Comment on R. L. Carneiro's Article on the Origin of the State: ‘The Circumscription Theory: A Clarification, Amplification, and Re-formulation’

Comment on R. L. Carneiro's Article on the Origin of the State:  ‘The Circumscription Theory: A Clarification, Amplification, and Re-formulation’
Author: Testart, Alain
Journal: Social Evolution & History. Volume 11, Number 2 / September 2012

The circumscription theory like other environmental or geographical explanations of the origin of the State, and Wittfogel's organisational theory, have two major flaws.

The first is that they just forget to ask: What is a State society? It is a society where the State has the legal monopoly of violence, according to Max Weber's well-known definition. Although one may refine this classical formulation (as I do, but it would be much too long to explain it here), there is nowadays a large consensus about the main idea it contains. Only American scholars (in social anthropology or archaeology, not in history or sociology) seem to ignore it. The State is not the same thing as social stratification, or population pressure, or large polities ... The State is something political. It means that you cannot take the law into your own hands: only the State (whether it be the res publica or the king) can inflict penalties, especially death penalties (hence the proverb found in pre-colonial African states: ‘the king has the monopoly of the knife’). To be a member of a State society means also that you are not allowed leading private wars: only the State can decide about war and peace. So what is a stateless society? It is a society where you can track down and kill the murderer of your son or father; you can do it with the help of parents, clan, lineage and so on. It is a kind of society where feuds are not only frequent; they are also considered as legitimate. This is obvious because there is no other way to get your cattle back if stolen, or to punish a murderer: you can only organize a party and, with using weapons and the sense of your rights, you try to do justice yourself. It is too long to explain here how these societies have certain customs and laws to limit outbreaks of violence. Stateless societies are not constantly at war, as Hobbes thought, but they are societies where feuds and private wars are frequent and normal. What is the main characteristic of this kind of society? It is the fact that each segment (generally, kinship groups, but this may be groups allied on another basis, e.g., on a local basis) of a Nation is independent from the others: each of these segments has sovereignty, with its own weapons at its disposal, using them when it decides to. The individual or the segment has political power in the full sense of the term including executive power. Now what does it mean when a stateless society becomes a State? It means that the several segments, into which a Nation is divided, lose this power. Why should they do so?

For a very simple reason. You retain political power as long as you have strength enough to maintain it. Now the strength of the clans, lineages, and other kinship or non-kinship groups, is variable, depending on the number of their members and their ability to form political alliances. Imagine now a powerful man who has many followers and bodyguards; the one who may even have a private army, since it is frequent in the ethnographical or historical records to find private armies composed of slaves. This man's total strength may be stronger than that of the most influential lineages, and if so, he may ask everyone not to make war without his consent, nor to judge or kill anyone without his own judgement, nor to use weapons outside of his armies. Doing this, he invented the State, he created one. There is nothing more to explain about this. What needs to be explained is: how is it possible that he found a sufficient number of followers and faithful dependants ready to obey him, and why should they be a better support than a kin? I treated this subject at length in my book La servitude volontaire (Testart 2004). The main points are as follows: 1) slavery is found almost everywhere among stateless societies (in ethnographical data), so there are grounds to think that slavery also existed (historically) before states; 2) slaves have a choice between rebelling (and be killed) or obeying; 3) those who obey may have more options, they may be faithful servants of their masters, and so partake something of their master's prestige and power; 4) masters have good reasons to prefer slaves to kin, because there is no risk of concurrency as in the case of the former; 5) for all these reasons, one of the main use of slaves – in stateless societies as well as in State societies – is military and political; 6) a man who has such
a manpower at his disposal has a lot of power in his hands, and is able to disrupt the society. Here I cannot develop other theses of the book which deal with other various forms of followers, with clients in the Roman sense (cliens versus patronus), with a specific form of dependency (the dependant being ready to die for his master, even uselessly, just to show how faithful he is), with geography and archaeology. Anyway, all this is about a way to the State, the way to a despotic State.

Another possible trajectory is when segments of the same Nation voluntarily give up their right to carry out feuds and thus create a kind of (or an incipient) democratic State. The American Indians of the Plains are the case in point, about which Lowie earlier, in a classic lesson, made outstanding comments – and I cannot understand how one can write on the topic of State origins without quoting him. Referring to how one of the so-called ‘secret societies’ among the Cheyennes acted as ‘police’ and in some cases kept the imposed penalties themselves (like a fine, which is to go to the State) rather than to the aggrieved party as reparation, he wrote: ‘They were the State in this case’ (Lowie 1948: 19).1

This first point may be summed up in two propositions. First, the emergence of the State is a political fact. And second, a political fact can only be explained by political facts. It cannot be explained by ecological factors or by considerations on the organisation of work, which are at best occasions of political struggles, catalysts, not causes. Strangely enough, Carneiro says in the course of his paper (p. 12) that environmental circumscription ‘accelerated’ the progress towards State but did not ‘initiate’ it; he also says (same page) that tight geographic constriction ‘is not absolutely essential to [State formation]’. I perceive these lines as confession that the circumscription theory does not and cannot give us the necessary cause – the one we are looking for.

The second flaw of such theories, both Carneiro's and Wittfogel's, is that they seem to be unfamiliar with archaeological invisibility. They quote repeatedly Mesopotamia and ancient Egypt.2 Sure they were linked with circumscribed environment and with water. But how do we know they were the first states? State is a political and sociological fact which, like almost all other sociological facts, is normally not recorded in archaeological remains. African pre-colonial interlacustrine states, like Burundi or Rwanda, without towns, without any writing, without monuments, even without palaces (the kings usually went round their kingdoms in camps) are totally invisible through archaeology – except maybe by funerals. We know they were states only owing to history and ethnography. Nowadays one has reason to think that the Gauls just before their conquest by Caesar were organised in local states. But we know this only because Latin texts mention taxes, military obligations, and the like: archaeology says nothing about this. What we know about Mesopotamia and Egypt around the end of the 4th millennium BC is that there were cities, writing, and monuments. This is altogether what Gordon Childe earlier called ‘civilisation’ (adding social, or class, stratification), but this should not be confused with the existence of the State. We know of hundreds of states in pre-colonial Africa, without civilisation. And there is absolutely no reason to think that before the unification of Egypt in the conventional date of 3200 BC, before partial unification a little later in Mesopotamia, there were not already dozens of micro-states in existence. And which may have existed for millennia.

To sum up my second point: I say that the theories I am criticizing have no idea of the complexity of the archaeological criterion of what a State is. They confuse the origin of the State and the origin of civilisation. Or worse: origin of states and origin of empires. These concepts are obviously not the same. And the bases needed for grounding a State or for grounding an empire are also different. To establish an empire, you need a good economic basis – and so it will always be important to discuss the natural environment when considering empires or big states. You also need something like a civilization, by which you can easily prove your ideological superiority. But to establish a State, you need much less. Only acceptance of a strong leader who will prevent crime and theft, or the people's mutual consent to give up the old traditional feuding customs.


1 In this article which was a Huxley Memorial Lecture, Lowie, writing at the end of his life, went much farther than in his book of 1927 on the origin of the State.

2 This is truer in the first articles of Carneiro (1970).


Carneiro, R. L.

1970. A Theory of the Origin of the State. Science 1970: 733–738.

Lowie, R. H.

1927. The Origin of the State. New York: Russel & Russel.

1948. Some Aspects of Political Organization among the American Aborigines. Journal of the Royal Anthropological Institute 78(1–2): 11–24.

Testart, A.

2004. La servitude volontaire. 2 vols. Paris: Errance.