Indigenous Knowledge and Development Monitor, March
2000
Indigenous Knowledge and Development Monitor, March
2000
Contents IK Monitor (8-1) | IKDM Homepage | Suggestions to: ikdm@nuffic.nl | © copyright Nuffic-CIRAN and contributors 2000.
Review: Modelling and indigenous knowledge
Reality in its full complexity can never be fully understood. All of us create our own images or models of reality. Holders of indigenous knowledge (IK) make their own models in their minds. And when research scientists conduct studies or experiments, they formulate hypotheses and construct formal models (thought models, flow diagrams, mathematical models of various kinds, etc.). I was asked to review recent publications on modelling of 1) the dynamics of agricultural development and 2) vegetation dynamics in semi-arid grazing systems. I found both books to be quite interesting but could see little relationship with IK. After discussion with the journal editor, we agreed that I would review the books only briefly and would add some general comments on the usefulness of scientific modelling for farmers and development workers.
Bontkes, Tjark Struif (1998) Modelling the dynamics of agricultural
development: a process approach - The case of Koutiala (Mali). Tropical
Resource Management Papers 25. 233 pp. ISSN 0926-9495. NLG 40. Wageningen
University and Research Centre Liaison Office, P.O. Box 9101, 6700 HB
Wageningen, the Netherlands.
Tel.: +31-317-484293.
Fax: +31-317-484292.
In its own words, this study 'explores the suitability of dynamic simulation
modelling as a tool to help decision makers …' For a cotton-growing area in
southern Mali, the author presents two models-one for the farm level and the
other for the regional level--which try to integrate biophysical with
socio-economic components, in this case crop and animal production with labour
and farm economics. Farms are classified according to wealth, family size,
crops, farm equipment, soil quality, etc.
The models reveal the limitations of holistic approaches. They are too complex
for most people to understand quickly, but are still very simplified views of
reality. As examples of the non-captured complexity: the models include only
cattle as livestock, but the farm families also keep goats, sheep or chickens;
no differentiation is made in division of labour according to gender; and a
leguminous crop, dolichos, is viewed only as forage although it produces edible
beans.
The recent shift to cotton production in Mali has increased the wealth of many
farmers, but the models show that unless land-use practices change, the yields
of both cash and food crops will eventually decline. An important conclusion
from the model for the region is that it cannot support a growing human
population and provide for higher incomes at the same time.
The models are then used to explore some alternatives, such as keeping livestock
in stalls. The choice of scenarios tested (e.g. lower fertilizer price and
higher cotton price) is somewhat optimistic. What would happen if the cotton
price falls and the fertilizer price rises? For an understanding of the possible
impacts of change, it would have been more helpful to investigate farmers'
current practices more deeply and, as far as the farm model is concerned, to
concentrate on fewer but key components, to do this in more detail, and not to
attempt to explain an entire production system in equations.
Rietkerk, Max (1998) Catastrophic vegetation dynamics and soil degradation
in semi-arid grazing systems. Tropical Resource Management Papers 20. 155
pp. NLG 40.
Wageningen University and Research Centre Liaison Office, P.O. Box 9101,
6700 HB Wageningen, the Netherlands.
Tel.: +31-317-484293.
Fax: +31-317-484292.
This doctoral thesis is a collection of papers dealing with vegetation
dynamics and soil degradation in semi-arid parts of Africa. The catastrophe
theory assumes that certain events, such as droughts, can push the vegetation
beyond a threshold from which it is difficult or impossible to recover. The
proof for this is based on some observations and much theoretical reasoning and
model calculations. Already in the introduction, the author states that he does
not agree with a recent theory that vegetation dynamics in drylands are
determined mainly by rainfall and its variability, rather than by grazing
pressure. This refers to the 'non-equilibrium theory' that variability in annual
rainfall leads to great variability in annual vegetation yields and composition,
and that managing pasture by adjusting stocking rates therefore cannot, on its
own, achieve equilibrium in vegetation.
The examples chosen to disprove the non-equilibrium theory are derived from the
more humid areas of the 'drylands' (seasonally dry tropics, with annual rainfall
of 600-800 mm). But proponents of the non-equilibrium theory say that where
annual rainfall varies by more than 33 per cent, vegetation will seldom reach
equilibrium. This is found in areas of West Africa with mean annual rainfall of
400 mm or less. The author's own observations show that outside cropping zones,
vegetation recovers quickly after rain, especially if most of the plants are
annuals and soils are light and sandy. A catastrophic collapse of vegetation is
more likely if perennial herbaceous or woody plants predominate.
The author of the thesis does not show any link between his ecological theories
and indigenous knowledge and practices. Nevertheless, for readers interested in
ecological theory, the thesis provides stimulating reading.
General comments: how can scientific modelling be relevant to IK?
At the International Rangeland Congress (IRC) held in July 1999 in Townsville, Australia, a special session was devoted to the modelling of rangeland systems. Although many Australian ranchers have college educations and are computer-literate, they complain that scientific models are too complicated to be useful. Even the widely known decision-support models are rarely used in practice. The data required are not easily available, and only a specialist can really understand the models. One conclusion reached at the IRC was that two broad types of models should be differentiated:
If we add graphic and thought models (e.g. flow diagrams and various kinds of
maps and calendars), the link between models and indigenous knowledge can be
more clearly seen. Graphic and thought models can pinpoint essential strengths
and weaknesses in an existing system. Identifying patterns of activities
(calendars) and movements of resources as inputs and outputs (e.g. bio-resource
flow diagrams) can help both land users and participating scientists to gain a
better understanding of a local farming system and environment.
Scientific (mathematical, computer-based) models can help verify indigenous
knowledge or show its limitations. This is particularly true where scientific
parameters are closely correlated with indigenous concepts, e.g. of soil
fertility. Where the parameters in the model require extensive laboratory
measurements (e.g. of micro-nutrients), model calculations can help identify
trends that are important for the sustainability of local practices in the long
run but may be less obvious to farmers in the short run. For example, farmers
may know that their traditional practice of burning vegetation before planting
makes plants grow better in the wet season (burning increases the short-term
availability of nutrients to the crops) but may not know that burning could lead
to an overall decrease in nutrients in the soil over several years or decades.
This could be revealed by a computer model based on soil analyses.
There is no intrinsic contradiction between the use of complex models and
indigenous knowledge; in many instances, these can be complementary. Both of the
above-mentioned books dealt with scientific, largely mathematical models which
would be beyond the understanding of anyone not specialized in this type of
modelling. If such scientific models are to be useful for farmers in southern
Mali or in the African drylands, they must be constructed not in isolation but
rather in close cooperation with the land users. Mathematics can be useful if it
can be made clear to the farmers and development workers what the outcomes mean.
Unfortunately, in neither of the two books were serious attempts made to bridge
the gap between indigenous knowledge and formal science.
Wolfgang Bayer (Dr Agr.)
Consultant in pasture and forage management and extensive animal husbandry, and
part-time lecturer in tropical forage management at the University of Göttingen
Rohnsweg 56, D-37085 Göttingen, Germany
Tel.: +49-551-485 751
E-mail: wb.waters@link-goe.de
PHOTOS:
Back to: top of the page | Contents IK
Monitor (8-1) | IKDM Homepage
Suggestions to: ikdm@nuffic.nl
© copyright Nuffic-CIRAN and contributors 2000.