“War made states and states made war” is a great slogan of Charles Tilly’s – but it is not universally true. War long predated the state. And it looks like the state is thriving while war is in perhaps terminal decline in much of the world. So it applies from about 5000BP to about 1945 or 1989. But that’s not bad coverage!
If you step back and ask what kind of theory is Tilly’s non-unilinear, relational, bellocentric theory, it looks very much like a special case of group selection theory.
The theory of group selection aims to account for the trend of upscaling in social cooperation: essentially, the less cooperative were selected out. Likewise, Tilly aimed to account for a particular upscaling in cooperation from the petty states, cities, and tribal chiefdoms of the early middle ages to the national states of European modernity. He invokes a similar mechanism: competition selected out the less cohesive.
(Meanwhile, Asian states in same period Tilly covered (medieval-early modern) went through a greater upscaling into large empires. Why? They were in striking range of nomadic horsemen. There was nowhere to hide from the fierce selection power of nomad geopolitics.)
As a zillion sci-fi stories have correctly pointed out, a further upscaling of cooperation to a world government is most likely when an alien military force appears.
Currently, in the belt going from the Sahara to Afghanistan, we are seeing a case of “war made clans and clans made war.”
In short, Tilly’s largely right, but his approach is form of group selection theory.