ABSTRACT
For simplicity, we will confine ourselves to the simple causal network illustrated in Figure 13.4. Our problem is the following. We have a disease Z which is known to have two indeterministic causes X and Y. It would seem sensible in such a case to try to avoid Z by eliminating through our actions at least one of X and Y, and preferably both. It would seem to be a good strategy for doctors to advise such a course of action. However, as the pregnancy, contraceptive pill, thrombosis example shows, this advice would not always be correct. Can we then formulate some mathematical condition on X, Y and Z such that the “common-sense” advice of eliminating at least one of X and Y is in fact correct advice?
