It is self-evident that treatment and critique of the oral sources is of crucial importance to the construction of any history of the Nyiginya kingdom and it is fitting to say something more about them. The oral traditions can clearly be divided into a set of official sources transmitted and controlled by the court and a set of popular sources. The first include a very large number of traditions, rendered in a much wider fan of genres than has been common in other African kingdoms. Merely to draw attention to this fact is already to underline how exceptional the role of history has been in the ideology of Nyiginya royalty.

Samples of most of the literary genres in which these official sources were composed have been published and certain genres have benefited from general critical analysis. Yet useful as it is, such a critique of the different types of sources tends to obscure a fundamental truth, to wit, that all these sources belong to but a single reservoir of information that circulated throughout a single social network at the court. Hence in-formation stemming from one genre of sources, especially those of a narrative nature, was easily used to enrich or even to construct a tradition in any other genre. Thus one cannot really privilege any literary genre above any other. Kagame did privilege three categories of sources because they were learned by rote and because great care was taken with their transmission owing to the exceptional ideological or ritual importance of their content. He believed that their uttering had not been altered by a single sound since their composition, which virtually turned such sources into written texts and rendered them more credible than others, in particular more credible than ordinary historical tales whose precise rendering varied with each performance. He did accept that the date to be assigned to the testimony of ordinary traditions must be the date of their collection, always after 1900, but claimed that the text itself of his privileged categories dated to the time of their composition so that certain texts would date to the seventeenth century. The three genres to which he accorded privileged status are the liturgy of the rites of kingship (utterances and practices), dynastic poems, and the list of the dynastic succession.

It has been demonstrated, however, that even within those three genres certain utterances have been altered by interpolation, deletion, or substitution. The texts recorded on paper do therefore not date to the moment of their first composition. Moreover, both the dynastic poems and the liturgy are allusive. They do not include a straightforward setting out of historical events and their allusions cannot be understood without glosses. Since those glosses or narrative commentaries are not learned by rote they should in any case be handled as if they were merely ordinary historical tales. At the most one might argue that an allusion in an old poem “proves” what one also finds in an historical tale but even that is not safe. For the tale could well only be the narrative development of a speculative gloss aiming to account for an allusion that could not be understood. Moreover, we should never forget that the performances of these three genres of sources were always closely and directly monitored by the court because they were crucial to the legitimacy of both king and political regime and also served to justify political actions taken by the court. This role evidently did favor their being remodeled according to the taste of the moment for instance after a coup d’état or their being used to justify a political innovation in search of precedents. In the final analysis there is no reason to privilege any genre over any other since they all draw from the same reservoir of information. At the most one might accept that in certain cases where there is no suspicion of a later interpolation, historical allusions made in certain dynastic poems do refer to the narrative historical traditions about the king to whom the poem was addressed. These traditions could then safely be assumed to have existed during the lifetime of that monarch.

By 1907 at the latest Europeans began to record some popular traditions. But the specialists at the court, the Abacurabwenge, still refused to give any information. Ten years later that stance had changed, for in 1917 de Briey obtained the first official list of dynastic succession dictated by the “Bachurabwenga.” Then Pagès mentioned these “official historians” in 1927, but unfortunately he did not name the one who had been his informant. Soon a single set (and always the same set) of Rwandans began to serve as major collaborators—the expression is Schumacher’s—and transmitted their version of Nyiginya history to Schumacher, Delmas, and the young A. Kagame. Thanks to Schumacher’s synthesis we know what the state of knowledge was in 1936, the year he left the country. As it happened Kagame began his research in the same year. At the outset he obtained a general overview of the official traditions from Sekarama, official performer and dynastic poet, who had been one of Schumacher’s collaborators. But soon the number of his interlocutors grew so as to progressively include nearly all the competent officials at the court, whose data he used to further enrich the already established framework. Kagame published his first synthesis in 1943 and again in 1947.

Because Schumacher’s synthesis had been written in German and remained unpublished it found no echo, whereas Kagame’s, written in the Rwandan language and published locally, was avidly read all over the country. It became the “true” version for the leaders of Rwanda, for its official historians, and for a good many popular storytellers as well. His second synthesis, the Abrégés, essentially maintained both the frame-work and the overall interpretation from his earlier Inganji Karinga, but was enriched by a great mass of details.

The gist of the single reservoir of traditions was thus written down between 1910 and 1936 and stemmed from a handful of specialists at the court. Schumacher in particular based his work on the testimony of four main collaborators, two of whom also appear in Delmas and three of whom appear in Kagame. The crucial persons among these were Rwanyange, the official keeper of the official list of dynastic succession, Sezibera, chief of the guild of dynastic drummers, Kayijuka, a senior chief and ritualist, and Sekarama, official performer and composer of dynastic poetry as well as a great expert of the traditional tales. In the last analysis, then, these four people, all trained at the court of Rwabugiri, have imposed their views on the whole historiography.

Yet one must still account for the differences found in the traditions as presented by various authors. According to Pagès, for instance, the King Ruganzu Bwimba campaigned against Burundi or Bugesera, not Gisaka as later authors claim, and Ruganzu Ndori raided Karagwe, an unimaginable situation for Kagame who claims that a pact of eternal peace existed between Rwanda and Karagwe. Researchers have usually accounted for such differences by claiming that the first historiographers were not very knowledgeable and often misunderstood what they were told or alternatively obtained only bits and pieces from incompetent persons. Therefore, the reasoning has gone, one should rather rely on more recent information and especially on what Kagame tells us, since he himself was both a Rwandan and a ritualist. Although this explanation is clearly valid with regard to the very early and hasty collections— especially with respect to the official list of dynastic succession—it does not hold for the later historiography beginning with Pagès. Actually Schumacher’s synthesis is particularly valuable because it gives us a detailed state of the question on the eve of Kagame’s research. A comparison of his work with the syntheses of Kagame allows one to see how much the latter owes to the framework of Sekarama and his colleagues as well as to verify a series of detailed assertions made by Kagame, but by no one else, and to place these in the context of the whole traditional body of evidence.

Bearing in mind that there was only a single official reservoir of information at the court, irrespective of literary genre, that the information given by a few historians, who were the recognized official experts at the court of Musinga, was asserted to be authoritative, and that these experts knew each other, as did the historiographers they informed, it becomes clear that one must have recourse to another dynamic to account for the divergence of information among authors of different dates. Yes, the historiographers drew their knowledge from conversations and interviews with the historiologists (oral historians) who were their collaborators. But little by little the latter discovered that different versions existed from the one they knew and they saw the need to account for the existence of these different versions. For any divergence among different versions only becomes glaringly apparent when a text is written down, an action that freezes it and allows for detailed comparison with other versions. Therefore the foremost historiologists consulted with each other and used their authority to impose one version and eliminate the others. They sometimes did so by invoking still secret esoteric commentaries and explanations (intekerezo). Secondly the historiologists took note of the questions and the objections lodged by the historiographers during their meetings and invented answers to these that were then incorporated into later performances. Finally the historiologists also adopted new elements introduced by the historiographers, who had found them in writings about neighboring countries.

Thus the history of the kingdom took shape step by step as the result of the dialogue between historiographers and historiologists. The clearest example of this process is the harmonization of the chronologies of Burundi and Rwanda. Schumacher and his collaborators went to visit the literate chief Pierre Baranyanka to exchange data and to set up a common chronology. Later on Kagame, who was not happy with the results of this encounter (which Sekarama had told him about), renewed the dialogue with Baranyanka until he reached satisfactory results, which later on became the received truth.

To sum up: the active official traditions in Rwanda began to be standardized in 1917, thanks to a collaboration between historiologists and historiographers, and by around 1936 a permanent and definitive version of Rwandan history had been arrived at. Hence the divergent versions of yore were not the product of misunderstood information or even incompetence, but represent variants that were accepted before standardization, at a time when for certain Rwandans, Ruganzu Bwimba had indeed campaigned against Bugesera and Ruganzu Ndori against Karagwe.

Faced with these now-standardized traditions, what can a historian do to recover the content of the traditions before 1920? A very severe critic will conclude that they cannot be recovered since most of them were not written down, that we have no idea as to how variable or how credible they were, and that it is therefore futile to attempt to write a history of the Nyiginya kingdom. That is too much! After all, the standardized version is not sheer invention. It is based on older versions. Should we not then after all privilege the dynastic poems and the liturgy of the royal rituals since they were learned by rote? For even if they were not totally invariable and even if they were subject to interpolations and deletions, the bulk of their content as written down certainly is mainly what it was by the end of the nineteenth century. Moreover, one readily finds numerous passages that were altered as the result of the dramatic events of 1895–96. But still, one must keep in mind that most of the historical information of the poems and liturgy stems from explanatory oral glosses, glosses that drew on the general reservoir of oral information, that is, when they were not a product of speculation pure and simple. Hence they too did probably not escape the overall standardization.

Actually, the historical narratives remain the keystone of any reconstruction and to establish their value one must find their variants. To do so one can, on the one hand, compare the versions of Pagès and Schumacher with the canonical version established by Kagame, and, on the other, compare Kagame with the more popular versions gathered after 1943, even those of storytellers who were somewhat influenced by Kagame. This has turned out to be a fruitful approach. In most cases one finds a variant in each narrative, and the magnitude of the difference from the other versions gives a good idea of the relative credence one can accord to any given narrative. On the other hand, the total absence of variants does not inspire confidence. In those cases one can only arrive at an opinion about reliability by both focusing on internal criteria and by observing how well the tradition fits into the context of all other known data.

How did these oral traditions envisage the past? For them the history of the kingdom is that of its kings. They begin by telling us how the first King Kigwa (“fallen”) descended from heaven and organized the aristocracy. They continue by giving a series of names of successors that stops at Gihanga (“creator”), the cultural hero who founded the first empire, which he left to his children, one of whom was Kanyarwanda, an eponym for the country. Then follows a new series of royal names until RUGANZU Bwimba is reached. He is the first monarch about whom historical tales are told. From then on the list of dynastic succession constitutes the framework around which the whole body of oral traditions has been organized.