the performance, aided by a technique of delivery that heightens its artistic effectiveness. There are also the cases when the performer introduces variations on older pieces or even totally new forms in terms of the detailed wording, the structure, or the content.
The extent of this kind of innovation, of course, varies with both genre and individual performer, and the question of original composition is a difficult one. It is clear that the process is by no means the same in all non-literate cultures or all types of oral literature, 4 and between the extremes of totally new creation and memorized reproduction of set pieces there is scope for many different theories and practices of composition. There are, for instance, the long-considered and rehearsed compositions of Chopi singers, the more facile improvisation of a leader in a boat- or dance-song, the combination and recombination of known motifs into a single unique performance among Limba story-tellers. There are also occasional cases, like the Rwanda poet just mentioned, where there is interest in the accuracy and authenticity of the wording (at least in outline) and where memorization rather than creation is the expected role of the performer.
In spite of the very real significance of these variations, one of the striking characteristics of oral as distinct from written literature is its verbal variability. What might be called the ‘same’ poem or prose piece tends to be variable to such an extent that one has to take some account at least of the original contribution of the artist who is actualizing it—and not simply in terms of the technique of delivery. Take for instance the case of Ankole praise poems. Since the ideas expressed in these poems are stereotypedand repetitive, the omwevugi (poet/reciter) must change the wording to obtain variety.
[He] has to rely to a great extent upon the manner in which he expresses these ideas in order to give beauty and interest to his poem. Herein lies the art of the accomplished omwevugi who, by the ingenious choice of his vocabulary, can repeat identical themes time and time again, always with a different and startling turn of phrase.
(Morris 1964: 25)
Again, there is the production of stories among the Thonga. It is worth quoting Junod’s excellent description of this at some length. Having postulated the ‘antiquity’ of Thonga tales, he goes on:
This antiquity is only relative: that is to say they are constantly transformed by the narrators and their transformations go much further than is generally supposed, further even than the Natives themselves are aware of. After having heard the same stories told by different story-tellers, I must confess that I never met with exactly the same version. First of all words differ. Each narrator has his own style, speaks freely and does not feel in any way bound by the expressions used by the person who taught him the tale. It would be a great error to think that, writing a story at the dictation of a Native, we possess the recognized standard form of the tale. There is no standard at all! …
The same can be said with regard to the sequence of the episodes; although these often form definite cycles, it is rare to hear two narrators follow exactly the same order. They arrange their material as they like, sometimes in a very awkward way …
I go further: New elements are also introduced, owing to the tendency of Native story-tellers always to apply circumstances of their environment to the narration. This is one of the charms of Native tales. They are living, viz., they are not told as if they were past and remote events, in an abstract pattern, but considered as happening amongst the hearers themselves …. So all the new objects brought by civilisation are, without the slightest difficulty, made use of by the narrator ….
Lastly, my experience leads me to think that, in certain cases, the contents of the stories themselves are changed by oral transmission, this giving