Chiefdoms, and especially states, developed independently in several places around the world; but in most places, humans continued living in bands or villages. What made the difference? In 1970, Robert L. Carneiro identified three kinds of circumstance that seemed to foster political evolution. The first is environmental circumscription--fertile land more or less hemmed in by mountains, deserts, or water. Here, as agricultual intensification made land ever more scarce, defeat in war increasingly would leave the losers with nowhere to go to escape subjugation. Chiefdoms, and eventually states, would result. A second circumstance is resource concentration--productive resources, such as lakes or streams rich in seafood, so attractive that people try to stay near them. A third circumstance is social circumscription--being hemmed in not by geographical features but by other societies. 

The term "social circumscription" Carneiro borrowed from Napoleon Chagnon (1992). Chagnon had observed that population growth among the Yanomamo Indians of the Amazon Basin led to villages splitting and spreading deeper into the tropical forest around them. Due to such splitting, the average village size--around one hundred people--seemed fairly stable through time. At any given time, though, the more centrally located villages were the largest. Perhaps, Chagnon suggested, this was because, being surrounded by other villages (usually hostile), central villages were less able than peripheral ones to resolve internal conflicts by splitting. 

A time-honored idea held that human societies, like organisms, had a natural tendency to grow larger. (This analogy was elaborated especially in the 19th century by Herbert Spencer.) From Chagnon's work, though, it seemed that the natural tendency was for societies instead to stay about the same size, even when overall population was growing, due to splitting. Was it possible that the tendency to resolve social conflict by splitting was in fact a deep and universal propensity that had had to be suppressed before large societies ever could evolve in the first place? Could inhibition of splitting be the key to human social evolution? Suppose it was. Under what conditions would splitting be easy, and under what conditions would it become difficult? It seemed to me that the most important factor would be the presence or absence of opportunities for geographic expansion. If a growing population was surrounded by rich, unoccupied territory, it would expand easily into that territory; but it would do so Yanomamo-like, and the average society size would remain nearly constant due to splitting. Increase in this average size would be expected only when the opportunity to expand was somehow inhibited. Inhibited expansion would lead to inhibited splitting; if societies could no longer split fast enough to offset population growth, larger societies-- chiefdoms, states, empires--eventually would be forged, and culture would have to be transformed accordingly. What kinds of conditions would cause this process to unfold? The very circumstances Carneiro had identified: environmental circumscription, social circumscription, and resource concentration. 

It proved possible to formulate mathematical definitions for inihibition of both geographical expansion and political splitting. The assumption that splitting would not be inhibited until expansion was inhibited proved fruitful, and led to an exact mathematical theory of the relationship between population density and political evolution. I began presenting these ideas at meetings of the American Anthropological Association in the mid 1980's; my first article on the subject appeared in 1988; and a book, A Scientific Model of Social and Cultural Evolution, came out in 1995. I continue working along the same lines today. I like to think it is the richness of the ideas that has kept me absorbed so long; but I'm afraid another reason is my slowness at figuring out even rather elementary mathematical implications!