This paper proposes to test the ethical acceptability of four styles of agricultural resource management: (1) contemporary industrial integrated systems agriculture, (2) modern industrial input dependent agriculture, (3) continuous traditional agriculture and (4) non-continuous (or swidden) traditional agriculture. The test of ethical acceptability is whether or not these styles of agricultural resource management embrace or are even compatible with that pattern of practical reasoning and interaction among ethical agents which we have independent theoretic grounds for preferring. The preferred sorts of (...) practical reasoning and interaction are those which we find operating in ethical theories which are strongly committed to letting the discretion of ethical agents construct what is right for them to do. Thus the discussion distinguishes several different strengths of constructivist ethics relating them to the work of John Rawls, Immanuel Kant and Onora O'Neill. Then it argues for the theoretic preferability of one particular strength of constructivist ethic. The paper winds up by arguing that only traditional continuous agriculture embodies the preferred sort of practical reasoning and interaction among ethical agents. Further, I argue that this is the only style of agriculture which can embody such reasoning and patterns of interaction. Thus, as we consider the role of agriculture in our plans of international development, we have one reason to try to favor traditional continuous agriculture. To do otherwise would ignore the ethical superiority of the practical reasoning and patterns of interaction of traditional agriculturalists. (shrink)