A Probabilistic Model of Causation
Robert Charles Koons · 2000
Abstract In this chapter, I will develop a quantitative, probabilistic model of causation, building on the work done in the preceding two chapters. As before, the desiderata for the model include securing the transitivity of causation and the statistical independence properties associated with the Markov properties. At the same time, I want to faithfully represent the full complexity of the relationship be tween causation and probability, including the possibility of causes with negative statistical correlation to their effects, the possibility of independent over determination, and the possibility of cases of the preemption of causality, in which positive correlation is not sufficient for causal connection.