Mainstream agricultural research has focused primarily on technical and biological aspects and is aimed at controlling or manipulating nature through the use of external inputs, such as
agricultural chemicals or super seed. In developing countries, the results of this research have benefited some resource-rich farmers in well-endowed areas, were suitable to only a limited
extent for poorer farmers in the more favourable areas, and were - in most cases - completely inappropriate for small-scale farmers in marginal areas, e.g. in the mountains or the drylands.

The findings of a Nigerian case study discussed in this paper indicate that the notion of wives of leisure is really not applicable to most women in Nigeria, as women have always worked. Even those in purdah engage in income generating activities within the confines of the compound. It is therefore wasteful to continue to by-pass or displace women in development. Selective mechanisation of difficult processes in agriculture could become a useful method of integrating women into, and enhancing their contributions to, development.