The African soil fertility ‘problem’ (I am thinking of dryland soils) is of course a management problem, as after many decades of expanding cultivation and grazing, the basic characteristics of virgin soils have been significantly altered nearly everywhere, or stand to be altered soon. Management is based on knowledge, which is fragmented. At least three levels can be discerned:

• Science-based knowledge, drawing on soil science and related natural science disciplines, which has enjoyed dominance since the beginning of the colonial period and has therefore led policy makers to search for technology-driven solutions
• Policy-makers’ and donors’ perceptions, linked to that of field professionals, which has been marked by top-down and generalist tendencies that result from attitudes obtained from educational institutions, the influence of influential stakeholder groups, and donors’ home constituencies
• Local peoples’ knowledge, which consists not merely in picturesque representations of the properties and potentials of local soils, inherited from the past (‘indigenous’ knowledge) but also in experiential and adaptive knowledge from project successes or failures as found relevant to their livelihood circumstances

Each of these crude categories has its own social ambiance. The first flourishes in universities and research stations, entangled with institutional structures and priorities and often lacking adequate ‘off-station’ inputs, often for want of resources rather than inclination. The second is driven by political targets and prejudiced in favour of grand scale interventions that attract publicity and funds. The third – insufficiently recognised – positions soil management as one component in a complex livelihood system where natural resources compete with wide-ranging livelihood objectives for the limited labour, skills and finance available.

It is only at the third level that knowledge properly confronts the complexity of local ecosystems, which have recently been characterised as ‘co-evolving human and ecological systems’ in the ‘Drylands Development Paradigm’. This level is also the only level at which the diversity issue is confronted on an everyday basis. It is at this level that well-known ‘success stories’ characterised as ‘area development’ (rather than project successes) have been worked out. There is a great gulf fixed between scientific knowledge patiently acquired from research at this level and the sweeping generalities promoted by the continental surveys and projections, and ruthlessly repeated in support of politically acceptable grand programmes in the soil fertility debate. Divergences between understanding obtained from macro- and micro-scale research should be a cause of concern. And such micro-scale research as has been undertaken is far too limited.

What is ‘success’? Given the current trends in food prices, fuel and other inputs, demographically-driven demand, urbanization, and climate change (or increasing variability), sustainable soil productivity is surely the only acceptable indicator of successful management. As such, it comes quite close to the perspective of a great many small farmers, who only ‘mine’ nutrients when their resources are constrained, and who are acutely aware of their need to pass on a productive asset to their heirs. Provided that the inheritance is assured, they invest – often with labour rather than with finance – in small-scale, intermittent, incremental  inputs over time.

In this context, the search for the ‘right’ policies continues, each with its own proponents. A question worth raising is whether the difficulties faced (so far) in hitting on demonstrably ‘successful’ strategies reflects a failure to come to terms with the fragmented and under-developed state of understanding of African soils management. Beyond the commendable use of participatory methods in projects (which pursue an external agenda) and a new emphasis on knowledge partnerships between farmers (or livestock herders), researchers, professionals and policy makers, two awkward concerns are:

• The near-universal popularity of a diagnostic-prescriptive framework for designing intervention and promoting change. This mode, inherited from colonial forbears and an unequal exchange between scientific and local knowledge, suggests that every intervention begins afresh, as if no-one had been there before. This cannot be so, after many decades of agricultural policies and interventions affecting most of Africa. It is a consequence of the nature of development projects – nothing yesterday, funded today, impact (and withdrawal) tomorrow. Is this shallowness acceptable, or does the diagnosis need to be positioned beyond expert opinion in a more sophisticated analysis of project precursors, policy impacts, and long-term trends (for example, in rural population densities, markets, technology transformation, ecological or landscape evolution)?  This is how local people see it. Their memories are often longer than those of the institutions that seek to turn their lives upside down! Projects should be positioned through long-term understanding of transition in the countryside, not only in environmental management but also in livelihood circumstances.
• Livelihoods approaches, although widely acknowledged to be relevant to soil management, are quite difficult to implement. How can development policy or project design deal with the possibility that investment in a bag of fertilizer may have to compete with the cost of taking a sick person to hospital? Agriculture is traditionally managed at national and donor level as a sector, but at the local level, no sector division is made. Investment decisions reflect such variables as education, attitudes, state of health, access to labour and knowledge, markets, social priorities, as well as financial resources. All these are embedded in a slow process of change that may influence how local people evaluate the prospects of technologies being promoted.

This may be a caricature of issues already familiar. But they are not always reflected, it seems, in policy debates leading up to grand programmes. Beyond the local scale, and the inspired action-research project agenda, there are methodological difficulties in scaling up temporal depth and systemic breadth, which remain as outliers in the policy debate, if recognised at all.

Mike Mortimore, Consultant
Drylands Research
mike@mikemortimore.co.uk