ABSTRACT

Many, if not most scientific theories of causation are given in terms of changes. The idea can be expressed in an astonishingly different number of ways, but it all comes down to saying that a change in x in some sense produces a change in y. In medical research it is virtually an article of faith that one can only begin to demonstrate causality if one "intervenes" by producing changes. The definition of an intervention is that the researchers cause a particular change, rather than that change having been produced by some other causal law.