Underlying stories: The structure of causal talk in chemistry and in everyday life

Chemistry uses a concept of causation continuous with our commonsense concept. Moreover, a concept of causation is essential for chemical projects.;I consider problems with regularity and counterfactual accounts of the concept of causation. I suggest an alternative, the underlying story account, which better avoids these problems. In everyday causal talk, "A caused B" means there is a suitable underlying story, a description of A and B and how they are connected. Some causal connections, paradigm causal relations, are primitives needing no further underlying story to support their causal status. Others are chains of paradigmatic causal links.;Chemistry grounds causal processes in underlying stories. The causal status of mid-level relations depends on stories about lower-level causal relations which are chemically basic, attractions and repulsions. More causings in chemistry are chains because chemists feel a pressure to analyze chemical phenomena in terms of these two basic relations. Although attraction and repulsion are relations unanalyzed by chemistry, chemistry accepts that the causal status of attraction and repulsion depends on further stories physics provides in terms of causal relations which are truly unanalyzable primitives. One ingredient in chemical stories resists being recast in terms of attractions and repulsions, thermodynamic regularities. However, chemists are committed to the existence of causal stories about how the structure of chemical systems brings about these regular patterns, and look to statistical mechanics to provide that causal story. As statistical mechanics is not yet a completed theory, parts of the expected causal story remain black-boxed.;Causal talk in chemistry cannot be excised. Replacing causation with mathematical regularities might improve prediction, but would be at odds with explanatory projects and leave descriptive projects involving new measurement techniques largely unmotivated. Replacing causation with unifying patterns would force chemists to accept as explanatory certain unifying patterns which chemists judge are not explanatory at all. To advance chemical projects of description, prediction, control, and explanation, chemists rely on a concept of causation which looks for structural connections in the world.