The most important force behind the process of continual change that characterises human reality is agency. An agent is not a passive recipient that is destined to be influenced by the dynamics of the world. He is the constitutive part of the world, the very dynamics that characterise it. Agency means a freedom to construct and create one's own life. It stands for independence in constructing one's own reality. Attempts to monopolise this process are against agency. They stop the movements of invention. In this article the author tries to examine whether economic poverty research allows for the development and preservation of agency in Africa. Is such research adequate to describe the situation in Africa? Can it increase the agency of Africans? The article describes the history of poverty in Africa and discusses the issues of power relations that are the background of economic poverty research. Its first conclusion is that the economic poverty studies undertaken by international organisations are only partially adequate for a proper description of poverty in Africa. This is because Africans only partially underwrite their basic presuppositions. Despite this fact there are many features of African poverty that are not taken into account within this research. Agency in Africa is therefore partially undermined by economic research. Nevertheless attempts to open up the economic poverty research so as to allow Africans to express their agency have been made. More intensive, case-specific studies might, to a certain degree, broaden the scope of economic poverty research. It is however argued that as long as such locality specific research needs to be incorporated into the established development orthodoxy, thepotential of increasing the agency will not be realised. A locality specific knowledge will always be distorted by the "universal" development discourse. It must be concluded that the agency in Africa does not have much chance of developing within the North Atlantic economic body of knowledge. Africans can be in control of such research only as far as they agree with its presuppositions.
Read
