What is clear is that becoming agriculturalists does not automatically mean “abandoning tribalism.” Most remaining tribal peoples are in fact agriculturalists. Even living in farming villages doesn’t involve abandoning tribalism. In other words, adopting agriculture doesn’t always result in the formation of a non-tribal “state.”
When a state does evolve (for whatever reason—the reason given by Carniero or any other), tribal identity is necessarily lost in the larger whole. Thus “complete dependence on agriculture” is not the issue. You can depend completely on agriculture and still be a tribe. It’s entering a larger, non-tribal “state” that costs you the tribe. My own theory about the origin of the state deals only with the bare beginnings: when food surpluses become substantial enough to need guardians, these guardians of the food eventually become controllers of the food and hence a ruling class.
(Mark Meritt’s masters, linked on this site in a few places already, provides a fairly comprehensive critique of Carneiro’s circumscription theory in the context of critiquing the ecological dynamics of Taker culture as a whole. It can be read at, The Unsustainability and Origins of Socioeconomic Increase. Mark has since posted an update to some of his thinking, The Social Full House: Circumscription and the Evitability of Complexity.
updated: 02 Oct 2004