To provide a slightly different twist on John’s post below, one useful way to think about the relationship between violent rhetoric and violent action is to borrow from arguments about climate change. Very often, people engaged in debate over climate change either argue that a specific event (e.g. a cold winter) disproves or demonstrates the reality of climate change. But this is to misunderstand the debate. As I (as a non climate scientist) understand it, the scientific consensus about climate change suggests changes to average temperatures (and changes to the associated likelihood of certain weather events), but it is usually going to be next to impossible to tell whether any given event is ‘caused’ by climate change (it may simply be the result of random fluctuation). Testing arguments about climate change involves multiple data points and the usual problems of statistical inference etc.
Similarly, it is probably a bad idea to attribute any particular violent action to an overall climate of violent rhetoric without some strong evidence of a direct causal relationship. E.g., if the assassin had quoted some of the violent rhetoric that has been widely criticized as an inspiration, had listened to Michael Savage’s radio shows several hours a day or whatever, one would not be able to prove a causal relationship, but it would not be an unreasonable inference. There does not seem to be evidence of that sort in this case. John points to some evidence that is suggestive of a broader statistical relationship between violent rhetoric and attitudes towards violence. This is obviously much weaker than the kind of evidence that climate scientists have gathered pointing to global warming. But, to the extent that it does point to a possible relationship between violent rhetoric and violent action, it is to a probabilistic relationship. One can say that there is (moderate) evidence supporting the argument that violent rhetoric makes violent action more likely. But this does not and cannot show, in the absence of other evidence, that any particular violent action is the product of a general atmosphere of violent rhetoric.