Teaching Oral Tradition: What Type of Professional Training for the Mande Bards?

This article deals with education to oral tradition in West Africa. Its author analyses the training of Mande bards or griots in a small village – kela (in Mali) – through a didactic point of view. The general organisation of that village reveals two types of training which are complementary. The first one is embedded to everyday life, it provides what the author refers to as a type of socio-general training which covers some basic cultural topics emerging from social interaction with elders. The second one is clearly distinct from everyday life situations. Unlike the first one, it is planned and it provides a type of professional training which is disciplinary (e.g. music, public oral expression). As concerns the analysis of this second training which focuses on the teaching of pane-gyrics, the author describes the specific didactic gestures displayed by the local teacher to make trainees memorize the praises they are dealing with. Further analyses show that this is just the first step of an oral process where the object of teaching is first shared (bet-ween teacher and learners), made to be mastered (by learners) before launching, upon the object, some interaction and regulation aimed at strengthening that specific teaching by explaining some meaning, and showing how to melt praises into narrative sequences or how to handle a recitation of praises in real professional contexts.

In its instituted form, the action of education would mainly refer to any planned and organised communication that aims at developing learning through the transmission of information, cognition, knowledge, know-how whose efficiency is all the more important that it matches the education permanently transmitted by the social body / group (Escot, 1999, p. 26). This system of communication is mostly characterized by a type of organisation that only a school institution embodies best; as a result of the management of a state apparatus that regulates everything (Hofstetter, 1998, p. 93).
As opposed to other educative systems -such as informal and non formal education (Maulini & Montandon, 2005) -, the school institution is usually thought to be formal because everything within this system is carefully detailed: It defines in a much precise way the conditions its actors must fulfil, be it the teachers (in terms of training, career) or the students (in terms of competence). Participation here is compulsory. The organisation is made in a way that all the aspects of the system are specified and the control of the process insured (Escot, 1999, p. 47). The state management of the school institution provides a given credibility to this harmonized system based on a rigorously planned and scheduled organisation on a long-term basis. Everything is achieved through a kind of scriptural memorisation (for instance, state teaching guidelines, curricula and textbooks) that is made necessary by the complex character of knowledge itself; as well as the need to organize this knowledge in teaching contents, which is part of the didactic transposition process (Chevallard, 1985(Chevallard, /1991.
However, what do we know about the other above-mentioned educative systems? How do they manage to achieve education? Is there actually anything which is ever done or instituted to achieve that goal?
In an attempt to answer these questions, we shall focus on the training of griots (or traditional bards) which involves another form of education, stemming from a different context -the oral tradition -usually considered as informal education.
To do so, in the following lines, we must first define our theoretical and empirical context, so as to provide our readers with the background elements that sustain our analysis. Secondly, we shall describe the griots' craft, and also try to identify their training principles according to what they and their oral tradition say. In the third part, we shall provide an overview of the various types of training pertaining to the oral context that can be found in the village of Kela where our data has been collected. Finally the fourth part then focuses on the professional training, before trying to draw the didactic configuration of the whole setting of Kela.
Drawing a methodology and Seizing the Griots' Training: What Type of Analysis? Theoretical background From what has been mentioned above, it is relatively easy to trace the organization of school education, because it is planned, explained and carefully documented by means of written documents. But it seems very complicated when it comes to the griots traditional context, where things are done orally. As with most of the cases of learning outside of schools (Greenfield & Lave, 1982), the training of bards has always been considered to be informal because it seems to be carried out inside the family without any form of organisation. We consider this to be a popular belief that needs to be questioned through an adequate methodology. In fact, the category of informal education (Brougère, 2007;Dasen, 2004;Poizat, 2003) is usually the result of everyday cognition sustained by an «informal pedagogy, although it often remains unstated and even unconscious by those who practice it» (Dasen, in press). Chamoux (1983, p. 40) distinguishes two types of learning inside the above category. The first one is considered to happen by immersion; whereas the second one involves a clear transmission by a teacher. This author suggests that if the sole observation of a technique can lead to imitation, then no teaching is necessary; the only thing happening here is cultural immersion. In the case of a much more subtle knowledge, the help of an expert or a teacher would be essential for transmission. In a much more general perspective, Dasen (2004) considers immersion to be part of enculturation; and transmission by a teacher to be part of socialization (with active inculcation).
As far as the training of griots is concerned, many researchers (Camara, 1996;Hale, 1998;Johnson, 1986) assert that these bards follow a special training to become professional. However, Okpewho (1992) and Jansen (2000) highlight the fact that only a few researchers have provided a detailed analysis of this oral tradition system. While some suggest that these griots learn by immersion and imitation (Jansen, 2000, p. 23); others -like Camara (1996, p. 767) -mention that the only way to learn major oral texts is to go through a type of training where systematized teaching is organized under an institutional form like in Kela (a village of traditional bards in Mali).
What lacks in these researches is a case study based on a much more precise theoretical framework based on consolidated concepts meant for the study of the knowledge transmission process. Studying this transmission implies that the object of teaching, involved in the process, should be clearly identified and specified by the researcher in order to avoid hasty generalizations. As far as the griots are concerned, their profession consists in perpetuating oral traditions. It should be noted that this general goal is achieved through one of the multiple traditional functions in which they choose to specialize. For instance, if a griot chooses to specialize in praise singing or public speaking, the object of his/her 1 training would consist in referring to one of their epics that contains the panegyrics that are needed for the sake of teaching. (Toulou, 2005).
Our contribution aims to provide the analysis of such training (between an expert and an apprentice interacting on panegyrics) using certain concepts of the didactic transposition theory.
The didactic transposition theory studies how the contents that are to be taught are effectively reconfigured and adapted by mean of a didactic triangle involving teacher -learners -knowledge (Chevallard, 1985(Chevallard, /1991. During teaching, there is an interaction between these three poles. The human poles of this triangle are tied by a didactic contract (Brousseau, 1988) that defines the mutual expectations between the people involved in the teaching/learning process.
As far as the teacher's didactic action is concerned, some researchers (Sensevy, Mercier & Schubauer-Leoni, 2000;Toulou & Dolz, 2005) identified a few fundamental professional gestures that can help to analyse given teaching processes. Among others, we have the following gestures: definition (posing the objects and establishing the framework of the situation) -regulation (reformulating, designating, interacting, reframing or resetting in order to obtain winning strategies from students) -devolution (allowing the students take their responsibility and work on their own) -and institutionalization (validating some answers on behalf of the institution). Apart from these fundamental gestures that are general, there are some specific gestures (Bucheton & Dezutter, 2008;Toulou, 2008) that are closely related or dependent on the object of teaching. These specific gestures would therefore emerge from teaching interaction.
In general, the didactic transposition theory would only apply for the study of school phenomena; but as Mercier, Schubauer-Leoni & Sensevy (2002, p. 13) suggest, by adopting an anthropological point of view, the domain of didactic transposition theory can be enlarged to the study of all the cultural objects which display some observable didactic relationship between people. Therefore, the main focus of our analysis would be directed on one of the didactic transposition ramifications, the Anthropological Theory of the Didactic (ATD) that has been introduced and developed by Chevallard (1991). This theory stresses very important aspects in rejecting some generally established views of teaching and learning. It tries to establish new attitudes towards 'the didactic', seen as an anthropological dimension of social life. As Chevallard (2007) suggests: In widening the scope of didactics, as I suggest to do, we also have to generalise its object: didactics not only cares for the knowledge recognised as such by some authoritative institutions -e.g. the institutions of higher learning -, but it has to broaden its object of study, just because in the life of institutions, bodies of knowledge appear intricately linked, from the point of view of ecological analysis, with entities that some authorities would refuse to call knowledge, although we need to take them into account in order to explain the fate of «true» knowledge. (p. 23) In the ATD, any «body of knowledge» is simply a praxeology (or a complex of praxeologies) which has gained epistemic recognition from some culturally do-minant institutions. These latter do not all function in the same way. A school institution for instance would conceive and plan things differently from a small traditional institution. Thus, by borrowing some school theory to describe what goes on in an oral tradition context, there is possibly a risk of betraying the essence of the local action and discourse by a de-contextualisation. Therefore, how can one account for an oral tradition transmission process without betraying this ongoing culture by some borrowed ethnocentric theoretical models (Dasen, 1991, p. 221)? We suggest controlling this aspect by using an emic approach (Jahoda, 1983).
In this type of approach, Jahoda (1983) recommends a continual referral to the corresponding concepts of the symbolic system of any given context in order to obtain a better appraisal of the object being studied. In doing so, one avoids falling into the trap of stereotypes imported from so-called elitist cultures. Therefore, this paper will be developed from an internal perspective that gives priority to the cultural values conveyed by the society or group of people analysed, so as to provide an insight of what is carried out in that specific context. As Jahoda (1992) puts it: «any behaviour is undeniably cultural. And it therefore appears useless to try to decontextualize it» (p. xi). Likewise, we postulate that, by taking the cultural context of an institution (like the one sustaining oral tradition) into account and by adopting the local viewpoint, one can, not only gain a better understanding of the oral tradition issues, but also ideally analyse the praxeology that supports griots training in their specific context.

Data collection
In order to achieve our aim, we went to Kela, a small village of griots situated approximately 100 km away from south Bamako (Mali). For the data collection, we were concerned about two things: first watching attentively the young griots' timetable on daily bases, so as to elucidate what we would call the possible moments of education in general (i.e. providing a wide perspective of training); second, focusing much specifically on the teaching of oral skills between a trainer and its apprentice (so as to get an insight of a specialized training). One of the recommendations we gave to Kela griots was to continue to act as usual in spite of the presence of the researcher. Our research plan had also planned two groups of interviews: the first one before teaching (carried out with the elders in charge of training and based on the general philosophy of oral tradition transmission in their village, as well as some topics dealing with what we had read in books, or observed during our wide perspective survey); the second group of interviews would take place after teaching (carried out with the trainer, so as to ask questions about some aspects of teaching that required more details).
Three complementary means have been used to collect our data. The first established a written report of the observation of young griots in their interaction with their peers, parents and mostly with the elders of the village in charge of education. The second means concerned audiotaping of the interviews that we conducted in the main vestibule of the head of griots' compound. The third one concerned videotaping of the teaching sequences; the recommendation here was to let the griots indicate themselves to the researcher what they had planned and eventually what and when he could videotape.

Method of analysis
For this paper, the analysis shall be focused on the data concerning the general framework as well as the specific moments of training interaction concerning panegyrics. Analysing this data implies defining a method that enables to seize the essential part of education within that context. As mentioned above, it is mainly on the ATD (Anthropological Theory of the Didactic) that our analysis would be based. Within this framework, we use the concept of praxeology. From its etymology (praxis and logos), Chevallard (2007) suggests that each of the two parts consists of two components: The praxis part [which] is the union of a type of tasks (such as solving quadratic equations, blowing one's nose, composing a fugue, for example) and a technique -a way of doing -which purportedly allows one to carry out at least some tasks of the given type -those in the «scope» of the technique. The logos part [which] is the union of a whole set of notions and arguments arranged into a more or less rational «discourse» (logos), the so-called technology of the technique, which is intended to provide justification for the technique -why does it work (at least sometimes), where does its effectiveness come from?, etc. -and a more abstract set of concepts and arguments arranged into a more general «discourse», the praxeology's theory, supposed to justify the technology itself (p. 24). The praxis part of the Kela setting will be essentially analysed through what could be identified as interaction upon an object (or knowledge co-construction) between youth and elders.
As far as the logos part of the ATD is concerned, since there is no written document in this traditional context aimed at providing the so-called technology of the technique or the praxeology theory, it is therefore through the oral investigation, the discourse emerging from the interviews and what elders say during their educational interaction with youth that we shall reconstruct the logos of this specific type of training.
However, when it comes to analysing empirical data, the concept of praxeology needs to be made more operational. If, on the one hand, this concept clearly indicates the existence of a praxis that can be analysed, it does not, on the other hand, provide the tools to do so. Therefore, it is through the notion of «professional gestures» (Sensevy, Mercier & Schubauer-Leoni, 2000 ;Bucheton & Dezutter, 2008) that we shall seize what is referred to as a technique to carry out given tasks. These professional gestures constitute our unit of analysis for language teaching interaction in this paper. They are characterized by what the teacher does with his/her hand or words as a professional during teaching interaction. The action of the teacher is actually meant for others (i.e. the learners), who can be cognitively and affectively influenced by these gestures which are also designed to make them act.

The Specificity of the Oral Framework: an Emic Analysis of Transmission Processes in a Malian Village of Griots
Within the framework of our PhD (Toulou, 2008), we have carried out a research 2 where we have scrutinized the training of griots in a Malian village (Kela). These griots are considered to be the oral specialists in West Africa. Thanks to them, the oral tradition of this area has been handed down over generations since the thirteenth century without writing. In most of the scholarly literature, they are known as African bards. Locally, they are called the «artisans of the word» because most of their activities are achieved through words. They are also musicians; but the biggest part is played on words. There is a long list of the griot social functions 3 . Among others, they are usually expected by their community to: recount history, act like diplomats, provide advice from a traditional perspective, provide people with their genealogy, mediate conflicts, perform by praise-singing, teach students interested in either of these functions, etc.
As far as training is concerned, it is worth mentioning that it is not open to everybody and is selective in the sense that one should foremost belong to a jeli (griot) family in order to be admitted (to go through that process). In fact, their function is hereditary (Hale, 1998). These jeliw (griots) are endogamous people that are in charge of perpetuating the tradition of the descendants of the former Manding Empire through jeliya (orality or oral profession of jeliw). As Camara (1986) underlines, it is this group of people that is the standard bearer of oral tradition, «for, they are those who, in the absence of archives, hold the customs, traditions and governing principles (left by their ancestors)» (p.17). The griots work for a large community of people who are highly dependent on that heritage, or even for scholars who are simply interested in studying this oral culture. Within this endogamous group, it should be stressed that their most important function consists in training the young generation to carry on their tradition (Hale, 1998).
According to the people we have interviewed in Kela, this mission is incumbent on their community elders whose organisation is locally erected as an institution (where given tasks are assigned to specific people) for whatever deals with their profession. This group of elders or wise people (Toulou, 2005) is made up of a jeli kuntigi (head of griots), a kumatigi (master of the word) and other senior griot citizens (jeli kôrôbaw). These people form the decision-making body of their community. It is also within this institution that lays the responsibility of making people apply the educational rules left by their ancestors 4 . In the domain of education, these people work with unchanging principles. Thus, for instance, for anything concerning public oral transmission or events (Jansen, 2000) to young generations, they are the ones that will decide, plan and concretely organise everything. We have an example with the case of the kamabolon-ti 5 (Toulou, 2006) where the public is not selected. As far as another mode of a much more specialised and direct transmission (teacher / apprentice) is concerned, one must, first of all, fill in certain fundamental conditions so as to be admitted within that educational framework.
What filters from the interviews we carried out with the governing body of Kela is that, according to the jeli kuntigui, «the interest shown by a young griot for any aspect of the profession» constitutes a basic prerequisite to aspire to training. It is the interest that a child would show towards certain elements of their tradition «through their presence, attitude or curiosity» that would determine their predisposition to receive training. As the head of griots states, «any child that comes into regular contact or naturally likes interacting with elders is a potential jeliba (talented griot) that training should enhance». Because this training is not compulsory, any aspiring trainee should first claim (or make an effort to ask for) it either directly or even indirectly for, as the elders of Kela specify, they «can only teach to people who are hungry; the young jeliw must first show that they are actually longing for the knowledge of their oral traditional culture». As they learned from their own parents «if you teach anything to anybody who is not interested, there is probability that they won't keep it in their mind; however, if they ever do, there is the risk that they'll never use it properly because they would never see or measure its importance».
When a person is motivated or shows some interest -these elders go on emphasizing -«we are much inclined to show or teach them important things, including some hidden features or secrets, because we know that they'll make good use of them. In addition, this makes us confident in the fact that when time comes, they will also transmit it to the right people according to our ancestors' will and principles». When asked about cultural transmission itself, they answer that knowledge is everywhere. It begins from the family house to the open playground; from the public events to the instruction with a specialised teacher. The education that a young jeli receives from his/her father would be enhanced by the uncles, or possibly the kumatigi, if this child shows some aptitude and interest in the acquisition of oral skills or verbal art. As the elders of Kela say: «the kumatigi knows what the children need. He also knows what we expect from him or how to transmit our cultural knowledge».
As we can see from the above interview excerpts, if the elders have the duty to transmit their tradition to young generations, basically this should be done under certain conditions. In Kela, this can be observed in several types of training whose main aspects will be described in the following lines.
In the village of Kela, and mostly during the long dry season when many public activities are organised, there is a type of cultural life that feeds or enriches the local population. In addition, the influence of the performers creates a type of real emulation in front of the young people who would start dreaming and projecting to become professional jeliw when they get older. In that case, they would request the help of the local structure (family and/or teacher) to achieve their goal that could be related to musical or verbal skills.
In fact, the emulation can be observed at two levels that, in some case, could be complementary. The first level is the one of being a simple spectator, by watching what the other people do in the course of their profession (as performers). Practice constitutes the second level; it could take place in the form of games (when kids imitate professional at playgrounds), or during training and eventually in real professional situations (when kids who are being trained are admitted to perform in public). This local situation (with two levels that have their ramification) produces a type of education where many things are interrelated. We suggest outlining this actual situation by making a distinction between a type of socio-general training on the one hand, and a professional training on the other hand as we can see in the following figure 6 :

Socio-general Training in Kela
With regards to the first mode of training, we consider it to be a socio-general training because it is a matter of general culture, and it is embedded into the context of everyday social life; generally-speaking, this training corresponds to what Dasen (2004) describes as enculturation. In this type of configuration, we notice that young griots would globally learn by observing and imitating their elders. However, there are other instances showing that teaching also influence the socio-general training. When learning occurs, the different situations observed in this specific context enable us to argue that the particular teaching sequences happening here can either be indirect, implicit or incidental.
First of all, an act of teaching is perceived here as indirect in the sense that exposure to knowledge happens by the simple presence of young griots in a setting of teaching that was not explicitly or rather priory meant for them. This type of indirect transmission/learning generally occurs with the presence, in the bakkground, of these young people in settings where a teaching/learning process meant for other trainees is in progress. The sole presence of these young griots, in these different contexts not overtly meant for them, makes them automatically responsible for their own acquisition. In this sense the acquisition here is essentially the result of their own initiative. What the young jeliw (griots) learn here is just the result of their curiosity and quest for cultural knowledge.
Secondly, still within the socio-general training, some teaching/learning can be considered as incidental. Even if this type of teaching is strongly embedded in daily life, it can be opposed to the first type in the sense that it is explicitly meant for the person who is learning, but nevertheless happens by chance. Generally-speaking, it is a random situation, event or an element of the social context that will bring on an explanation, or raise a punctual process of teaching by an adult. In other words, in the present case, the didactic relationship is operated through the means of a direct need raised by a current daily life context. Nothing is planned here, it is the situation itself that creates or generates teaching or learning. This situation can be illustrated by Kesteloot (cited by Okpewho, 1979) who indicates in her writing that the initiation of young people could happen at any time or at any place. One would not be surprised to see teaching occurring in agricultural fields; for, «on the occasion of the teaching of an agricultural technique… a teacher could explain its myths of origin…; any topic of conversation or observation is a pretext for digression, and any digression is a pretext for teaching» (p. 254).
The third type of transmission in the case of socio-general training concerns implicit teaching. This type of transmission usually happens at the occasion of given oral tradition public events. All the didactic interest in this specific case is inherent to the type of events being reproduced in the presence of the potential depositaries of the oral tradition. From the data collected in Kela, this has been the case between February and April 2004 during most of the rehearsal sessions of the kumatigi in preparation for the big septennial ceremony of the Kamabolon which is a sacred hut containing the relics of the thirteenth century Manding Empire (Jansen, 2002). If the main objective of these rehearsal sessions was basically meant for the oral preparation of the master of the word, it is also possible to point out another underlying need for the tradition bearers to transmit their knowledge to the young generation. This need results in a sort of dramatisation that -within the framework of an article written on these sessions that we had identified as mixed 7 (Toulou, 2006) -evolves into a quadruple didactic triangulation which can be described as follows: a first triangle depicting the general conception of education (made up of: Elders -ancestral Knowledge -Young people); a second triangle representing the training of the kumatigi (consisting of: Jeli Kôrôbaw -Sunjata epic -Kumatigi training); a third one showing how during his training, the kumatigi also directly trains his potential heirs by sometimes asking them to take on the narration from where he would stop (made up of: Kumatigi as trainer -Sunjata epic -potential Heirs as trainees); a last triangle revealing how the kumatigi and his potential heirs would play a part in the process of knowledge appropriation by the other young people (consisting of: Kumatigi/Heirs -Sunjata epic -young Jeliw).
Beside the above types of socio-general training, we have distinguished another mode of training that prepares for the job by the transmission of precise knowledge that we have considered to be as professional.

Professional Training in Kela The preliminary palavers in preparation for training in Kela
This training is based on explicit terms that are negotiated within the framework of irenic palavers that precede training. As we mentioned elsewhere (Toulou, 2005), the irenic palaver and the teaching/learning sessions constitute the two indisputable pillars of this training that is mostly solicited by other jeliw coming from their villages to Kela either for an initial training or for specialization. The period of the year which is favourable for this type of training is the one that goes from December to June when griots take advantage of the long dry season (which is inadequate for tough agricultural activities) in order to carry out cultural activities.
Generally-speaking, the jeliw that report to Kela would do that on their own initiative. In the case where they have been recommended to someone who could become their host, they would report straight to them in order to have a first informal palaver. Most of the time it is open to the other members of their local contact family. In the case where they had no recommendation, in such a village of traditionalists, it is customary the first adult griot whom these visitors come into contact with, who should become their local reference and host. These visi-tors would then have the same type of informal palaver with their potential host.
The second step would then be to summon a second palaver or meeting, at a suitable time, aimed at introducing the applicant to the elders. It is within this presentation palaver that the host would have the responsibility of explaining the visitor's aims while sustaining his talk with the traditional cola nuts meant for this type of quest. This should be formulated in the very best way: which general type of training (verbal or musical); what level (beginner or advanced); in case the visitor has an idea as to which object of studies; for what sake? Etc.
Depending upon the case, the oral tradition holders would either ask for a delay in order to conduct a consultation, or give their answer right away. If the answer were positive, the head of the jeliw or his deputy would let the applicant know who their teacher would be. This person is generally chosen in reference to the object of studies. In the case of musical training for instance, they will send him to the person in charge; whereas anything else that deals with language will be assigned to the kumatigi (Toulou, 2005).
Since we have essentially followed the kumatigi's schedule, our data show that before anything happens, he will plan concretely how to teach in relation to the object negotiated with the trainee. It is the kumatigi who decides if he will be integrated into a group or if he should receive individual teaching. Now that we have seen what surrounds training in Kela, what can we notice about the act of teaching itself? Is it just one topic which is taught or are there many things that are transmitted by the kumatigi? How does this latter manage to teach?

Teaching verbal arts in Kela: the case of panegyrics
With regards to the verbal art in Kela, three topics (panegyrics, wedding procedure steps, and types of conflict mediation) were taught during the period of our data collection. We shall just focus on the first one in this paper. The teaching of panegyrics is essentially based on memory. These panegyrics are drawn from the Sunjata epic. They have a very fixed form that forces the griot to remain faithful to the original version. This is something the teacher would always remind his apprentices when they made a mistake. He organizes his teaching through some gestures that favour immediate memorisation from one person to another in the didactic system.

The specific gestures sustaining the teaching of panegyrics
From the didactic setting point of view, it is possible to draw some regular specific gestures that are used by the kumatigi during the teaching sessions with its trainees. These gestures can function autonomously or can be combined to provide a deeper insight to teaching.
Among the basic autonomous gestures, we have distinguished: -Declamation Lecturing (DL): a solitary declamation made by the teacher to illustrate one or many examples of panegyrics to his apprentices. These latter play the role of simple listeners here; -Enunciation/Repetition (E/R): this usually happens when the teacher presents a panegyric to the learners who faithfully repeat after him. -Reciting (Re): this is a way for the teacher to evaluate the trainees by asking them to recite what has been taught during a current or past session. As previously mentioned, depending on the situation or the teacher's aims, these gestures could be combined in blocks as follows.
When the kumatigi takes up a new panegyric, we notice that it is the block made up of E/R (Enunciation/Repetition) and Re (Reciting) that is used to make the learner meet the object of teaching for the first time (see figure 2). In this case, the trainer presents a model (fair copy) to the trainee who repeats each verse after him. Afterwards, the task is handed down to the trainee who should recite the whole set of verses on his own. This recitation has to be faithful to the fair copy. In other words, there are two steps of work here: the first step is a phase of collaboration where the learner is drilled by the teacher (the first repeats what is being enunciated by the second one); the second step represents didactic devolution, it is the phase of autonomy where, on his own, the learner tries to solve the task handed down to him, by reciting what has been presented in the phase of collaboration.

Figure 2: two phase-block
The second block is characterized by another form of combination. Here, the first block evolves into something different because the teacher will intervene at the autonomous (or devolution) phase to regulate what goes wrong. In this block (figure 3), the first step is the phase of collaboration (E/R, as described above); then comes the phase of devolution in second place (Re). But, the mistakes of the learner will compel the teacher to interrupt (at the intersection zone) the on-going task and to resume teaching by regulating the mistakes through explanation and repetition (E/R). This new phase of regulation will lead to devolution, where the trainee is given the opportunity to recite faithfully, without collaboration (Re).
Revue suisse des sciences de l'éducation 30 (2) 2008   337 T h e m a E/R Re Figure 3: four phase-blocks The third block (figure 4) begins like the previous one. There is a presentation that is first done in collaboration; then the task of reciting is handed down to the learner. In case of mistakes, the teacher launches a third step made up of E/R where he will make the learner repeat after him. Sometimes, the repetition happens to be difficult for the learner who, for unexpected reasons, is unable to follow the teacher's fluency. In the current case, the teacher reduces the number of verses so as to facilitate the learner's task. This partial E/R (because of the reduced number of verses) will then evolve into a new devolution where the trainee is asked to recite the restricted number of verses.

Figure 4: five phase-blocks
The fourth block highlights a successful sequence (see figure 5). After a first phase of E/R, if the second one consisting in autonomous reciting (Re) works perfectly, the teacher will then add some more verses (in a new phase of E/R). Usually the teacher, step by step, moves from one stanza to another. This block functions like a cycle that can be renewed by the teacher if need be.   6) begins in the form of verification of what has been previously done. The teacher first asks his apprentice to recite the last lesson (Re). If this is successful, he will present new features to the learner (E/R). But in case of mistake, he will make the latter repeat the same set of verses or stanzas, before assigning the task of recitation once again to him (Re).

Figure 6: three phase-blocks
The last block (figure 7) is a type of preparation for the final devolution. Here the kumatigi first presents a set of verses in collaboration. Secondly, he asks his apprentice to recite alone. In case the recitation is successful, he will add a new set in collaboration, which he will then ask the learner to recite alone (i.e. after repetition). In case of success, the cycle evolves to the end of the verse, where the teacher will eventually ask his apprentice to recite the whole panegyric. This enables the teacher to have a general control of the trainee's learning gesture by checking if, in the end, the whole praise poem could be considered as mastered.

Figure 7: two blocks
Everything happening throughout this combination of the above specific gestures function on a traditional basis made up of solicitation from the teacher and answer from the learner. This teamwork also reveals other aspects of knowledge co-construction that can be understood from the progression within one or the whole set of lessons upon the same object of teaching.
Revue suisse des sciences de l'éducation 30 (2)  From what we have just described, the way the kumatigi teaches the object can be analysed at two levels. At a micro-level, there is a kind of progression between the work which is cooperative (i.e. done in collaboration between the teacher and his apprentice) and the one that is individual (when the apprentice tries to recite the verses alone). Further analyses (Toulou, 2008) show that apart from mere memorization, there are some other activities that aim at strengthening the teaching of panegyrics. First of all, it is only when the learner succeeds in reciting the set of verses or stanzas assigned to him that, sometimes, the teacher will explain the content of the recitation to him or elaborate on some points that are important for the recitation. The kumatigi's approach seems to function here like a type of reward; but in addition, we see this like an approach inferred by the object of teaching. The kumatigi first tries to share the object or make it happen between both of them, so as to be able afterwards to point at some of its aspects through some explanation or comments. The explanation can also result from some learner's repetitive mistake that would make the teacher come back to the problem and provide more information in order to help his apprentice. The type of regulation displayed here shows how the kumatigi manages to improve the level of his trainees. Apart from the explanations after mistakes, the teacher also makes some comments that essentially take place after the learner has made a successful recitation. These comments show some important additional steps. It is only when the expert is convinced that the trainee recites the sets of praises correctly that he moves on to another level. There he actually shows how a piece of narration can be combined to a sequence of praises. It is also during this phase that the teacher provides some advice on how to handle the recitation in real professional contexts. Here, our analysis (Toulou, 2008) shows that the kumatigi teaches his apprentice how to move from the status of a simple speaker (just reciting) to the one of a real representative of a given school of tradition (for instance, the case of Kela through the way of introducing oneself, using some ancient words, pronouncing like the griots of that village, etc.).
At a macro-level 8 , we notice that after work based on the memorization of verses, then stanzas, the teacher adds another layer of information pointing to some aspects of the object being taught. These bits of information are completed by some comments after the assigned praises have been mastered by the learners. The teacher orientates his comments towards the use of the praises in real professional situation. Generally speaking, it is worth noting that the teaching movement seems to move from something simple into something which is relatively much complex. The trainees are first made to learn the content of the panegyrics by heart, before getting some explanations and then some practical information on how to efficiently use in real life contexts what they have learned from the teacher.
Notice should be made of the fact that, this teaching can be supervised at any time by any other of his peers or local wise people (who are the bearers of tradition). Within the framework of that jeliw training for the verbal art, it is also worth mentioning the existence of some public talk (locally called Jeli tolon) sessions -where after a given period of training, the apprentices are invited to entertain the audience. This takes place under the control of their teacher and the scrutiny of his peers. This is considered to be part of the training aimed at testing their level as well as assessing the type of job done by their trainer. This way of functioning shows certain forms of didactic configuration that shall be illustrated in the following lines.

The Didactic Configuration in the Setting of Kela
It is possible to point to a form of social organisation where a group of people claim to be assigned the mission of transmitting oral tradition to a younger generation. As we have already argued, this transmission can be done under a socio-general form embedded in daily life, or under a professional form detached or distinct from unplanned everyday-life activities.
As far as the professional training of jeliw is concerned, it is worth mentioning that notwithstanding its size or the lack of written documents in the system of Kela, there seem to be a type of local institution ruled by a chief (Jeli kuntigui) assisted by his peers. As we have described, it is this governing body that is in charge of the good functioning of their community. In the case of education, they are the ones making the decision to accept some trainees that they would then recommend to a trainer who is supposed to have the required body of knowledge (Chevallard, 2007). The simple fact of choosing an appropriate trainer in Kela reveals a type of work distribution. Moreover, it shows a type of specialization according to disciplines that is taken into account at the moment of assigning the training to a teacher. By accepting this mission, he knows exactly what is expected of him, as well as the other partners of the project they are involved in. It is worth noting that the ongoing training would be first based on contents carefully defined at the beginning of the process, then selected among others by the trainer who continuously goes on negotiating them (throughout some regulation) with his trainees in the course of didactic interaction. This whole process of setting the object is based on a didactic contract (Brousseau, 1988) aimed at sustaining the type of didactic relationship that is being set up. What characterizes this didactic contract is the fact that it is explicitly set on an individual basis; it is designed for a specific person according to his needs.
In this special context, the teaching/learning sessions are clearly dissociated from current life. Once engaged in this didactic relationship, the common roles and routines of the new partners change into something much closer systemati-cally to school tradition where each party plays the traditional role of the system: with a teacher (knowing much and having a given technique) and a learner (knowing less, but determined to learn from the type of tasks assigned to him), both of them interacting upon an object of teaching (which, in this case, is essentially defined through oral bases).
It is here that the ATD of Chevallard (2007) plays a major role because it widens the didactic system from school activities towards any type of tasks (between an expert and a learner) requiring explicit techniques to be able to solve them. As mentioned in the theoretical section of this paper, this type of situation, where interaction takes place between a teacher and a trainee, confirms the complexity of the object being taught and the need to go through a process of teaching aimed at helping the trainee to acquire that object throughout some means which are different from simple imitation (Chamoux, 1983).
As far as this object of teaching is concerned, we had already suggested in previous papers on the learning of oral skills that the most important part of this teaching was made possible through the channel of given genres of the manding taxonomy (Toulou, 2002(Toulou, , 2005. Thus, so as to teach the names of the descendants of famous families for instance, we see that the kumatigi actually needs to report to the Sunjata epic 9 . This kind of detour is not made by chance; rather it is guided by the awareness and, in a sense, the recognition of the richness of such a unique oral source that contains all the manding patronymics. However, since this epic is dense, the trainer would basically orientate his lessons towards panegyrics that constitute the object targeted for teaching sake. Thus, after this object has been taken from its original source, it would go into a process of transformation and simplification (Schneuwly, 2000) throughout the course of teaching for an optimal transmission to the learners. As we have seen throughout the main gestures with different blocks characterizing interaction between teacher and learners, this transmission process is done through some specific tasks based on the question/answer pattern, which Chevallard (2007) maintains as «the heart and soul of the social diffusion of knowledge».
The analysis of the interaction around these tasks shows that memorization constitutes the basic step of the teaching/learning process. This first step is implemented through some specific gestures (DL -E/R -Re) that are combined to let the trainees meet and learn the object first.
One thing leading to another, we can see how, within this system, a series of actions are gradually operated on the basis of what can be called oral archives. These offer opportunities to transmit and regulate given pieces of knowledge that, in this case, are systematised in an area where writing is excluded. Despite the absence of writing in this system, we find it interesting to investigate the type of traditional channels used efficiently in the above setting to promote and perpetuate the culture of the Kela community.
There is a work which is done here and it goes beyond mere memorization by providing some technique which would help the learners to make use of the oral archives which are transmitted to them. This transmission cannot be assimilated to what is done in other systems, because of its specificity. The tradition involved in this training has its own way of functioning that is logically different from others, such as in public schools for instance, since they don't use the same transmission medium. For instance, the logos part of this education is not written anywhere, but there is a type of shared discourse guiding the main steps of the teaching process. This whole effort of having a shared discourse (which is meant for the education of young people mainly) makes it possible to transmit their knowledge to their own children according to their assumed ancestors' rules (without distortion of facts).
There is a type of tradition going on here; a past tradition to which the concerned people attach a great deal of importance and which still needs to be analysed from inside so as to see if they succeed in achieving their goals efficiently on their own. As a matter of fact, the system, we have just explored, offers something different; it is made in such a way so that communication is insured through the only oral channel that manages everything, at least, as far as the education of oral tradition is concerned. It is within this oral process that the tradition of these local people has been perpetuated over the centuries.
Notes 1 Griots are both female and male. The double gender he/she or his/her will be used in this paper for general cases ; but for specific cases, like the group of teachers and learners we have studied, we shall only use the male gender, because they were all men. 2 The research has been sponsored by the «Fonds National Suisse pour la Recherche Scientifique» (FNRS N°101412 -102041) 3 Refer to Hale (1998), for more details concerning the griot function 4 This idea is quoted from the series of interviewed we had carried out with the elders of Kela within the framework of our PhD research. 5 This is a very big septennial event in the mande world with the renewal of the ceiling of a sacred hut (called kamabolon) where griots have to perform. For more details, see Jansen (2002). 6 In this figure, the two types of training are complementary. The figure should not be understood as a ranking in terms of hierarchy or size. 7 In that study, it is on the basis of a double configuration incarnated by the presence of a didactic contract (Brousseau 1988) when one contrasts the training of the kumatigi on the one hand, and the impact this activity has on young griots on the other (for more details, refer to Toulou, 2006). 8 It should be noticed that this level tries to indicate the general tendency. This does not mean that what is described in general terms happens all the time; we are only trying to draw the tendency of the whole teaching sequences dealing with the same object in order to establish the general steps or framework that the kumatigi follows. 9 Johnson (1986) and Jansen (2001, p. 206) also consider this epic to be a kind of encyclopaedia or a mental text where griots pick up and organise what they need for their verbal art.