The Corpus
This volume presents a bilingual edition of one performance of the so-called baja ni, a string of funeral songs that the Dogon ascribe to a poet called Abirè Goro who probably lived in the nineteenth century. The term baja as such means ‘problem’, but in combination with ni (song) it implies singing about the loss of life, so funeral songs. This volume complements and completes our earlier ethnographic and literary studies of this singing event that forms an integral part of Dogon funeral proceedings.1 In combining a full vernacular text and translation with a detailed historical and linguistic analysis of the songs, we aim to reveal the crucial cultural and historical dimensions of this baja ni singing tradition. In particular, this publication aims to enhance our understanding of Dogon religion and ethnic philosophy, for which this corpus of songs is an essential treasure trove.
As we shall argue, the texts are not meant to be a coherent tale about the past, but their historical relevance is of a different character, which is expressed in several ways. First, each song comes with its background history, orally transmitted information that is separate from the lyrics, so the songs contain a series of related mini-histories. The whole cycle and its context thus offer a window on the Dogon living in a specific temporal and geographical setting, zooming in on their reactions to crucial events in their lifetime. Second, since the corpus has a historically situated author, the text reflects the way individual Dogon experienced their subservience to the Fulbe,2 who at that time were the political rulers. Thus, many of the songs relate to the inherently unstable relation between horticulturalists and pastoralists, two ecotypes virtually destined to live together in the same area throughout the Sudan-Sahel zone; in this case both dwell the plains of what is habitually called ‘Pays Dogon’.3 The songs reflect on this precarious balance of power, which in the era of Abirè Goro had found a poignant political form in Fulbe dominance with its concomitant slave raiding. Within this framework the texts eternalise names of persons, either for their historical political importance, or as local cultural heroes who embodied admirable
What we present here are the texts of all songs that filled an entire night during one traditional Dogon funeral. There are more baja ni songs than can be sung in one night, and even the seven hours allotted for such a session are not nearly enough to perform all known pieces. Therefore in each rendition of the baja ni the singers have to make a selection. A core corpus of songs has to be performed always and forms the highlight of the singing ritual, usually somewhere halfway the night. For the remainder of the performance the singers can select from the universe of the baja ni, though we will encounter some rules pertaining to these choices; the singing night has its own rationale. So this volume contains one such selection, in effect all songs performed during the night of 8 to 9 June 2005 in Tireli, a village at the foot of the Bandiagara Cliff in Mali.
The number of songs in this book, 88 in total, is larger than usually performed at a funeral, for two reasons. First, this occasion was a baja lugo, a ‘counted mourning’, implying that the ritual farewell of the funeral had already been done more than a month ago, and that this baja ni was organised well after the main festivities. When performed as part of the funeral proceedings, the singing time is limited to seven hours, as another type of funeral song must be sung first. In this case the deceased had been a singer, one of their ‘colleagues’, so they felt they needed more time to sing in his honour. The second reason for including a larger number of songs was the presence of the first author, Van Beek. As will be explained in the text, he and his host Dogolu Saye decided to have a small pre-session of singing in van Beek’s courtyard, in order to enlarge the catchment area of recordings of songs.
How large would this universe of baja ni songs be? The songs of the 2005 session are the first ever to be printed, so we cannot compare with earlier publications. But this session was by no means the first one recorded by the field team, consisting of the first and third author, with incidental assistants. The
However, in oral traditions any number comes with a lot of provisos. A major question is what constitutes one song. Each text has its variations, and exactly when a text crosses over into another song is somewhat arbitrary; as we shall see the dividing line mainly depends on the chorus, but these lines too can shift. Second, some songs are very short, just a few lines, and are linked to others, and in our 2016 sessions disputes arose over whether these would count as one song, or just as part of its predecessor. Third, any lead singer can insert large improvisations, commenting on events during the performance, on a bystander or on the ululating women, or he can recite parts of folktales as he sees fit; though the chorus may be recognizable as the core of the song, these diversions make it hard to identify the song as such. In effect, during such a large detour the lead singer sometimes gets off track and shifts from one song to another, mixing two well-known texts into a hybrid new one. We shall see this with some of the songs in this volume, like B – g. In folktales a similar process has been noted, where a narrator starts out with one tale, and ends with another.
A more fundamental problem in counting the songs is that the baja ni cannot be considered a closed corpus. As we shall see, the songs attributed to Abirè Goro are embedded inside a corpus of pre-existing songs, and as in any oral tradition new items are being added, novel themes addressed, and varying issues commented upon, a process the singers are quite aware of. Old songs will probably have been lost, but since this volume contains the first published baja ni texts, that dynamic is not yet visible. Not only are new songs added, but for the singers each rendering of an old one is new in itself. For example, during the 2016 analytic session the research team asked for a piece to be sung again, and immediately the singers stated explicitly that it would be different from any other rendition. They consider the performance of each song as a new creation.
All Dogon texts in this book are in Jamsay, the Dogon language of the plains. The orthography of the Dogon languages has been standardised by AMALAN,4 the Malian alphabetisation program, and we will follow this. No diacritics are used in this orthography, so even if all Dogon languages are tonal, tones are not indicated.
This orthography features four symbols from the International Phonetic Alphabet:
ɛ – the è in French, ‘men’ in English
ɔ – like ‘offer’ in English
ŋ – like long in English
ɲ – palatalised n, like mañana5 in Spanish
There is no schwa in Dogon, so the ‘e’ in Dogon terms and names pronounces as the vowel in the English ‘bay’. In this AMALAN orthography the nasalisation of vowels is expressed by adding an ‘n’, as in French.6 The nasalised ‘e’ and ‘ɛ’ converge as ‘en’, and the nasalised ‘o’ and ‘ɔ’ as ‘on’.
Personal names always demand separate attention, also for non-linguistic reasons. The Dogon singers and informants insisted on having their names spelled out in writing; they had given their wholehearted support to the project and are proud of their cultural heritage. Using pseudonyms was out of the question, and anyway for them privacy was for those who have something to hide. Especially in photo captions they wanted to have their names mentioned correctly and in full.
Also the text contains many Dogon names, of both persons and villages. As for their spelling, in the Dogon texts we follow the Dogon orthography. When performing the singers often take some liberties with personal names, adding vowels for poetic effect; we note this in the transcription but ignore it in the translation. In the English translation, whenever possible,
Sung sentences flow easily into each other, so we had to provide interpunction. This was done in two phases: the first was during the transcription and translation into French, most of which was done by the third author with a small group of singers present, during the 2016 analytical sessions. Preparing the text for English publication required a second round of applying interpunction, this time by the first two co-authors, giving special attention to the interpunction of the Jamsay text, a language for which both the rules for orthography and interpunction are still developing; the latter do not always run similar to English conventions.
Van Beek 2008; Van Beek 2012b; Van Beek, Saye and Ongoiba 2020.
Fulbe is plural; the singular is Pulo. We also use ‘Fulbe’ as an adjective.
This quite common indication tends to ignore the fact that the Fulbe belong there as well. The name is of colonial origin and has been reinforced more recently by tourism, at a time when the balance of power has been swinging in the direction of the Dogon. But with the recent jihadist insurgencies the scales drop into the other direction again.
Académie Malienne des Langues, the successor of DNAFLA, Direction Nationale de l’Alphabétisation Fonctionelle et de la Linguistique Appliquée du Mali. This government agency suggested labelling Toro so as the standard Dogon language, but that would be impossible in this text. Moreover, Jamsay is very much a living language (Ongoiba 1988; Heath 2008) and there seems little reason to suppress this.
We thank Jeffrey Heath for this suggestion.
Together with the absence of diacritics for tones, this sometimes generates a rather awkward spelling, such as when a nasalised vowel is followed by an ‘n’ (e.g. atunwunnu, spirits) or any kind of double nasalisation (e.g. unrun – instead of another possible spelling such as ûnû). For instance, in the name Yannhire, the first ‘n’ nasalises the ‘a’; the second is a real ‘n’.
All supplementary material to this volume has been made available online at figshare.com. They may be accessed via this QR code and the following link: https://doi.org/10.6084/m9.figshare.27377406.