historical records—different strata—to discover bits and pieces of symbolic elements that are not disconnected as they first appear but interconnected with each other in a highly structured system. The basic armatures—the structuring components over time—are revealed by structural analysis. Hunt’s analyses show that Levi-Strauss’s stance that history erodes structure needs to be reconsidered in view of how she is able to show how deep structure endures and ‘remains sharp and clear’ (Turner 1977: 9), an observation that resonates with de Certeau’s discourse on remnants of the past in the present as well as other recognitions of the presencing of the past (see Lane, Aguilar and Preucel, this volume; Schmidt 1997, 2006).

Most useful in Hunt’s analysis is her examination of arbitrariness and elaboration, both being clear signposts for change such as the growth of new cults and the overthrow of institutions such as priestly hierarchies. Since much African oral tradition touches on institutional history interpenetrated by ritual and myth, the methods proposed by Hunt resonate with the goals to isolate major periods of change that may have been inscribed in the material record. I have suggested (Schmidt 1997, 2006) that an analysis of tropes—symbolic language held in metaphor and metonymy—will reveal moments of elaboration or increasing arbitrariness.

Marc Gaboriau once presciently remarked that Levi-Strauss was misunderstood, that he did not see an antimony between diachrony and synchrony—that, in fact, ‘It should be possible to study in terms of structure the passage from one state to another in any system or society’ (Gaboriau 1970: 162). Certainly those who have sought to put such a perspective into practice and derive historical insights from structural analyses—thinkers such as Leach, Hammel, Feierman, Atkinson, and Willis—have handed us a legacy that has remained subaltern in the study of African history and archaeology. Now that we are more aware of its potential for unveiling deep time histories in the present, it is time we revisit its application to appropriate texts, particularly those that may be linked to places and material remains that reach into antiquity, whatever the age, period, and time segment.

5.7 FINAL REFLECTIONS

The Uganda and structuralism case studies discussed here help us understand that boundaries between prehistory and history are meaningless within an African setting as well as the other settings discussed in this book. History’s flow as process has not been interrupted by an artificially inscribed boundary, no matter how unthinkable the domain on the other side of the boundary may be considered. The conflation of myth with prehistory and its subsequent transformation into history in Uganda teaches us that when political expediency or missionary agendas triumph, separations become meaningless and give way to fabrication and downright fiction. When this occurs, it carries with it profound implications for alternative renderings of local history that are subalternized and eventually disappear because of that status and the power of the published, popular meta-narrative, a process witnessed in many indigenous settings within the Americas (e. g., Gould, Mrozowski, this volume).

We have also learned that the capacity of structural analysis to disentangle deep time texts from those of later historic times is tainted by stereotypes of timelessness and as a consequence now occupies a subaltern position within anthropological analysis of oral traditions. Regardless of the subaltern status of structural analyses, when it is successfully put to work for historical anthropology and historical archaeology (see Yentsch 1988), it is important to celebrate those successes and continue trying to find ways of understanding the entanglements of recent history with deep time history. Insofar as structuralism can be shown to be an effective tool for unravelling historical processes, it is compelling to follow what is effective and ignore the intellectual dead ends proffered by critics.

As we reflect on the practice of historical archaeology in North America, we see that the entanglements of deep time history with more recent history warn us that if we inscribe our practice only within thinkable colonial history, we consign most of global history to the unthinkable. If we forget the unthinkable and open ourselves to new ways of seeing how other cultures think history—as the application of structural analyses instructs us—we may find historical processes unfolding in ways not previously imagined, in ways that enrich how we make history.

As we continue to explore possibilities for seeing history as process extending into the deep past, it should be inevitable that prehistory as a concept will fade, leaving bits and fragments of its former hegemony behind to remind us that we once celebrated divisions that occluded the flow of human history rather than opening it to better comprehension.