Reflections on the Asymmetry of Causation

Abstract

The asymmetry of causation is arguably the most important of a cluster of temporal asymmetries that characterize our experience of the world. Early science treated causation as the fundamental ordering relation of the world, but when Newton replaced causes with time symmetric laws of evolution and Russell argued that causation was a folk notion that should be eliminated from mature science, the status of causal notions became contested.  In the last few decades two developments have shed new light on the asymmetry of causation: clarity in the foundations of statistical mechanics, and the development of the interventionist conception of causation.  In this paper, I ask: what is the status of the causal arrow, assuming a thermodynamic gradient and the interventionist account of causation? I find that there is an objective asymmetry rooted in the thermodynamic gradient that underwrites the causal asymmetry. Along a thermodynamic gradient, interventionist causal pathways – scaffolded intervention-supporting probabilistic relationships between variables – will propagate influence into the future, but not into the past. The reason is that the present macrostate of the world, in the presence of a low entropy boundary condition, will screen off probabilistic correlations to the past.  The asymmetry, however, emerges only under the macroscopic coarse-graining and that raises the question of whether the arrow is simply an artefact of the macroscopic lenses through which we see the world. The question is sharpened and an answer proposed.