There is a temptation in the public discourse, I think, to reduce this to monocausal explanations, and clearly, religion can be one of them. There are so many people who are religious, and virtually none of them become violent. Many of the people who become violent, for instance, are not religious or don't adhere to, say, any ideology of any particular form. Understanding that the predictors are multi-faceted and combine in different ways in different cases is critically important.
I think this is where some of the structural work that Dr. Perry is doing can help us. It can help us to identify some people who may be at higher risk, but then ensuring, for instance, that when we're talking about youth—and as any of us who have children of our own or are familiar with youth know, youth do things that we would prefer them not to do—we get them the right intervention and the right help, rather than necessarily and immediately criminalizing viewpoints that in and of themselves we find objectionable but that aren't necessarily criminal.