The Bounds of Possibility

THE DEPTH-STUDIES WHICH ANTHROPOLOGISTS MAKE are a dimension in our perception of the human condition. In its ecology, its social life, its beliefs and values and artistic achievements, each society is a microcosm. It can be studied, evaluated and compared to other microcosms in an attempt to gauge the bounds of possibility. In this way we can enlarge our understanding of social systems, past and present, our own and others. Is government necessary? Is law? Or religion? And, if so, in what forms? In this article I propose to discuss some interesting points raised by a recent anthropological book. Lucy Mair's new Pelican, Primitive Government, surveys the political life of some East African societies, several of which have been regarded by the anthropologists who studied them as being without government. Dr. Mair's contentious point is that the "ordered anarchies" described in such books as The Nuer and Tribes Without Rulers, are not really anarchic at all. Government is ubiquitous. This, and one or two other cavils, aside, her book is a most useful introduction to its subject; and we can only hope that Penguin Books will go on to balance their series on history, archaeology, philosophy and psychology with one on anthropology.

Dr Mair divides her book into three parts: "Government Without the State", "African States" and "Primitive Government and Modern Times". There is also a short introduction in which she discusses her conceptual framework. The reading list, with which Primitive Government concludes, is not quite so complete as it should be. Tribes Without Rulers is omitted, though two of the essays from it are cited in the text. If African Political Systems deserves its place in the reading list, then so, too, does the later collection. And I think Max Gluckman's Custom and Conflict in Africa, based on his Third Programme talks, should have been included.

The title of Dr Mair's book, and her treatment of the "ordered anarchies" (a term she does not apply to them herself), as stateless but governmental, raises problems of definition. What is a "primitive" society? What is meant by "government", "the state", "law"?

Societies are primitive, says Dr. Mair,l when they are small-scale, lack writing, and have only rudimentary technology and forms of government. She rightly stresses that to describe a society as primitive is not a reflection on the innate character of those who belong to it. Ralph Piddington2 uses these criteria, but adds the importance of kinship and locality in determining social relations and the lack of specialisation. Leonhard Adam3 denotes as primitive, the societies not included within modern European civilisation or the great Oriental civilisations. The definition I find preferable in that of Leslie White.4 He confines the term to societies organised on the basis of kinship, and ecologically dependent upon human energy alone. The harnessing of non-human energy — plants and animals — constitutes the Agricultural Revolution and lays the ecological foundation of civil society, in which kinship diminishes in importance. Centralised political institutions and class-divisions appear. In White's sense, some of the societies dealt with by Dr. Mair are not

However crude and ineffectual primitive cultures were in their control over the forces of nature, they had worked out a system of human relationships that has never been equalled since the Agricultural Revolution. The warm, substantial bands of kinship united man with man. There were no lords or vassals, serfs or slaves, in tribal society. In social ritual one man might make obeisance to another, but no one kept another in bondage and lived upon the fruits of his labour. There were no mortgages, rents, debtors, or usurers in primitive society, and no one was ever sent to prison for stealing food to feed his children. Food was nat adulterated with harmful substances in order to make money out of human misery. There were no time clocks, no bosses or overseers, in primitive society, and a two-week vacation was not one's quota of freedom for a year.
Crude and limited as primitive cultures may have been technologically, and wretched and poor as life may have been far many — but far from all — people living in tribal organisation, their social systems based upon kinship and characterised by liberty, equality, and fraternity were unquestionably more congenial to the human primate's nature, and more compatible with his psychic needs and aspirations, than any other that has ever been realised in any of the cultures subsequent to the Agricultural Revolution, including our own society today.
—LESLIE A.WHITE: The Evolution of Culture, 1959.

primitive at all, but represent early forms of civil society (this would be true of the "African States" dealt with in Part Two. The Part One societies would, I think, be transitional between primitive and civil. There is, however, no unanimity on the use of "primitive"; Dr. Mair's definition is conventional, but White's has the merit of denoting a logical class and thus is helpful in arranging social systems in some kind of evolutionary sequence.

"Government" is another term whose boundaries are amorphous. To anarchists, of course, it has emotional connotations, but in considering the use of a word by a writer who defines what his meaning is, attention should be directed to the referent, not to the emotional undertones. Dr. Mair defines government functionally: "What then does government do? It protects members of the political community against lawlessness within and enemies without; and it takes decisions on behalf of the community in matters which concern them all. and in which they have to act together."5 I should say that in such societies as the Nuer, people protect themselves and regard decisions as binding only the person who makes them. What Dr. Mair considers governmental functions could pertinently be interpreted as co-ordination without compulsion, the dream of the anarcho-syndicalists.

Dr. Mair's novelty is her conceptual separation of government and state while maintaining the ubiquity of the former. In African Political Systems, M. Fortes and E. E. Evans-Pritchard define the state by presence of government. In dividing the political systems with which they are concerned into two categories, they write:

One group … consists of those societies which have centralised authority, administrative machinery, and judicial institutions — in short, a government — and in which cleavages of wealth, privilege, and status correspond to the distribution of power and authority…. The other group … consists of those societies which lack centralised authority, administrative machinery, and constituted judicial institutions — in short which lack government — and in which there are no sharp divisions of rank, status, or wealth … Those who consider that a state should be defined by the presence of governmental institutions will regard the first group as primitive states and the second group as stateless societies.6

In Anarchism, Paul Eltzbacher gives a definition of the state which, in essence, conforms to what Fortes and Evans-Pritchard have said about government:

Some inhabitants of a territory are so powerful that their will is competent to affect the inhabitants of this territory in their procedure, and these men will have it that for all the inhabitants of the territory, for themselves as well as for the rest, the will of men picked out in a certain way shall within certain limits be finally regulative. When such is the condition of things, a State exists.7

In The Evolution of Culture, White makes essentially the same points, but reminds us that government has two faces:

Civil societies are characterised by a number of diverse parts and specialised functions, on the one hand, and a special mechanism of co-ordination, integration, and control, on the other. This special mechanism should have a name, and we have decided to call it the state-church. We do this because this mechanism always has both a secular and civil aspect and an ecclesiastical aspect; state and church properly designate aspects of this co-ordinative, integrative mechanism rather than separate entities.8

Mikhail Bakunin, too, understood the dual nature of government;9 and Dr. Mair discusses the relation of divinity and ritual to kingly office.10

"Law", again, is a semantic nightmare. At one pole, there is A. R. Radcliffe-Brown's well-known definition: "social control through the systematic application of the force of politically organised society"11 — a definition he borrowed from the jurist Pound. In Tribes Without Rulers, David Tait states that law in this sense does not exist among the Konkomba.12 "In a strict sense Nuer have no law"13 says Evans-Pritchard, though he goes on to define the term in a sense he considers appropriate for this people:

We speak of "law" here in the sense which seems most appropriate when writing of the Nuer, a moral obligation to settle disputes by conventional methods, and not in the sense of legal procedure or of legal institutions. We speak only of civil law, for there do not seem to be any actions considered injurious to the whole community and punished by it.14

At the other pole, is Malinowski's definition of law: "The rules of law stand out from the rest in that they are felt and regarded as the obligations of one person and the rightful claims of another. "15 Such definitions as his are unsatisfactory, for they blur the differential classes of sanction existing for the varied rights and obligations of members of a society.16

"Law" in Malinowski's sense exists in every society; in Radcliffe-Brown's sense it exists only in some, e.g., in the African states discussed in Part Two of Primitive Government. Dr. Mair understandably refrains from offering a new definition; she sensibly confines herself to discussing the ways in which disputes are in fact settled in the societies with which she is concerned.

Before leaving the semantic morass in which concepts of "the state", "law" and so on are embogged, we can note that the classical anarchists were agreed in their negation of the state, in the sense in which Eltzbacher defined it, but were divided on the issue of law. Godwin, Stirner and Tolstoy were anomistic, they negated law. Proudhon, Bakunin, Kropotkin and Tucker were nomistic, they affirmed law, though without being unanimous on its content.17 Law or not, therefore, the Nuer, the Konkomba and the rest can accurately be called "ordered anarchies".

I do not accept Dr. Mair's view that government can exist independently of the state, or, rather, I think that the term government is best jettisoned in describing how co-ordination is effected in societies without the state form. In reference to the Tiv, Laura Bohannan wrote in Tribes Without Rulers:

… a segmentary system of this sort functions not despite but through the absence of an indigenous concept of "the political". Only the intricate interrelations of interests and loyalties through the interconnection of cultural ideology, systems of social grouping, and organisation of institutions and the consequent moral enforcement of each by the other, enables the society to work. To isolate part of it as "political" may be correct, insofar as our definition of the political is concerned, but to do so robs the society of those very factors which endow it with vitality.18

Dr. Bohannan's approach seems to me to be the most fruitful, and the one which is most faithful to the reality of the stateless societies. In the remainder of this essay, I will take a look at the principles underlying the everday conduct of life in some of these societies.

The Nuer, a people of about 200,000 in the Southern Sudan, were the subject of Evans-Pritchards' classic study, The Nuer. He is the first social anthropologist to have given a detailed account of what are now known as segmentary lineage systems. The characteristic of such systems is that the parts are in conflict and opposition; yet the whole does not disintegrate. Why this seeming paradox should be so, I now hope to show.

The Nuer consist of several tribes, each of which segments, according to circumstance, into smaller and still smaller sections. Thus the Lou tribe segments into the Mor and Gun primary sections; Gun into Rumjok and Gaathal secondary sections; Gaathal into Leng and Nyarkwac tertiary sections. In turn, the tertiary sections segment into Nuerland's basic units, the village communities. Mor and Rumjok also segment in this way; and what is true of the Lou is true of every Nuer tribe.

Corresponding to the segmentation into territorial sections, and providing a moral or ideological framework for it, is a segmentation into lineages. (A lineage is a social group whose members are united by an ideology of common descent. A number of them form a clan, the founder of which is believed to be the common ancestor of the various lineage founders. Lineages, themselves, segment into smaller lineages.) Associated with the tribe is a clan. The clan segments into maximal lineages; maximal lineages into major; major into minor; minor into minimal. Each maximal lineage is associated with a primary section; each major with a secondary; each minor with a tertiary; each minimal with a village community. Although there are several clans within a tribe, one is regarded as dominant, and it is this clan which is associated with the tribe and its lineages with the tribal sections. "Dominance" has a mythological referent, and does not imply any ruler-subject relationship.

A tribe, and each of its sections, has a distinctive name, a common sentiment and unique territory. Indeed, in a sense, tribal sections are like tribes. What distinguishes them is the manner in which intergroup relations is conducted. Between sections of the one tribe, feuds are fought, and compensation is paid for homicide and other torts. Between tribes, however, wars are fought, and compensation is not due. The system of arbitration by which disputes between sections of the one tribe can be settled with minimum blood-shed, has never been extended to regulate intertribal relations.

The tribe does not disintegrate into the sections of which it is built up. For one thing, the division into territorial sections is cut across by many bonds of kinship. The clan, and each lineage within it, is exogamous, i.e., its members must take their spouses from other clans. Thus everyone is necessarily kin to many Nuer of other clans, and the presence of such kin in other villages, other tertiary sections, other secondary and primary sections, constitutes a network binding the tribal sections together. Too great a development of intratribal hostility is thus inhibited. Indeed, Evans-Pritchard compares the cross-cutting bonds to elastic bands which stretch out in times of injury by one Nuer to another, only to pull the opposing fractions, together again.

Ritual is also a mechanism of cohesion. Persons belonging to groups between which there is a blood feud cannot eat or drink with one another. Social relations are severed. This is naturally a further incentive to mend the breach. In negotiating settlements, a ritual officer, the "leopard-skin chief", plays an important role in mediating, though he cannot impose his will on the parties.

Finally, the tribe coheres because the system is such that sections aligned against one another on one occasion will be aligned with one another at some other time. Thus two villages of the same tertiary section may fight, but will close ranks against a village of another tertiary section. Two tertiary sections will fight, but they, too, close ranks against a tertiary section of another secondary section. And so on, until the whole tribe unites as one unit in the event of any of its parts becoming involved against a part of another tribe.

The ecological framework, within which the Nuer are born and live out their days in glad anarchy, necessitates the kind of balance I have described, for in the dry season the village communities must migrate to the rivers where water and pasture exist for their cattle. In migrating, the land of other sections must be crossed, and this is why no part of the whole can be an island entire of itself.

The Tallensi of West Africa19 also exhibit a segmentary lineage system, in which the cross-cutting network of kinship ties inhibits the whole from disintegrating into its component parts. Again, the equilibrium is added to by the balance of like units for or against one another according to circumstance. Again, ritual is important in helping to stabilise intergroup relations through giving people an incentive to reconcile the offended parties. Ritual, however, does not operate in quite the same way among the Tallensi as among the Nuer.

Homicide is considered a sin against the ancestors and the Earth, as well as an offence against the victim's group. To expiate the deed, the families of killer and killed must offer sacrifices to the ancestors and the Earth; unlike the Nuer, compensation is not paid over to the victim's kin as bloodwealth. And each year, a cycle of festivities takes place in Taleland, a cycle in which each group has an indispensable part to play. All conflicts and disputes must be set aside, for otherwise the people would cease to prosper. Thus the common good is symbolised in ritual, and sectional interests are transcended.

Not all of the politically uncentralised societies of Africa are segmentary lineage systems. I would therefore like to describe how the Ndembu of Northern Rhodesia manage to cohere in spite of, or perhaps because of, the absence of any dominating central authority. The Ndembu are organised in villages, each consisting of men who are matrilineally related. The principle of matrilineal descent (descent traced through females) contradicts the principle of virilocal marriage (women live in the village of their husband), and tension is thereby inherent in the Ndembu system. Men grow up in their father's village but have their rights to property in their mother's. Tension is reflected in, for instance, the extreme instability of marriage.

The Ndembu have been studied by V. W. Turner,20 who makes a detailed examination of how conflict and cohesion are promoted by the dominant structural principles of their society. He uses the term "social drama" to denote the procession of events in which the structural contradictions are reflected in particular instances. Four steps characterise the social dramas. First, occurs a breach of one of the norms. Secondly, conflict mounts until the whole village is aligned on one side or the other. Thirdly, aghast at the split opening up in the village, an attempt is made to correct the balance and restore harmony. Finally, either the breach is mended or it is not. If reconciliation is effected, then the conflict is resolved, but only temporarily, for the contradictory principles inevitably give rise to further disputes. If reconciliation fails, the opposing factions split apart and one founds a new village. Here, too, harmony will be short-lived, for the principles on which Ndembu society is based are in contradiction always.

The culturally-approved solution to inevitable conflicts is the parting of the opposed factions, a never-ending process of fission. And, as among the Tallensi, ritual beliefs and observances function to mend breaches (sometimes) and to restate common values.

The Nuer, the Tallensi, the Ndembu are not representative of all politically uncentralised African societies. The ways in which the individuals born into them handle their problems of social relationships without the intervention of the agents of some centralised institution, gives us an insight into the practicability of decentralism. Which is not to assert that Nuer or Tallensi or Ndembu principles of organisation would be viable for any save the Nuer and the Tallensi and the Ndembu. It is our imaginative grasp of the human condition, our comprehension of the bounds of possibility, which is enriched and furthered in understanding how such societies function; there is no practical lesson for us. Indeed, the whole primitive or near-primitive world is in swift transformation to something else.

I have departed from Dr. Mair's frame of reference in discussing the Tallensi and Ndembu, for she limits herself only to East Africa. I have described the shell, not the kernel; the structural principles and ritual sanctions of the Nuer and Tallensi and Ndembu, not what is feels like to be a Nuer or Tallensi or Ndembu. Empathy can only be truly conveyed by those who have steeped themselves in the life of a primitive people, and, lamentably, not all who have shared the life of primitives have cared to share that experience with their readers (one who has is W. R. Geddes, whose Nine Dayak Nights is a more valid guide to the charm of anthropology than a dozen ordinary monographs). And I have discussed only decentralist societies, for the African states of which Dr. Mair writes are infinitely less interesting than, for instance, the wonderfully anarchic Nuer. Listen to Evans-Pritchard on these Sudan negroes:

The ordered anarchy in which they live accords well with their character, for it is impossible to live among Nuer and conceive of rulers ruling over them.

The Nuer is a product of hard and egalitarian upbringing, is deeply democratic, and is easily aroused to violence. His turbulent spirit finds any restraint irksome and no man recognises a superior. Wealth makes no difference. A man with many cattle is envied but not treated differently from a man with few cattle. Birth makes no difference. A man may not be a member of the dominant clan of his tribe, he may even be of Dinka descent, but were another to allude to the fact he would run a grave risk of being clubbed.

That every Nuer considers himself as good as his neighbour is evident in their every movement. They strut about like lords of the earth, which, indeed they consider themselves to be. There is no master and no servant in their society, but only equals who regard themselves as God's noblest creation.21

Compare this with the state societies, of which Dr. Mair notes:

Another characteristic which is shared by rulers, great and small, in East Africa is that they are, or have been when they were independent, privileged to inflict punishments on their subjects of a nature which would not be tolerated in the "anarchic" societies, and for actions which would not be considered offences in the relations between other people.22

The transition from primitive society to civil society is a transition from primitive anarchy to the servile state.

1. Lucy Mair, Primitive Government, 1962, Harmondsworth: Pelican Books, 7-9.
2. Ralph Piddington, An Introduction to Social Anthropology, Vol. 1, 1962, Edinburgh: Oliver & Boyd, 5.
3. Leonhard Adam, Primitive Art, 1954, Harmondsworth: Pelican Books, 26-7.
4. Leslie A. White, The Evolution of Culture, 1959, New York: McGraw-Hill, 367-8
5. Mair. 16.
6. M. Fortes and E. E. Evans-Pritchard (eds.), African Political Systems, 1950, London: OUP, 5.
7. Paul Eltzbacher, Anarchism, 1960, London: Freedom Press, 20.
8. White, 303.
9. The Political Philosophy of Bakunin, 1953, Glencoe: The Free Press, 207-8, 346.
10. Mair, 214-33.
11. A. R. Radcliffe-Brown, Structure and Function in Primitive Society, 1952, London: Cohen & West, 212.
12. David Tait, "The Territorial Pattern and Lineage System of Konkomba" in Tribes Without Rulers, 1958, London: Routledge & Kegan Paul, 186-7.
13. E. E. Evans-Pritchard, The Nuer, 1940, London: OUP, 162.
14. Evans-Pritchard, 168.
15. Bronislaw Malinowski, Crime and Custom in Savage Society, 1926, London: Routledge & Kegan Paul, 55.
16. cf I, Schapera, "Malinowski's Theories of Law" in Man and Culture, 1957, London: Routledge & Kegan Paul, 139-55.
17. Eltzbacher, 186-91.
18. Laura Bohannan, "Political Aspects of Tiv Social Organisation" in Tribes Without Rulers, 65-6.
19. M. Fortes, "The Tallensi" in African Political Systems, 239-71.
20. V. W. Turner, Schism and Continuity in an African Society, 1957, Manchester: University of Manchester Press.
21. Evans-Pritchard, 181-2.
22. Mair, 197.