Varieties of scientific knowledge and their contributions to dealing with policy problems: A response to Richard Nelson’s “The sciences are different and the differences matter”

The substantial differences between the kind of knowledge characteristic of classical mechanics and that found in much 20th century physics, biology and the human sciences have made the belief that there is a single model of scientific knowledge best represented by Newtonian physics increasingly untenable. The more that sciences deal with phenomena that are complex in terms of their internal structures, are partially constituted by their connections to their contexts, and whose causally influential processes are subject to endogenous change, the more important it is to recognise the diversity of the kinds of knowledge they produce and their varied connections to material and social technologies. Theories and models that differ in their contextual independence, complexity of internal structures and specificity of causal processes contribute in different ways and under different conditions to the resolution of policy issues. Applying causal models of what worked in other situations to policy problems often fail because key contingencies not included in such models have changed and/or their success has itself altered actors’ perceptions and preferences.