A third is that self-control can be adjusted adaptively according to its costs and benefits in the local environment. Self-control, after all, is not an unmitigated good. The problem with having too much self-control is that an aggressor can use it to his advantage, anticipating that you may hold back from retaliating because it’s too late to do any good. But if he had reason to believe that you would lash out reflexively, consequences be damned, he might treat you with more respect in the first place. In that case people might adjust a self-control slider according to the dangerousness of those around them.

At this point in the story, the theory of the Civilizing Process is incomplete, because it appeals to a process that is endogenous to the phenomenon it is trying to explain. A decline in violent behavior, it says, coincided with a decline in impulsiveness, honor, sexual license, incivility, and boorishness at the dinner table. But this just entangles us in a web of psychological processes. It hardly counts as an explanation to say that people behaved less violently because they learned to inhibit their violent impulses. Nor can we feel confident that people’s impulsiveness changed first and that a reduction in violence was the result, rather than the other way around.

But Elias did propose an exogenous trigger to get the whole thing started, indeed, two triggers. The first was the consolidation of a genuine Leviathan after centuries of anarchy in Europe’s feudal patchwork of baronies and fiefs. Centralized monarchies gained in strength, brought the warring knights under their control, and extended their tentacles into the outer reaches of their kingdoms. According to the military historian Quincy Wright, Europe had five thousand independent political units (mainly baronies and principalities) in the 15th century, five hundred at the time of the Thirty Years’ War in the early 17th, two hundred at the time of Napoleon in the early 19th, and fewer than thirty in 1953.