Understanding the Nature of Inference: Correlation and Causation
Professor and author Carlo Rovelli, Theoretical Physicist, Director of the Quantum Gravity Group, Centre de Physique Théorique, Université de la Méditerranée, Marseille, France
The physics of the time orientation of causation is more subtle than what it looks like superficially. For a long while it was reduced to a mere linguistic issue (a “cause” is just the term of a correlation that happens earlier, Hume). As emphasized by Russell, there is no time orientation in fundamental physics. But causation was later better understood as an essential notion in the context of an agent having choices, which after all is our own common context. This traces the time orientation of causation to the time orientation of agency. This realisation has been used by Nancy Cartwright to argue for the incompleteness of fundamental physics. However, agents are, in turn, physical systems and they must derive their time orientation from physics. Does this bring us back to square one? Remarkably, the answer is no because agency is a macroscopic notion and macroscopic physics is, contingently, time oriented, in a way compatible with the fundamental absence of an arrow of time. Hence the time orientation is causation is a consequence of the thermodynamic arrow of time, but in a peculiarly indirect way, that goes through the physical nature of the agents with respect to which causation is meaningful.
Please note that a moderated discussion with Professor Eric Winsberg of the University of South Florida will take place on Saturday, February 27, at 3 p.m. EST.