I have been taught that Mill's methods for identification of causes (agreement, difference, etc.) only apply when we can define our universe of possible causes very strictly (i.e., when we know all the possible causes) and when we need to identify just one cause. Thus, complexities like mixed causes and models with more than one cause aren't suitable to be treated with Mill's methods.

My question is: what are some alternatives to these methods that can deal with more complex situations of causality?

  • Science does not use Mill's induction method. On the philosophical issues regarding induction, you can see The Problem of Induction and Inductive Logic. – Mauro ALLEGRANZA Mar 10 '16 at 8:31
  • Thanks for the links, Mauro. Regarding to the use of the methods in science, at least in political science they're widely used, though with certain limitations. That's why I want to go a bit further and find some kind of rules like Mill's. – numberfive Mar 10 '16 at 17:06
  • 'widely used' is overstating things. The most frequently used theory of causal inference in political science, sociology, epidemiology, etc. is counterfactual: either as Pearl's graphical framework (e.g. Pearl 2000) or equivalently as Rubin's potential outcome framework. The qualitative methods community seem to like Mill (and related techniques like QCA), but are in this respect, a minority. – user19742 Mar 10 '16 at 19:41

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.