Psychologist Robin Dunbar believes there are numerical ranges of "friend categories" that most efficiently support us, but also biological lower and upper limits to these categories. In others words, we are all wired with the same social limits.
What caught my interest is that Robin Dunbar believes these divisions may also be the basis for infantry unit structures, from squads to platoons, etc. The nomenclatures and weapons vary across the ages, but at least in West Eurasia, seem to follow similar social blueprints, roughly falling in numerical categories with high and low limits.
Of course, each unit can be reinforced or have detachments, and units were sometimes perfected by sexagesimal or decimal factors. It's also worth pointing out that specialized units (artillery, calvary, oarsmen, dragoons, archers, aviation, engineering companies) may be numerically structured for specific tasks.
This got me thinking about Harald Meller's paper looking at ceremonial weapons deposits at an "Unetice barracks" at Dermsdorf in Sömmerda District, Germany. There, we see what may be strangely reminiscent of modern units, complete with a social division of arms. Meller's ideas are by far, the most logical and intelligent interpretation of weapons depositions from the Bronze Age. They are military deposits, possibly of a ceremonial or religious natures. Practical is not impossible either. (everything other theory is rolled eye whites, foaming at the mouth, holding snakes and murmuring devil words)
If Dunbar is correct, then maybe we shouldn't be so surprised by Dermsdorf after all. For that matter, what precludes sufficiently-numbered Neolithic people from organizing war units this way?