Discussion of: 'Causal Diagrams for Empirical Research' by J. Pearl