Penn Arts & Sciences Logo

Colloquium: Holly Andersen

Friday, December 9, 2016 - 3:00pm

Cohen Hall, Room 402

I will begin by providing a general introduction to the information-theoretic approach to causation, both its metaphysical commitments and its methodological consequences. This account of causation involves the claim that causal relata and relationships should be understood as patterns, in a technically defined sense, that are instantiated in the rich causal nexus, as construed in a manner similar to that of the Salmon-Dowe process account of causation. Against this background, we can ask about what I’ll call Laplace’s Pattern, the basic pattern that just is the description of every single ‘pixel’ in the causal nexus through time. A challenge to my view is this: isn’t Laplace’s pattern the only ‘real’ causal pattern, and everything else just a stand-in? Or, put another way, isn’t the real causal story still just the lowest microphysical level? I answer this challenge by drawing on some of Salmon’s own work and extending it further. Genuine causal relationships cannot be exhausted by Laplace’s pattern, I argue, because that would require that causation itself be a separate conserved quantity. This implies that the space of possibility for causation changes through time, undermining the Laplace’s Pattern challenge to genuine higher-level causation. I conclude by noting the interesting consequences this has for how to conceive of the time evolution of the universe.

Paper Title

Causation, Information, and Laplace’s Pattern