Inconsistency, Generic Modeling, and Conceptual Change in Science

Abstract

Traditionally inconsistency, as viewed from the perspective of logic, is held to be something detrimental in reasoning processes. From the perspective of the history of scientific development, though, inconsistency can be seen to play significant heuristic roles in the processes of conceptual change in science. I have argued in previous work that various forms of “model-based reasoning” specifically, analogical modeling, visual modeling, and thought experimenting—figure centrally in concept formation and change. In these forms of reasoning, physical and formal inconsistencies can serve as the basis for model revision in an effort to eliminate them, such as is the case when thought experimenting reveals an inconsistency in a representation. However, they also can be ignored in provisional models and the representations derived from them in the service of exploration and progress in a domain. That reasoning can be productive under these circumstances presents a problem for classical logic since once one has discovered an inconsistency in a model or representation all inferences from it are meaningless. Here we will consider a case of productive reasoning in concept formation involving inconsistencies: Maxwell’s construction of the electromagnetic field concept.