Indigenous Knowledge and Development Monitor, July 1998

Contents IK Monitor 6(2) | IKDM Homepage | Suggestions to: ikdm@nuffic.nl | (c) copyright Nuffic-CIRAN and contributors 1998.

Learning local knowledge of soils: a focus on methodology by Deirdre M. Birmingham.

There is little published literature on methods for studying local knowledge of soils. This scarcity is surprising, given that local knowledge is increasingly recognized as necessary for effective agricultural and rural development, and the livelihood of farmers. This article describes and analyzes the approach and methodologies used in recovering local knowledge of soil and land typologies among two ethnic groups in two different agroecological zones in West Africa. This analysis is of particular importance, as the manner in which research is carried out is critical to the quality and type of results obtained. Not only the results should be published, but also an analysis of the methods used, with a view to improving future research.

There is growing interest in local knowledge of soils, as evidenced by the increasing number of studies (Niemeijer 1995; WinklerPrins 1998). These studies focus mainly on indigenous soil classifications, often termed ethnopedology, but few provide detailed information on the methodologies employed. I did a comparative study of the local soil and land typologies, as well as the land management practices of two ethnic groups: the Bété in the equatorial forest zone and the Senufo in the Guinea savannah zone, both in Côte d’Ivoire (Ivory Coast) (Birmingham 1996). There was an advantage in studying two cultures: the distinguishing features of each culture’s soil typology were highlighted by the comparisons and contrasts. The Bété consistently described 11 mutually exclusive soil types, while the Senufo claimed to have three different soil types which were not mutually exclusive. The conclusions I drew concerning the effectiveness of certain approaches and methods employed by one group were often tempered by contrasting, and sometimes puzzling, experiences with those same methods in the other group. I will start by describing the approach and methods used.
Soil identification exercise (with Bété in the forest zone). The recorder and paper used to represent toposequence are visible.
Photos: Deirdre M. Birmingham

Description
I chose two villages in each of two agroecological zones, each inhabited by an ethnic group indigenous to that zone. In each of these four villages the West Africa Rice Development Association (WARDA), with whom I did this research, conducts village-level research on rice production systems. WARDA also participates in an agroecological characterization of inland valleys in West Africa. The use of these sites allowed me to compare my results with those of socioeconomic studies conducted by WARDA, and also with the findings of agroecological characterization (Windmeijer et al. 1994).

Anthropological approach
I opted for an anthropological approach based on semi-structured interviewing, observation and participation in agricultural operations during a period of one year. Introductions to village leaders and interviews began during the dry season following the harvest in each agroecological zone, when there is less farmwork. I used group and individual interviews, and incorporated visual participatory aids into the interviewing (e.g., maps, calendars and diagrams). The day and time selected for the interviews were those most convenient for the participants I wanted to talk to. Separate group interviews were held for men, for women, and for younger men (les jeunes), in deference to village social structure. The interviews with individuals continued until no new information on major topics was forthcoming. Some 250 individual semi-structured interviews were held with approximately 192 different people; in each zone similar or representative numbers of men and women took part.
Interviews were conducted in the local language. I hired a research assistant in each zone to translate the local language into French and to help with soil sampling. The assistants were crucial to the quality of my data, as they were my access point to local farmers. All had worked for WARDA and came well-recommended; there was an initial trial period of one week, and all the assistants were given a manual on the field methods used in rapid rural appraisal (Gueye & Schoonmaker Freudenberger 1991).
All interviews were recorded on audio cassette, whether conducted in the village or in the farmers’ fields. I reviewed the cassettes daily with my assistants and made written notes, using English paraphrases rather than French transcriptions.

Specific methods
A local soil typology was elicited in each group meeting (Birmingham 1996)1. Participants sometimes asked for prompting questions about the specific information needed. The Bété tended to make this request more often, apparently to give them a mental checklist to help them describe their many soil types more efficiently and thoroughly. This checklist evolved gradually on the basis of the way in which the various soils were described.
The participants volunteered to show me the different soil/land types in situ, which allowed me to further assess whether they were describing soil or land types. Each soil/land type was repeatedly identified by villagers in different age groups, both men and women. As this took place during the dry season, I was able to auger soil samples for laboratory analysis before the rainy season; in this way, quantitative measures could be compared within and between soil types, as well as with local qualitative descriptions.
Soil mapping, ranking, and identification exercises were done with groups and individuals, in order to explore more deeply the bases for local soil typologies. For example, months after collecting soil samples, Bété villagers were asked to name unlabeled plastic bags, each containing a different dried soil sample. To aid in identification, they were allowed to handle the samples and moisten a portion of the soil. As they named each soil, they also gave the reasons behind their determination. They then indicated the usual location of that particular soil type within the toposequence: each bag of soil was placed on a large piece of paper on which a sloping line had been drawn to represent the toposequence. In another soil identification exercise, one participant selected a bag of soil and described it to a second person, who then determined the soil type on the basis on the description alone. This allowed me to hear the key characteristics of each soil type. Identifying soils in plastic bags rather than in situ also confirmed that the Bété and the Senufo do distinguish and name the soils separately from the surroundings.
Methods were iterative and innovative, depending on the situation and the different responses among social and, more particularly, ethnic groups. I will return to this point below.

Analysis
I typed the notebook entries onto computer disks before leaving Côte d’Ivoire and later analyzed them, using HyperRESEARCH™ software. This software facilitates the analysis of qualitative text-based data through the computerized coding, retrieval, and manipulation of data. Interviews were coded according to name, gender, age category, date, village and agroecological zone.
Using information on the Senufo noun-class system and semantic analyses provided by missionaries who had worked among the Bété or the Senufo, I gained a better understanding of the local soil/land typologies, particularly that of the Senufo. Linguistic variation sometimes occurs among Senufo villages; this ranges from different pronunciations to complete languages which are mutually unintelligible between neighboring villages.

Analysis of approach and methods
By opting for an anthropological rather than a structured survey approach, I was able to gain a more thorough understanding of the epistemology of local knowledge of soils, and of the reasons behind differences in land-related knowledge and changes in attitude. The use of semi-structured interviews gave villagers greater freedom in describing soils than a questionnaire or a survey approach with structured response categories. Leaving the descriptions open-ended also helped the participants to see conceptual linkages between such aspects as vegetation, animals, organic matter, soil quality and land quality.
Questionnaires or more structured surveys are useful when more specific or quantified information is needed. For example, I developed a questionnaire for my assistant in the forest zone which focused on the incidence of individual female plots by age and marital status, as well as on motivation and rituals pertaining to the land. He was encouraged to add questions he found pertinent or interesting.
Farmers appreciated the fact that the group interviews were conducted during the dry season and on days of rest. The rapport built up during the dry season made farmers more willing to take part in interviews during the rainy season, when they were busy. Group interviews were best done on regular days of rest, avoiding holidays, initiations, funerals, etc. The division of agricultural labour by gender meant that men and women were available at different times of the agricultural calendar. Appropriate scheduling yielded a more representative sample of those actively engaged in farming, as opposed to those readily available because they were not physically engaged in agriculture. The women had domestic responsibilities at specific times of the day, which were best respected.
Iterative interviewing and observation right across the growing season facilitated a deeper understanding of local soil typologies and differences in the management of certain soil or land types. While the Senufo claimed only three soil types, they pointed out variations within plots of the same soil type; however, these were only described during interviews in their fields. Observations made in farmers’ fields revealed links between what they said (theory) and what they did (practice), and also illustrated practices and beliefs which either could not be readily described or represented tacit rather than cognitive knowledge.
The quality of translation work needs to be constantly monitored. Researchers who speak the local language or translate into one language rather than two are at an advantage when studying local knowledge. Linguistic information on the Senufo peoples was crucial in that it helped me to understand why nearly 20 terms were elicited for only three soil/land types. Access to this linguistic information during the fieldwork would have clarified things at an earlier stage, giving me more time to refine my understanding of the farmers’ land management practices.

Flexible and creative methods
Methods are needed which are flexible, responsive and creative, so that they can be adapted to cultural preferences and village priorities, enabling researchers to make optimal use of the available time. A few examples will serve to illustrate this. While waiting for a farmer whose fields I was to visit, I interviewed several senior men to learn about the Senufo calendar and seasonal activities. At one point, a respected elder described the different soil types so extensively that I asked if he could map them for me. He agreed, and a date was arranged when other elders could participate. The map was drawn on the ground at the chief’s compound, and local materials were used to designate different soil types and geographic features.

Group versus individual interviews
Group interviews worked differently among the Bété and the Senufo. Culturally, the Bété were relatively easy to interview; they are an open and verbal people who readily adopt Western ways, and some of them speak French. Individuality is one of their main characteristics, and when one Bété was chosen as spokesperson, the others were free to supplement or question his descriptions. As a result, the group interviews furnished more complete and accurate information on all soil types than the individual interviews. This is understandable, since there are so many soil types that a single individual can easily forget one or two of them. Group interactions enlivened the process, and made the interviews less intense and tiring. More soil identification exercises were done among the Bété than the Senufo, due mainly to the fact that the Bétés were receptive to these techniques, while the Senufos took no interest in them.
In Senufo society there is strong pressure to conform. The women in particular are suspicious of outsiders, and it was more difficult to build up a rapport with them and to interview them. As a rule, they did not like to meet in groups, a reluctance particularly apparent in one village and even experienced by a development worker who is himself Senufo. One person was the group’s spokesperson and the others remained silent, even when I requested their verbal participation. Consequently, only nine group sessions were held with the Senufo, as opposed to 20 with the Bété. The latter frequently had 20 or more participants, whereas only a handful of Senufo generally took part. More individual interviews were held with the Senufo, which was due in part to the fact that the agricultural practices of the Senufo were more varied.

Social and cultural characteristics
Interviewing men, women, and young people separately allowed age- and gender-based differences in knowledge and practice to surface. The women were often reluctant to speak in the presence of older men, especially men of status. In general, men and women had more similar than dissimilar knowledge of the various soils in each culture, despite the differences in their agricultural, domestic and social roles. The ease with which information was shared varied according to age and gender, especially among the Senufo. Older persons were more accustomed to instructing younger people, and thus more disposed to teach me about their soils. The men, who were often more educated, were frequently more articulate.
I came to see how agriculture fits in with broader cultural characteristics, and how labour is socially organized and performed. This gave me insight into both the epistemology of land-related knowledge and its inter-generational transfer. A survey approach might miss these aspects, especially the links between them. Teaching and learning about agriculture and land-related knowledge was better observed than would have been possible by interviewing the villagers. In both cultures learning is highly socialized and experientially based: children learn practical and social skills through their families and through "poro", the traditional structured educational system maintained to varying degrees in Senufo villages. Senufo competence in, and esteem for, agriculture is transferred not only through copying the agricultural practices of parents and relatives, but also through group activities; thus participants in poro classes might perform farm tasks for respected elders. Group work and hoeing competitions among young Senufo men are occasionally performed to balafon and drum music. Not once did I witness such work being done among the Bété, whether performed by a group or in the form of a ritual. Observing these major cultural differences helped me to understand differences in the degree of interest in, and knowledge of, the land and its soils, in particular between Bété and Senufo youth.

Gifts and compensation
Published studies on local knowledge of soils have not addressed the subject of compensation or gifts; yet villagers understandably want to know what benefits the research will bring them. After exploring the cultural norms and potential ramifications of gifts or compensation, I did present gifts. These were intended to express my appreciation and feelings of friendship for the village as a whole. Separate gifts that were appropriate for both men and women and could be easily divided and distributed were given in honor of the new year and again at the conclusion of my work. These were presented to village leaders in the presence of other villagers. It is also customary for younger adults to provide elderly chiefs with small gifts, and I did so when I periodically returned to the village during the year I spent doing my research. Sometimes I gave an individual I had just interviewed a small token of appreciation in the form of coins, cola nuts, or aspirin, which I always had in my backpack.

Conclusion
My purpose in describing and analyzing my methods was to contribute to the existing literature on the subject, and the methodology employed was as important as the data obtained. Greater efforts in the area of research, extension, and development are needed in order to learn about, work with, and strengthen local knowledge. Only then will we see improvements in local livelihoods and the land resources on which they depend.

Deirdre M. Birmingham, Ph.D.
905 Yukon Drive
Alpharetta GA 30022
USA
Tel.: +1-770-993 9651.
Fax: +1-770-993 7061.
E-mail: deirdreb@mindspring.com

1, For more information on the results of the research, as opposed to the methods, please contact the author.

References
- Birmingham, D.M. (1996) Local knowledge of soils: the case of contrast in Côte d’Ivoire and its considerations for extension. Unpublished Ph.D. dissertation, University of Wisconsin-Madison.
- Gueye, B. and K. Schoonmaker Freudenberger (1991) Introduction à la Méthode Accélérée de Recherche Participative (MARP) (2nd ed.).
- Niemeijer, D. (1995) ‘Indigenous soil classifications: complications and considerations’, Indigenous Knowledge and Development Monitor 3(1).
- Windmeijer, P.N., N.V. Duivenbooden and W. Andriesse (1994) Characterization of rice-growing agro-ecosystems in West Africa. Semi-detailed characterization of inland valleys in Côte d'Ivoire (Technical Report No. 3). Wageningen, the Netherlands: DLO Winand Staring Centre for Integrated Land, Soil and Water Research (SC-DLO).
- WinklerPrins, A.M.G.A. (1998) ‘Local soil knowledge: a tool for sustainable land management’, Society and Natural Resources 11(7) (October issue).

Back to: top of the page | Contents IK Monitor 6(2) | IKDM Homepage
Suggestions to: ikdm@nuffic.nl
(c) copyright Nuffic-CIRAN and contributors 1998.