37

thus very n- - is (Photo: P. Richards) pest control. describes over two hundred current versions of describes over two hundred current Mancala board in use as a means of scaling farmers' investment priority preferences concerning weed and Ayo is a Yoruba word meaning 'the game' and is played on a wooden board which has two rows of six holes carved into it, as in figure 1. Other of weeds, pests and farm methods. One of the of weeds, pests and a traditional Yoruba game using the format of which originated in generically term Mancala, of all games still played Egypt and is the oldest Mancala is derived from in the world. The name 'to move' an Arabic word naqala which means across the and the movement of the game itself game to globe is interesting; it is the only major dominoes, spread from west to east (unlike ludo, during the dice, etc.). It diffused west and south to the West spread of Islam and was also carried trade. Its Indies at the time of the African slave present geographical distribution where it is extensive and covers most of Africa, game, the claimed by some to be its national (1952) Middle East and south east Asia. Murray the game. different methods of measuring their perception different methods successful of these attempted most interesting and and stimulate discussion to generate information Ayo. ancient game of Mancala The diffusion of the version of a game ethnologists Ayo is the Yoruba and cultural behavioural 'appropriate pre-testing and of familiar Oguntoyinbo in preparation for the to use type and careful Nigeria. Nigeria, examined several effort this Barker, in significant 1975; locally Many of the ideas expressed in this paper are the result of fieldwork undertaken in conjunction with Paul Richards. Joyce Tait, David Atteh with the assistance of Sammy Ogunjimi, Robert Olanipekun and Sunday, at the village of Workshop on Perception of Environmental Quality, organised by the International Geographical Union and held at the University of Ibadan, Nigeria, July 22-29 1978. The author gratefully acknowledges financial support for this fieldwork from the Hayter Travel Award Committee of the University of London. Oluwatedo, Oyo State, 1 community in which he works. The danger is community in which he works. The social that an uncritical reliance on transferred and possibly misleading. music and Verandah story telling and traditional modes poetry are examples of locally important reservoirs of social interaction in Africa, and are incorpor- of community expertise which can be of this ated into research designs. The purpose simple, short paper is to illustrate how a similar yet a research pastime can be adapted and refined as tool. Field trials with local farmers at Oluwatedo village, Oyo State, The recent growth of interest in the utility of The recent growth knowledge in Africa indigenous environmental (Richards, Richards, 1976; O'Keefe and Howes, infra) has Richards, 1976; O'Keefe into focus the cross-cultural brought more sharply conventional geographical limitations of many perceptual and behavioural methods for collecting perfected in industrial data. Elicitation techniques suitable in develop- societies are not automatically ing countries and so role in evaluation of methods assumes a vital a more conscious methodologies' is part of a research. The data collection process and local social contract between a researcher to ensure people, and the onus is on the former cultural his techniques are appropriate to the and are context in which they are to be used, of the tailored to the abilities and requirements culture- science methodologies, often embodying can bound assumptions about people's behaviour, be partial produce research conclusions that may Appropriate methodology David Barker Appropriate Methodology: Appropriate Game African Board a Traditional using An Example Images1 Environmental Attitudes and Farmers' to measure advocates research design. Indeed, Barker (1978) Africandesignssometimes includeadditional to produce from local materials; it can even be storage holes, offset at the end of the board for played by scooping holes in the ground and using captured pieces. Where the game differs more stones. In addition to being a popular pastime (it substantially there may be more than six holes was once the Royal Game of the Ashantis), it per row, as in some East African versions of Bao, is also a device for teaching young children skills and perhaps even four or six rows. Ayo is played in mental arithmetic. with 48'pieces',normally cannalilyseeds, although in general any locally available items such as seeds, shells, beans or stones are used. The Atteh board Although some versions of Mancala, like the The original idea and adaptation of Ayo as a Akan game Oware in Ghana, can be played with research tool came from David Atteh and Paul four or six players, Ayo is for two people and the Richards who devised a board whose basic objectistocapture more seeds than your design is shown in Figure 2. Atteh, and also opponent. The game proceeds alternately, each Flynn and Knight, have applied techniques which player removing the contents of one of his own utilise this board as part of carefully integrated holes and placing seeds, one per hole, serially in surveys2 of farming behaviour and indigenous an anticlockwise direction. Different versions of knowledge in south west Nigeria. During the Mancala have different rules about how seeds course of these interviews a farmer is asked to may be captured. imagine he has N100 to invest in his farm and The game is very skilful, played at high speed, and is widely available in African communities 2 This research is in progress at lITA, Ibadan, Nigeria. The Atteh board used at Oluwatedo was kindly made available to because the basic equipment is easy and cheap the participants of the field trials by the lITA team. Figure000000 1. The basic layout of an Ayo board 000000 Figure 2. The basic layout of the A tt eh board

small scale farmers f- 000000000000-6040 20 0 20 40 60 80 100 120 140 160 150 100 50 0 50 100 150 200 250 300 350 400

large scale farmers

38 to indicate his likely returns over the next five struct can be defined as each end hole of a par- years. He is given five Ayo seeds and drops them ticular row on the board. All the repertory grid into the holes representing these returns. Next, elements, which might be a setof specimen he is given three seeds and asked to show his weeds, can now be scaled along the construct. best, worst and expected returns over this five To do this, each weed is represented by an year period. Both procedures lead on to more Ayo seed, and the farmer drops them into general questions from the interview schedule those holes whose positions reflect how easy or and discussion about his decision making. difficult each weed is to clear. Thus the construct can be scaled in a way which permits the farmer A variation tried during the Oluwatedo field trials to view each scaling decision in the light of his using this board was to examine farmers' pre- previous decisions about the other weeds in the ferencesfordifferentcombinationsoffarm set. returns over three year periods. One farmer, for example, preferred returns of N 100, N150 and The procedure can be taken a stage further and N200 to N400, N200 and N 100, and explained another construct defined and scaled alongside his liking for the smaller net gain of the first the first using the second row on the board. Con- three year combination in terms of his not want- structs may then be compared directly and ing to make a loss for any given year. Similarly, methods devised to determine their superordinacy he preferred N200 and N250 over two years to relative to each other. Again, weeds can be com- N50 and N400,a arguing that the first was more pared by asking questions about their different 'progressive' because it yielded a steady income. positions on each construct. Only slight modifications are necessary to con- Perception studies and the Ayo board struct A yo boards with a variety of different pro- perties in a similar way to the Atteh board. For If the Ayo board is viewed as a two dimensional example, a board could have enough holes to matrix, with the holes being equivalent to the correspond to the months of the local calendar matrix cells within which frequencies (i.e. num- year. This might be a quick method for determin- bers of seeds) can be represented, there are ing the proportion of time spent on different farm innumerable methods and scenarios for eliciting operations throughtheyear.Also,semantic information, and the information need not be differential scalings could be obtained by using restricted to farming behaviour. For example, boards whose rows contain an odd number of a row canbedefinedas an attitudescale holes, or boards might be embellished by paint- of six divisions where the two end holes are ing them green and black to indicate the favour- assigned values of +3 and 3 and intermediate able and unfavourable parts of a scale. Another holes are the discrete positions between. Using strategy is to give a handful of Ayo seeds to this idea, farmers at Oluwatedo were able to scale farmers to represent a given amount of money the severity of the grasshopper pest, Zonocerus and ask them to allocate the seeds between holes variegatus, for each of a number of preceding to indicate their expenditure over the next year, years by simply dropping a seed (for each year) where each hole represents a different variety of into the appropriately positioned hole, and then yam or rice. describing and explaining their decisions. As an extension to this technique, a second pest might be selected and its severity scaled in the same Some implications for research methodology way along the second row of the board; com- It could be argued that any search for 'appro- parisons between pests are then possible by com- priate' research methods will result, implicitly, menting on their relative positions along the scales. in a compromise between the field techniques of anthropologyandthoseof'western'social An alternative adaptation of the Ayo board is to science. The constraints under which many social elicit and scale personal constructs from a reper- scientists work in the developing countries often tory grid.4 Suppose a farmer is asked to contrast preclude the use of the classic field techniques of a set of three weeds (the grid elements) and he the anthropologist (Crane and Angrosino, 1974), provides a construct such as 'easy to clear- which normally take many years to bear their difficult to remove'. The two poles of the con- richfruit. On the other hand, much of the impetus to engage in the experimental work described here arose from a felt need to eschew 3 The Nigerian currency unit is the Naira, roughly equivalent to £1. the central role of the questionnaire, traditionally 4 For an introduction to the literature on Personal Construct important in the design of geographical research Theory and repertory grids, see D. Bannister and F. Fran- sella, 1971. into agricultural systems.

39 The view which regards participant observation together with locally significant innovation in and survey research as two extremes of a con- techniques. This type of multi-method approach, tinuum of research techniques is, however, not given carefully designed research programmes, particularly helpful. Thus whilst Warwick (1973) could provide a variety of different learning for- provides a useful focus for an informal cost- mats and experiences for both research worker benefit analysis of the two, his conclusions are and farmer, and encourage mutual understanding disappointing in their self-evidence: the strength and co-operationinagriculturalresearchin of sample surveys lies in their greater potential developing countries. for quantification, replication and generalisation whilst participant observation scores on qualita- References tive depth, flexibility and more detailed analyses Bannister, D., and F. Fransella, inquiring Man, of social processes. Indeed, there are much more Penguin Books, 1971 usefulframeworks thanthesocialsurvey-- participant observation continuum for evaluating Barker, D., 'A note on research methods in the research elicitation techniques (see for example, studyofIndigenous Environmental Know- Whyte, 1977). ledge', background paper for the IGU Work- shop on Perception of Environmental Quality, Cole, Gay, Glick and Sharp (1971), studying the University of Ibadan, Nigeria, 1978 mathematical ability of the KpeIle in Liberia, faced similar methodological problems in develop- Barker, D., J. Oguntoyinbo and P. Richards, The ing suitable elicitation techniques, and advocated utility of the Nigerian peasant farmer's know- an 'experimental anthropology' in which they ledgeinthemonitoringofagricultural adapted and integrated methods from psychology, resources, MARC Report No 4,Chelsea anthropology and sociology.Theystressthe College, London, 1977. importance of setting people tasks within the Cole, M., J. Gay, J. A. Glick and D. W. Sharp, home cultural milieu (i.e.in familiar settings) The cultural context of learning, Methuen, and involving routine activities to elicit categories London, 1971 of response relating to everyday experience. Thus Crane, J. G. and M. V. Angrosino, Field projects they refute the view that posing hypothetical in anthropology, General Learning Press, 1974 questions and tasks to people in non-literate societies violates cultural norms and generates Lindzey,G.andJ. S.Thorpe,'Projective atypical cognition. Gay has subsequently used techniques', in D. L.Sills (ed), international other projective techniques, such as sentence Encyclopaedia of the Social Sciences, vol 14: completiontestsand wordassociation,and 561-568, 1968 recently Turner (1978) successfully applied the Murray, H. J.R., A History of board games same methods in studying the vernacular geo- other than chess, OUP, 1952 graphy of the Sesotho people in Lesotho. O'Keefe,E.and M. Howes,'Theusesof The Oluwatedo field trials, and the Ayo hoard indigenoustechnical knowledgeindevelop- experimentsprovidebutoneexampleofa nient: an annotated bibliography', infra technique which shifts the initiative in providing information to local people, or 'respondents' to Richards,P.,'Alternativestrategiesforthe give them their depersonalised form. This seems African Environment', inP. Richards (cd), to be very important inoral cultures where African Environment, Special Report No 1, questionnaire schedules can act as both steering IAl, London, 1975 wheel and brakes on the free flow of discussion. Turner, S., 'Sesotho farming: the condition and Paul Richards has subsequently used both ques- prospects of agriculture in the lowlands and tionnairesand Ayo techniquestointerview foothillsofLesotho',unpub. Ph.D.thesis, farmers in Sierra Leone, and has tape recorded University of London, 1978 very positive assertions that they prefer the game Warwick, D. P., 'Survey research and participant format. An appropriate methodology, therefore, observation: a benefit-cost approach', in D. P. would use techniqjes which can structure an Warwickand S. Osherton,Comparative interaction so that the initiative rests with a local research methods, Prentice-Hall, New Jersey: community, or an individual. A research strategy 189-204, 1973 might thus involve a range of different methods suchasprojectivetechniques(Lindzeyand Whyte,A.,'Guidelinesforfieldstudiesin Thorpe,1968),and morelimitedformsof environmental perception', M.A.B. Technical participant observation and questionnaire work, Notes 5, 1977

40