Novel information-theoretic measures of causation in cognitive science The aim of this talk is to present and analyse three different information-theoretic measures of causation that have recently been proposed from within cognitive science, and to argue that they are best understood as offering a deflationary or instrumentalist account of causation.

The formal measures of causal strength or influence that these accounts provide are useful tools for the analysis of complex systems, but such tools cannot necessarily do the philosophical work that more traditional, metaphysically robust accounts of causation are expected to. It would thus be a mistake to draw any strong conclusions about topics such as causal emergence or mental causation from these novel measures alone.

Nonetheless, once this limitation has been recognised, the tools developed by these accounts can provide a valuable foundation for a formal philosophy of the special sciences, guiding our analysis of existing issues such as explanatory autonomy and levels of explanation.