The sustainability of smallholder agriculture in the drylands of West Africa is a topic of long-lasting concern. This paper is focused on the scientific assessment of nutrient (mis)management affecting the long-term productivity of cropland. "Nutrient mining" is seen as a major mechanism for land degradation in the region. Two models will be the focus: that tied to the rangeland-to-cropland metric and the nutrient balance model. These tools of assessment follow a long history of environmental assessment and modeling that purposefully abstracts from the heterogeneity of the same farming practices that modeling efforts seek to assess. Abstraction is always part of modeling but the form of abstraction described here ignores any detailed understanding of farming practices and in so doing, demands very little information about farming practices. In this way, models serve as both the symptom and cause of continued ignorance of farming practices by most agronomists and soil scientists working in the region. The variables used in these models will be described and how these same variables are strongly influenced by the variation of farmers' practices will be discussed. In addition, it will be shown that needless model abstraction results in research outcomes that provide very little insights to guide land policy formulation and agricultural extension. In short, these models point to the ultimate resource limits of the regional agricultural system but provide few insights of what could be changed to reduce rates of land degradation. The paper concludes by presenting alternative approaches that through their engagement with the diverse positionalities of farmers, provides insights useful for the development of effective land policies.