The first section discussed the differences in the forms of the two genesis traditions. The second discussed differences in the areas of residence, the histories, and the contacts with others sustained by the Baloho and Babambo, and speculated on the influence these factors may have had on the different forms of the traditions. But of course, the local cultural systems from which the two narratives are drawn have themselves undergone significant change during the past 150years, and these changes may well have been reflected in changes in the forms of the traditions. This section explores how such changes in the traditions have influenced our understanding of the past. To do so, it will compare Ijwi variants of the traditions with those elsewhere. The similarities and differences between these narratives may provide some clues to the nature of change that each tradition has undergone.

The cultural milieu of the Baloho has evolved in two ways: at the level of specific politics—in terms of their relation to the royal family and how that affects their status on the island; and simultaneously at a more general level, in their changing relation to more general cultural norms on Ijwi over the last century. The change in their political status has already been examined above. It was noted that while this may have encouraged the retention of an earlier tradition, such alterations were not likely to have served as the stimulus for the etiological crea-tion of the genesis tradition in the form in which they tell it today. The second aspect, however, concerning the broader cultural changes on Ijwi, has not been considered.

It was pointed out that certain features of the east, where the Baloho live, indicate a former way of life similar to that of the Batembo peoples now living west of the mountains, in exactly those areas from which the Baloho claim to have emigrated. The cultural norms portrayed in the Baloho tradition—the fragmentary nature of the tradition itself, the forest motif, the sack theme—are associated both with regional differences and historical differences on Ijwi. It appears that the cultural norms for most of the island’s population, including many Baloho, have gradually shifted away from the norms portrayed in the Baloho tradition. The traditions may therefore be expected to retain certain forms and emphases that reflect this early period, and hence to bear witness to the changes. In fact, in a free-form tradition such as this, it is possible that there are elements of every period in the life of the tradition; the problem is to determine which periods are retained in which elements.(In fact, it is possible that a traditionretains cultural elements from a period prior to the events to which it appears to relate.)

There are two distinct types of Baloho tradition on Ijwi, the one fragmentary, the other (less common) a longer version. As noted above, because they overlap, the fragmentary forms appear to be part of a larger tradition: recounted by members of the same clan—and only that clan—their synthesis is reflected in the longer traditions. But though they are rare and show signs of more elaboration, the longer versions are nonetheless similar to a tradition now found in Mpinga, the area near Kalehe on the western mainland where the senior Basibula line (the descendants of Kamerogosa) is now centered.The Mpinga tradition, however, though similar in form to the Ijwi tradition, applies to a different historical episode than that of the Ijwi tradition, an episode distinct in time, place, and actors. Precisely because their forms are so similar but the historical referents in the traditions are so distinct, the Mpinga and Ijwi versions of the tradition must have drawn either on each other or on a source com-mon to them both to have attained such structural homogeneity.

To account for their similarity in structure, it will be necessary to outline briefly the Mpinga tradition, which relates to the last quarter of the nineteenth century, when Ntale, the Havu king, fled the attacks of the Rwandan king Rwabugiri, who had earlier killed Ntale’s father (Nsibula).As recounted in the Mpinga tradition, Ntale’s closest advisors placed him in a sack and traveled with him widely, feeding him in the sack as they traveled, even to the court of Rwabugiri, without Ntale being discovered. In its presentation, this episode is obviously very close to the Kamo-Mwendanga tale, except that the historical context (Rwabugiri’s attacks on Mpinga) is explicitly noted, whereas the enemies seeking Mwendanga are not specified. The parallel between the Mpinga and Ijwi episodes is strengthened by the fact that the threatened king (in the sack) in each of the traditions (Ntale in one, Mwendanga in the other) is referred to by the sobriquet of “Kali-mumvumba.”Several themes are repeated in the two traditions: the sack (notably), the mobility, and the context of protecting the king from his enemies.

In determining the exact relation of the two traditions, we are left with two alternatives: either one derives directly from the other, or they both derive from a common prior source. (For reasons to be discussed below, I exclude the possibility of independent invention in such a restricted area of such pervasive cultural interaction.) Though the data do not bear directly on this problem, certain historical considerations make it unlikely that elements of one of these traditions were drawn on directly in the formation of the other. Because the Mpingatale refers to events that occurred well over half a century after those associated with the dynastic schism at Nyabihunge,and because, as discussed above, it is most unlikely that the Ijwi version is of recent provenance, it is therefore also unlikely that the Ijwi tradition is derived from that told in Mpinga. This conclusion is reinforced by the localization of the Baloho in the far eastern portions of Ijwi—a single small group with no apparent recent contacts to Kalehe.

The Baloho tradition on Ijwi relates exclusively to the Ijwi branch of the Basibula family. In most variants, it ignores even Kabwiika’s relation to the senior branch of the Basibula line: Kabwiika’s ancestry, for example, is nowhere mentioned in the Baloho traditions on Ijwi. In addition, the Ijwi and mainland dynasties have had little interaction with each other since the schism, and even so, the royal line on Ijwi has no knowledge of the Baloho tradition at all—except that Basibula narratives do recognize“Kalimumvumba”as associated with Mwendanga. But Mwendanga is little known in Mpinga, and Kamo is unheard of. It therefore seems improbable that the Mpinga line selected the formal elements of the Baloho-Ijwi tradition and applied them to their own history, just as the con-verse is also highly unlikely.

What seems more plausible, given the distribution of this element, is that both the Ijwi and Mpinga versions derive their common form from an earlier stereotype present in the area. In one form or another, the sack theme, for ex-ample, is found over a wide area to the northeast, south, and west of Ijwi. To the south, the Rundi speak of a regent wrapped in a “mat” and transported thus.Another variant on the theme is found in Bugoyi, northeast of Lake Kivu.This region shares certain sociological, linguistic, and historical ties with areas west of Lake Kivu, and these elements help to distinguish this area from other areas of Rwanda.(In Bugoyi, however, the sack theme is incorporated into a popular tale, or umugani, rather than one that is supposed to relate to empirical events of the human past.) Therefore, the Basibula areas seem to be well within the do-main of the recorded presence of this tale across a wide region. Nonetheless, it remains true also that in such areas this form of narrative is not connected with court culture or with political (or at least royal) themes.

At first glance the presence of this stereotype among the Havu seems associated with Basibula royalty, because both the Mpinga and Ijwi traditions refer primarily to members of the royal family. But simply associating it with Basibula royalty does not account for the timing of its appearance, given that the Basibula dynasty preceded the emergence of the traditions now extant; nor would such an explanation conform to the distribution of the tradition on Ijwi, which isrestricted to the Baloho in the east and apparently unknown to the royal family or the royal court.

But both traditions also indicate ties to the forest culture similar to that of the Tembo peoples today, societies located west of the Mitumba Mountains. Formerly, it appears, certain elements of Tembo culture were more widespread in the region, both on Ijwi and on the mainland areas west and north of Lake Kivu.Therefore, despite the presence of this tradition among the two branches of the Sibula royal family, the sociological associations common to both the Baloho and Mpinga traditions relate more to Tembo contacts than to a narrowly defined concept of Basibula history. The Baloho on Ijwi have already been discussed in the context of their probable origins in the west, from an area today in-habited by the Tembo. On the mainland, the Basibula dynasty arrived in Mpinga only in the generation prior to Ntale;before their arrival the area was inhabited by Hunde and Tembo peoples—two categories closely related culturally— who in fact remained in the area of Mpinga after the arrival of the Basibula dynasty. Indeed, Mpinga today retains intensive contacts with peoples identified as Tembo, those who live west of the Mitumba Mountains.

Thus these two traditions—the Kamo-Mwendanga tradition on Ijwi and the Ntale-Rwabugiri tradition in Mpinga—both apply to periods when analogous historical circumstances occurred within similar sociological contexts: when Basibula royal flight from an enemy coincided with important Basibula contacts with Tembo-type peoples. The thematic features emphasized in the traditions are compatible with Tembo-type cultural values (at least as they exist today). Therefore, it can fairly be postulated that the present forms of these traditions have resulted from grafting elements of specific Basibula dynastic history to the tropes of a common Tembo-type tradition. Such a Tembo connection helps explain the restriction of the Kamo tradition on Ijwi so narrowly to the Baloho because, like the royal family, other groups in the east did not share the cultural orientation of the Baloho.

In traditions such as these, each narrative element is, of course, a part of both the form and content of the tradition. But where there exists a common episode that recurs in different historical contexts (as with the Kamo tradition on Ijwi and the Ntale tradition on Mpinga), it is possible to distinguish the com-mon form from the more local content of each tradition. In this case, clearly the sack theme, the mobility, and the theme of royal protection by a commoner are common to the two. Assuming the traditions have remained independent of each other, as these two seem to have done, it is possible to account for suchcommon factors either by the structural requirements of the traditions them-selves or by historical additions; where they are structurally required, the probability of their historicity at a literal level is diminished. Obviously, this is true in reference to the sack, which is clearly a literary device.

On the other hand, as noted above, the presence of that common theme may well apply to a historical situation at a more general level: while not literally true, the sack theme does indicate a common theme of protection of the king from his enemies; and this aspect of the episode, as shown above, is confirmed by supporting evidence in each case. Thus the structural elements of the com-mon episode may well be metaphorical references to a specific perceived historical situation.

Other aspects of the tradition, those not required by the common structure, relate differently to the historicity of a tradition and hence need be evaluated differently. Here it is what is specific to one context, not what is common to several variants, that is important. Although no definitive categorization is possible distinguishing specific historical referents from common cultural themes, individual elements may tend toward one end or the other of the spectrum between these two poles. Those elements that refer primarily to a specific situation rather than a common cultural theme will most likely refer either to some specific historical referent or to contemporary perceptions and situations for that particular place. Where the latter can be ruled out as a possibility and where there are supporting data for the historical situation, this element can be used as historical evidence.

However, this can best be done when two conditions apply: where the element in question can be dissociated from the broader structural requirements of the tradition, and where it finds specific support from independent data. For ex-ample, the Kamo tradition highlights the relation of the king to his benefactors as an important theme, emphasizing the friendship of Kamo and Mwendanga and the services rendered by the Baloho to the king. In contrast, the Ntale version almost entirely neglects these aspects. Because this particular element cannot be ascribed to the structure of the tradition, it may have been generated by more local considerations; whether or not these considerations are historical in origin, or generated from contemporary concerns, need be demonstrated from an analysis of the corpus of local data. The preceding presentation has shown that it is unlikely that this particular Baloho element developed out of present considerations, and that other evidence supports Kamo’s historicity. Therefore, it seems probable that Kamo existed as a historical character and that he did develop some kind of affective tie with Mwendanga.

Although the structure of a tradition may provide valuable clues for historical analysis, it cannot be taken itself as a guide to the historicity of a tradition. Just as great elaboration in detail need not guarantee historicity, so relative fragmentation in form need not in itself deny historicity. One might conclude, for example, that there is little of historical value in a tradition as fragmentary as that of the Baloho on Ijwi. But virtually the same structural elements are found in two apparently independent historical contexts (that of Kamo on Ijwi and Ntale at Mpinga), with no apparent interaction of the two lines of transmission. The geographical spread of these two versions and the regularity of their structural elements indicate that the prototypical form of the tradition is likely to be considerably older than either version discussed here. Because of this age factor, and given their present fragmentary forms, it seems likely that most extant variants on Ijwi are condensed forms of an earlier (and broader) narrative. Yet we have seen above that Kamo’s historicity is generally upheld by other evidence. The structure of the data alone, therefore, is not a reliable guide to the validity of its content.

The condensation of the Baloho tradition seems to have been more a result of the changing cultural context within which it is narrated than of the changing political fortunes of the Baloho (though these factors, too, may well have played a lesser role). The few items that remain in the narrative today emphasize the structural over the historical, thus reinforcing the distinctive Baloho cultural identification (which sets them slightly apart from Ijwi society today). The Babambo tradition, on the other hand, is structurally less unified; it is not so clearly a part of a single encompassing structural formula, which itself exists independently of the content in question, as the Baloho tradition appears to be. While the Baloho tradition in its present form seems to have been the result of compression, the Babambo tradition today appears to have resulted more from a process of elaboration, as seems apparent from the number of different and poorly integrated episodes—Kabwiika’s death, the Babambo attack on Kamerogosa, Gahindiro’s court, and Mwendanga’s arrival on Ijwi. The Baloho tradition avoids this type of additive, synthetic presentation. Mwendanga’s royal status, for example, is sim-ply inferred from the present situation (his descendants being of “royal” Basibula status); the tradition does not go out of its way to assert the fact.

The form of the Babambo tradition is determined more by the linking of successive discrete episodes than from a single overarching and consistent symbolic form that includes everything—as among the Baloho, where the sack represents, or refers to, flight (hence conflict), mobility, and protection (by the Baloho).

This additive aspect of the Babambo tradition may help explain its current spread, its more comprehensive content, and its detail relative to the Baloho tradition, for the process of synthesis that seems to characterize the Babambo structure al-lows it to be expressed in norms more in harmony with contemporary social norms. The Babambo tradition portrays events and political struggles in terms closer to more common narrative forms today on Ijwi and the mainland. It has also permitted the tradition to remain consistent with contemporary social structure and values.

It is not only the different structural flexibility found in the two tradi-tions that has favored the preservation of a more elaborate tradition among the Babambo. The recent development of Ijwi society has also done much to provide an environment in which one tradition could flourish while the other languished. One important factor of this changing history would have been the Rwandan occupation of Ijwi during the late nineteenth century and the simultaneous intensification of Rwandan central court control in Kinyaga, the mainland area south of Ijwi and east of the southern extension of Lake Kivu.  During the Rwandan occupation of Ijwi, important segments of the Ijwi population may well have adopted certain affinities with Rwandan culture, of which one element is a richly elaborated oral literature of many types.The form and length of the full Babambo tradition make it possible, even easy, to accentuate, embellish, or add selective elements to conform to different sociopolitical and historical contexts.

The intense population interaction with Kinyaga that occurred for some segments during this period, especially in the southern part of Ijwi (the area of greatest Babambo strength), undoubtedly contributed to a general Rwandan orientation of cultural evolution in this period just prior to colonial rule. In some respects this orientation may have extended into the colonial period itself, because “chiefs” were favored on the model of Rwandan chiefs; strong Rwandan cultural and political influences among the local Ijwi colonial elitewould also have reinforced Rwandan elements in the traditions during the early twentieth century. The colonially imposed chief on Ijwi, for example, had lived his youth in central Rwanda, and certain Belgian administrators (including the one with the greatest impact on Ijwi) had had previous experience in Rwanda and hence encouraged a Rwandan mode of “colonial comportment.” In general during this period, Rwanda was used as the administrative standard, if not norm, by colonial authorities, and the values subscribed to in this particular administrative context may well have filtered through to influence other social contexts. Consequently, there was a favorable climate to augment, elaborate, and extend the Babambo tradition; at the same time, such changes may have undermined the appeal and significance of the Baloho form of tradition.

The objectof this essay has been to present two traditions current on Ijwi Island relating to the arrival of the present-day dynasty, and to analyze these traditions for their historical contributions. These two traditions, held by independent possessor groups, are apparently mutually exclusive in content. While referring to roughly the same events and individuals, the content of these traditions appears otherwise nonintersecting and their forms are clearly distinct. However, it has been shown that their apparently independent historical data are complementary rather than contradictory. Viewed in time perspective, the traditions take on new light as portraying successive episodes within a larger historical transformation, even while they also reflect the differing perspectives of two social groups each tied to royalty through very different historical processes.

The distinct forms of the two traditions have been examined by reference to a tradition parallel to one of those found on Ijwi, and associated with another segment of the same royal dynastic complex. While diverging historically from the Ijwi royal family, this segment is found today at Mpinga, on the mainland west of Lake Kivu, in an area with very few contacts to Ijwi Island; from the time of the dynastic split there have been no continuing contacts of the mainland group with the Baloho, the social group that preserves the tradition on Ijwi. Comparison of the Ijwi and mainland versions has shown that a similar stereotype has been employed both on Ijwi and at Mpinga. From the histories of the two areas, it has been determined that these two traditions relate to events associated with the Sibula interaction with Tembo peoples, peoples who today live west of the Mitumba Mountains but whose culture formerly was probably more wide-spread in the area of Lake Kivu.

In addition, some attention has been given to the possible evolutions of each of the Ijwi traditions in light of the subsequent known historical events. From this, it was shown to be probable that the eclectic structure of the Babambo tradition helped it to adapt to various regions. But the more important reason for its spread—and indeed for its very eclectic nature—may have been the expansion of Rwandan cultural norms in the area from the late nineteenth century and continuing, perhaps with increasing intensity, into the colonial period. Despite the political influence of individual Baloho on Ijwi during most of theperiods, the Baloho genesis tradition apparently was confined to a few Baloho informants by its form and content.

Finally, the problem of the localization and specificity of the traditions can be seen in perspective. Their regional localization and their very clear separation from each other despite their common focus on the same person reemphasize their character as documents rooted in their specific social group by their very form. They are clearly part of the social group that holds them, not only because of this content but because of the way they are told—by their relation to a specific cultural area.

At the same time, they can be seen to belong to a given clan by their content. This accounts for the specificity of these traditions: their exclusive nature, which ties them to one clan; they document the unique claim of one clan to the royal establishment. The dramatic cultural differences in the two narrative forms prevent the two traditions from contaminating each other. Because each tradition refers to one clan only, and each details a specific interaction with the kings, they do not spread more widely through the region. In itself, this limited extent of such traditions testifies to the imperfect political integration of royalty on Ijwi and to the continuing strength of particular regional and social identities on Ijwi.

Yet even these differences help to prove the unity of the island. Both traditions, so different and so separate, refer to the same dynasty, the same king, and roughly the same episode surrounding his arrival. There is no division into different dynasties or disagreement on events. As noted above, these accounts are completely complementary. In the narratives, as in social relations, the kingdom provides a single focus transcending the cultural differences of the clans: king-ship brings them together without necessarily diminishing their status. To the contrary, as in the traditions, kingship on Ijwi can be said to have augmented the role of clan identities, formerly restricted and regional. Not every clan shares the traditions explicitly, but no clan or region stands outside the meaning of king-ship. Even the Babambo, after their expulsion, and the Baloho, in their relative isolation of the east, retain this focus.

Genesis traditions, especially those for which there are several variants of a single tradition, are increasingly suspect on several grounds. From the point of view of the historian, the conclusion of many of these analyses is pessimistic. In some cases historians relate the widespread variants uniquely to generalcosmological values in the area, interpreting these traditions as magnificently elaborate statements on belief systems and values. Alternatively, they tend to view the tradition as widespread core clichés to which all manner of extraneous data can be (and are) affixed, thereby making it impossible to decipher the historical elements from the structural, and complicating, if not negating, the historian’s efforts.

This essay has tried to show how some of these problems can be approached by the analysis of two traditions found within the same local context and per-taining to the same series of events. The approach consists of several strategies. By carefully examining the similarities and differences of the traditions, one hopes to distinguish the convergent episodes, and to isolate conflicting elements. By analyzing their distribution and the historical role of the possessor clan, one can arrive at a fuller understanding of the form and how the narratives may have been altered. By examining the traditions in relation to the larger corpus of oral literature—not simply historical traditions—one hopes to identify those elements that may have been incorporated within the traditions and presented as history. By looking at the presence of certain clichés and sociological patterns, one hopes to arrive at a better understanding of metaphorical elements within the traditions. By looking at the wider geographical spread of similar narrative structures to those in the traditions, one hopes to gain insight into the manner, and perhaps also the processes, of how such general themes became embedded in particular social contexts. Finally, by looking at the historical evolution of the society within which these traditions function (and an evolution that the traditions themselves may influence), one can arrive at a better understanding of the specific historical referents contained within these narratives.

Much more, of course, is possible when multiple sources are accessible, whether these be written, ethnographic, or archaeological. Material items, residence patterns, rituals, linguistic analysis, and the study of the general literary culture in the broader area, as well as within the specific society, all lend potential value to the historian’s search for the how’s and the why’s of the past that have led to the present. But for many areas of Africa, there is little possibility of multiple sources converging on a single problem, without the geographical field of inquiry being so expanded as to alter the nature of the historical phenomenon under study. For more limited areas, with more precise historical problems to be considered, one is all too often left with only one or two sources from which to draw the data.

The principal sources employed here are the oral traditions themselves. I have attempted to determine what, if any, historical value can be derived from a careful analysis of the traditions. Because independent sources are few in this re-gion, historical analysis is largely limited at present to oral sources. Therefore, my guiding approach here has been to treat oral narratives as functionally part of their society and culture and influenced by changes in these domains; but at the same time, oral traditions need not be considered exclusively in this light— in short, the cultural context is not determinant. Although these contemporary functions are real and important, the traditions are not perfectly integrated to the society: even as they are products of an evolving social milieu, they also bear witness to that evolution. In both senses, they are historical traditions.

One analytical aspect that becomes important in such an analysis is the interdependence of all historical sources and aspects of cultural change. So close is this that historians are faced with the dilemma that in order to understand how oral sources change and function, they need to understand how the society altered and how social perceptions have changed. But in order to understand societal change in an area such as this, one is dependent primarily on the traditions themselves. This dilemma is partially circumvented by employing many different —and independent—types of data. While these may not always provide precise historical knowledge, they can often help define the limits of what can be safely asserted.

In this perspective, what is of crucial importance in an analysis such as this is to explain the differences in the various traditions, rather than to seek the similarities among them. For the fact is that an episode or event is never perceived in exactly the same way by different groups or individuals, either initially or (especially) through the perspective of time. That there be differences, then, need not in itself be surprising, or reason to cast out all as invalid. Nor is it enough simply to accept the common core of variants, because this is potentially most clearly the result of structural influences.

Instead, the initial observations of similarities and differences need be viewed as the point of departure for the application of the historian’s craft: both similarities and differences need be explained. To do this may require reading into a variety of domains sometimes ignored—often because the data are insufficient or totally lacking—by one school of analysis or another. It is essential to know how society changed; it is essential to know how the individual traditions changed; it is essential to explain the differences in the different interpretations, if we are to understand their historical meaning. In other words, historians are engaged in—as are the sources themselves—a continual and multidirectional dialogue. All this is evident and often recognized; but it is less often practiced.

What appears in the preceding sections is only an initial attempt to come to grips with the problem, but I have tried to show one way that this kind of inquiry might be undertaken, in this particular area for these particular traditions. Unfortunately, the number of variables is so great, as is the potential importance of any one variable in understanding a tradition and its historical meaning, that at this stage it would seem impossible to draw up any guidelines of analysis but the most general canons. Until more is known of social change in many parts of Africa, and especially until we understand better how oral literature of all types functions in a changing society, there are no specific methodological rules to follow—there are only standards by which to evaluate whatever techniques are applied. The most that can be done, from the point of view of historians, is to keep in mind the complexity of the materials they are handling, as well as the delicacy and finesse of these invaluable and intriguing sources. of kingsSocial & cultureThe first section discussed the differences in the forms of the two genesis traditions. The second discussed differences in the areas of residence, the histories, and the contacts with others sustained by the Baloho and Babambo, and speculated on the influence these factors may have had on the different...AMATEKA