You feel a slight philosophical discomfort about this. You don’t like the idea of forced change, of intervention, being so integral to such a seemingly basic notion as causality.
Maybe it’s easier to think that Bayes nets without external data can only say “rain is strong evidence to wet” and “wet is strong evidence to rain” but not “rain causes wet” or “wet causes rain”?
Maybe it’s easier to think that Bayes nets without external data can only say “rain is strong evidence to wet” and “wet is strong evidence to rain” but not “rain causes wet” or “wet causes rain”?
Yes, a Bayes net is a representation of patterns of conditional independence in the data. On its own it says nothing about causes.