The researchers should be open to the idea that the main problem is the boring, standard idea that violent criminals have little empathy for their victims, and

The researchers should be open to the idea that the main problem is the boring, standard idea that violent criminals have little empathy for their victims, and little impulse control. Even so, perhaps having little empathy and little impulse control is not the root, but actually a byproduct of something else. A programming analogy would be like this is a bug in their code, and our current method of dealing with it is to kill their process or filter their "standard out" to avoid/ignore the unpleasant buggy output. What if instead, we could come to comprehend a deeper architecture problem that we could then prevent or fix by refactoring?