This case study is based on ethnological fieldwork which was carried out from February to October 1996, from January to April 1997, and from July to December 1997. The empirical research among the Wodaabe families of the Siiganko’en lineage was characterized by two features which are methodologically important to the study of indigenous knowledge. First, research activities took place in a situation which anthropologists call “participant observation”, i.e. the researcher lives in close contact with the people he is studying over a considerable period of time. This has some important implications:
By living together in the bush, sharing meals, seeking protection from the fierce heat under the same trees, and by accompanying the nomads during their migrations, the researcher shows that he holds the lifestyle of his hosts in high esteem. This factor nurtures that confidence in people that is a prerequisite to sharing their knowledge.
The researcher lives through all sorts of situations as they develop. Even the very best informant cannot anticipate all possible aspects of an event that might be important to the researcher, and, conversely, the researcher can never contemplate all the factors he may need to know in order to describe events and behaviours. Only direct experience will allow him to make observations about hitherto unknown themes, and enable him to pose meaningful questions.
There will always be a difference between what people say and what they do. Therefore, interview data should be cross-checked with observation data; discrepancies between the two sets often lead to interesting research questions.
The second point concerns how the Wodaabe’s environmental and herding knowledge was elicited. The nomads’ knowledge of geography, geology, climate, wild plants and animals, time and space relations, pastoral mobility, herding, and the social organization of pastoral work was captured using an “encyclopaedic” approach. Capturing means that the researcher puts himself in the position of a disciple who learns step by step from his instructors, the pastoralists. Knowledge here always consists of two overlying views of reality: empirical, factual statements about what is or what happens in the world, and normative statements by which the world is represented, or expected to be. Moreover, it is important to note that the entirety of these (and many other) themes are also a reference to any single theme. Therefore, knowledge of the whole (i.e. the context) is a prerequisite to an understanding of a particular topic. Take a simple example: The tree Ficus platyphylla takes its name kalkaldihi from the word which designates the breeding bull of the herd. Thus, a knowledge of herd structure is vital to understanding the significance of the Ficus platyphylla for the Wodaabe.
As pointed out above, not all knowledge can be verbalized. Language, though, is the first key if scientific understanding of indigenous knowledge is to be developed. Verbal statements were tape-recorded, transcribed and then analysed. Asking for names of things (e.g. plants) is a starting point. The classifications adopted can be obtained with analytical tools, like pile sorting and paired comparison (Martin 1995). Question frames are used to arrive at definitions of concepts and propositions (cf. Ellen 1993: 60). Preference ranking and triadic comparisons help to obtain evaluative statements, e.g. concerning the quality of fodder plants. These question techniques are very narrow in that they clearly pre-structure what the informant is expected to respond. Therefore, they are mainly suitable for the purposes of systematizing and comparing data.
However, in order to obtain less filtered data, more open forms of data collection are needed. Asking informants to describe whole processes or to recount events is one possibility. Just accompanying them and letting them comment on what is happening and what they are doing is another. Here, those under study still refer exclusively to the researcher. That is, their statements are structured by what they expect him to be able to understand - or worse - to be willing to hear. Therefore, just listening without interrupting (e.g. the herders discussing pasture information they obtain from their scouts (see photo 5), or the political speech of a local leader), is one way to have more authentic information. There is much more language competence needed to follow free conversations of the local people than to understand longer accounts that are directed towards the researcher. But it is with the first method that new and often very insightful information can be obtained, when other techniques have exhausted themselves. Data from these longer speech acts are usually less systematic and more difficult to compare, but give rise to a whole context of conditions, associations, side effects and causes that is relevant for understanding the issue at hand. None of the techniques cited for oral data collection is sufficient for investigating local knowledge. Rather, they need to be combined in order to form a reliable method. Information obtained with one technique is cross-checked with that obtained with the others, and insights gained by applying one technique are used to develop questions within the structure of the other techniques.
