There is probably considerable common-sense agreement among those interested in what is meant by a stateless society. Brief definition of the concept does little that is positive to increase understanding, but it may focus disagreement and so lead to clarification. To cover data so varied, a definition must be abstract and employ terms which are themselves the subject of controversy.
Positive definition is desirable, yet we will start out with an etymologically negative concept. Most of the current alternatives for defining a stateless society, such as acephalous, uncentralized, or anarchic, are equally negative; for example, Middleton and Tait (1958) called a book on the subject Tribes Without Rulers. Furthermore, “decentralized” suggests the distribution of something from the center, which in the present case does not exist.
What a stateless society is ultimately depends upon what are held to be the characteristics of the state. Thus, a stateless society is one that has no specialized political roles, let alone institutionalized political structures composed of a plurality of roles. By specialized we mean formal, named, and recognized roles that are played by specialists, not by any or every full member of the society. A specialist is also, properly speaking, full-time. However, this is ambiguous, since the conventional (though highly variable) full-time working day is specific only to industrial societies. Stateless societies do not differentiate between working and leisure hours in this way. Specialization also implies special qualifications—training for achieved roles or inherent qualities for ascribed positions. Specialists are at least mainly and most importantly involved in the concerns associated with their specialization, and they are publicly held to be so.
By political we mean that aspect of social activity which is primarily concerned with power— with getting certain things done, with making decisions and getting them carried out, or preventing things from being done, by and on behalf of some collectivity of persons. In the societies with which we are concerned, the control of violence both in its relatively internal and external aspects looms large, but we would not subscribe to a definition of the political that relates it directly to the use of physical force, for although this possibility is present somewhere in any system of political action, it is essentially an ultimate sanction, usually very indirect and significantly overlaid with other important and conventional mechanisms. In our definition, stateless societies have political action, which is universal, but they do not have either purely or even primarily political institutions or roles.
There is a very fine and at points ambiguous distinction between the ritual and the political. Ritual mechanisms often achieve certain ends, get things done, and even coerce and secure obedience, and it is difficult to see how the essential distinction can be other than that between physical and supernatural forces. Coercion is political when it is backed by a chain of command, when it is part of an ordered series of roles whose players recognize the obligation to support an institution in which the right to secular command is held to inhere. It does not matter whether the reason for this recognition is a perception of rational ends that are properly and regularly to be secured in that way or a perception of an organized following that will support the command and enforce it against opposition, thus involving the ultimate sanction of physical force.
Ritual coercion primarily depends not upon its setting in a chain of command, nor upon followers recognizing a leader’s right to give commands for the common good, but upon the recognized right and obligation to invoke or express the operation of supernatural forces. Backing comes not from recognition of the right to command as inherent in a role, but from recognition of the efficacy of the supernatural forces invoked or expressed and the necessity of bowing and conforming to them. On this basis an apparent or rudimentary chain of command may emerge, and certainly economic motivations may reinforce the process, but the roots of ritual coercion are different from those of political coercion.
The negative starting point of the usual definitional approach to stateless societies is no coincidence. It is due to the fact that these societies first attracted the theoretical interest of scholars who were themselves socialized in states, where the concept of the state had been a consuming object of cultural interest for at least four centuries, from the time of Machiavelli, if not for nearly two and a half millenniums, from the time of Aristotle and Plato. The stateless society has understandably never been such an object of intellectual interest for anyone socialized in one. We are therefore justified in endeavoring to ignore the whole body of philosophical speculation on the state while concentrating on theoretical implications of the empirical facts of stateless societies insofar as they are known and to seek a more positive definition from this point of view. We may then briefly note how, in principle, stateless societies may acquire state forms through endogenous or exogenous change.
The passage from stateless to state forms is difficult to document, because stateless societies have no history of their own and dubious efforts at reconstruction are involved. Migration, climatic change, or improvements in technology may lead to increase in food supply, growth of population, congestion, hardening of boundaries, competition for resources, and attempts at permanent domination by one group over others. These developments will be accompanied by the emergence of new roles and greater specialization. The routinization of charismatic leadership is the most plausible general characterization. Prophets may arise in response to external threats and can easily turn eventually into hereditary ruling dynasties. By such changes the quality of society is radically transformed. However, no doubt the spread of state forms has occurred to an important extent through direct conquest and also through emulation.
To see the state as metaphysically ubiquitous in society and even present in the subhuman herd (Meyer [1884-1902] 1953-1958, vol. 1, pp. 10-12) does not assist understanding in this context. Rather, I stand with Vinogradoff (1920-1922, vol. 1, p. 93) in holding that the state, which has assumed a monopoly of political coordination, ruling, and making laws and enforcing them eventually with coercion, did not exist in ancient times. The commonwealth was not centered in one sovereign body, and the necessary political elements that are never absent from any human society were distributed among many other formations. This emphasis confronts us with the wide range of empirical variation and also the variable quality and direction of interest found in field reports and analyses based on them.
Definition of stateless society. There is no a priori reason why societies negatively defined by their lack of state forms should have many significant features in common. Attempts to delineate the empirical variations of the state cross-culturally have not been very satisfactory so far. Yet despite the enormous cultural variety of stateless societies, it is somewhat astonishing how frequently certain formal characteristics echo through the literature about them. These may be grouped under five very general headings.
(1) Stateless societies are multipolities. (2) Ritual super-integration is achieved in diverse ways beyond the level of the political community. (3) Potentially disruptive action is channeled and constrained in the complementary opposition of groups and categories at several levels, and integrative action may also follow the same lines. (4) The several levels of action share a graded and distributive legitimacy. (5) Intersecting kinship ties provide an essential, pervasive framework. These five characteristics—multipolity, ritual superinte-gration, complementary opposition, distributive legitimacy, and intersecting kinship—are cumulative and interlocking rather than exclusive. It is their combination that is sufficiently frequent to be significant.
Multipolity . All serious investigators have described stateless societies as multipolities. However varied their approaches, they found that whatever level of grouping or crystallization of activity had the greatest clustering and concentration of political functions defined an entity which was to them essentially less than a “society” in the complete, though not autarchic, sense. When Evans-Pritchard used the term “tribe” in the clear yet culturally specific sense of the largest group recognizing a moral obligation to settle feuds and other disputes by arbitration, his definition resulted in splitting what had been and still are regarded as tribes into multitribes; in the case of the Nuer and the Kenya Luo, there were then dozens of different tribes in this new sense. When applied by his followers to other peoples, such as the Dinka (Lienhardt 1958) and the Lugbara (Middleton 1965), this anomaly became so excessive and contradictory to common sense—the Dinka consisting of literally hundreds of tribes—that there has been a reversal to the older, common-sense meaning of tribe, while qualifying terms such as subtribe (Middleton 1963, p. 82; 1965, p. 36) are used for its more precisely defined components.
The other part of this dilemma may be seen, perhaps, in Boas’ (1897) vacillating use of tribe with reference to the Kwakiutl: as applied to that society it refers variously to the dialect group, the subdialect group, the corporate territorial communities within the latter, and even on occasion the component local “clans.” Such ambiguity should be deplored, but it is an apparent consequence of seeking a single level of polity.
Even when major political functions cluster at a particular level of grouping that may reasonably be called the political community, so much of the fabric of social life, such a high proportion and frequency of activity in kinship, marriage, economic exchange, and ritual, and such important meanings and identities lie essentially outside or extend beyond this limit that for the member and for the investigator alike “society,” however defined, must be considerably more inclusive than the political community. This applies even more when political functions are distributed at several levels rather than clustering significantly at one, or when they attach to temporary, ad hoc, situational, and occasional groupings of persons according to recognized principles rather than to permanent or corporate groupings.
Society is that field within which the requisites of social life are effectively met. A further characteristic of the stateless society is that it need not be an absolutely and uniquely bounded entity in space or time, but that its effective field may differ from one person to another and certainly from one family or primary settlement to another. Categorical limits may obtain where natural factors such as seas, deserts, or mountains provide them, or where the movement of peoples results in clearly juxtaposed boundaries of language and culture, but both social intercourse and most cultural elements show a remarkable capacity for transcending the apparent boundaries of mutually unintelligible languages. As this relativity is essential, it is important that it should be so recognized and not explained away as exceptional. However, there are usually at least relative discontinuities in the distribution of social and cultural features, so that for any local group society has effective working limits, although they may differ somewhat for different purposes and have distinctly different limits for several neighboring groups, which nonetheless regard themselves as essentially belonging together.
Ritual superintegration . Ritual superintegration is the most general basis for this wider-than-political effective social field. Of course, such effective social fields exist for persons and for groups in states also, but it seems logical and useful to distinguish between those situations where the effective social field lies for most, if not the vast majority of purposes, conclusively within the boundaries of the state, and those situations where it extends significantly beyond any zone of common political action. Thus “. . . there was no one who had authority over all the Tallensi; no one who could exact tax, tribute or service from all. They never united for war or self-protection against a common enemy” (Fortes 1940, pp. 240-241). Although fighting and warfare between sections and clans of the Tallensi were comparatively frequent, there were nonpolitical sanctions and ceremonies that brought them all together and expressed their unity. “These festivals are periods of ritually sanctioned truce, when all conflicts and disputes must be abandoned for the sake of ceremonial co-operation. ... In this festival cycle, therefore, the widest Tale community emerges. ... It is not a fixed political entity but a functional synthesis” (ibid., p. 263). Tallensi chiefs had no authority beyond the lineage, but they had wider ritual duties. Their office was modeled on the remote chiefship of Mampurugu, but neither that chief nor any of his subchiefs had any economic, juridical, administrative, or military rights over the Tallensi chiefs (ibid., p. 257). There is the same extension of common ritual action beyond the furthest limits of effective and legitimate political authority in all the societies from which examples are drawn here. The mechanisms of ritual superintegration vary, but very frequently they have a heavy kinship component, which is sufficiently important to be singled out under the heading of intersecting kinship.
Complementary opposition . The idea of complementary opposition arose from the work of social anthropologists on segmentary lineage systems, publication of which began in 1940 and continued during the following two decades (Evans-Pritchard 1940a; 1940b; Fortes 1944; 1945; 1953; Mayer 1949; Southall 1952; Smith 1956; Middleton & Tait 1958). In this work there is a constant reiteration of certain closely interlocking concepts, social processes, and structural features, and there is a gradual crystallization of terminology. Evans-Pritchard (1940a, p. 147) wrote of “the principle of contradiction in political structure,” stating that there is “always contradiction in the definition of a political group, for it is a group only in relation to other groups. . . . A section which from the point of view of its members comprises opposed segments is seen by members of other sections as an unsegmented unit.” Elsewhere (1940b, pp. 293, 296) Evans-Pritchard wrote of the “balanced opposition of political segments” and of “complementary tendencies towards fission and fusion which . . . enable us to speak of a system and to say that this system is characteristically defined by the relativity and opposition of its segments.” The term “complementary opposition” gradually came to be accepted (Middleton & Tait 1958).
We may distinguish three dimensions of meaning in the concept of complementary opposition— the classical, the composite, and the extended. The classical refers to the original context of super-ordinate and subordinate segments in a lineage system. But even in the prototypical Nuer case, the actual segments through which the principle of complementary opposition operates are usually territorial sections, not lineage segments as such. The composite may be derived from Fortes’ study of the Tallensi (1945, p. 7), where the idea of complementarity is used with the dual foci of the external Bogar cult and the Earth cult, which have an asymmetrical relationship to the major cleavage between Talis and Ñamóos. Complementarity here refers not only to lineage relationships but also to the ways in which lineage segments are tied together in many complementary and cross-cutting interconnections. For example,
... in spatial relations every maximal lineage belongs to one set of adjacent lineages in the Bogar cult and to a different set of adjacent lineages in the Earth cult. It has, therefore, two intersecting fields of politico-ritual relations so adjusted that its loyalties to the other component lineages of one field are counterbalanced by its loyalties to the component lineages of the other field. . . . Their organization in two complementary configurations around polar symbols checks the dangers of disruptive conflicts that might spring from them. (Fortes 1945, pp. 107-108)
Analogous interconnections of lineages that are not those of clanship appear in Winter’s account of the Amba (1958, p. 154). Thus, the Tallensi Earth cult and external Bogar cult are complementary, as are the Talis and Ñamóos, and the manner in which both cleavages are transcended in unity during the cycle of the Great Festivals is formally analogous to the transcendence of cleavage between subordinate segments by superordi-nate segments in lineage structure.
Intersecting kinship. It seems convenient and logical to extend the concept of complementary opposition to embrace analogous cleavages and transcendent interconnections in nonunilineal systems. In the first place, even in unilineal systems much of the countervailing motivation that transcends segmental oppositions is derived not only from superordinate agnatic bonds but also from cross-cutting cognatic ties. It is the latter ties that are paramount in the complementary oppositions of ego-centered categories in nonunilineal systems such as those of the Ifugao and Kalinga. Here personal kinship ties individuals to series of other individuals who are not in corporate groups. However, the Kalinga local community has something of the corporate identity, based on a dense clustering of cognatic ties, which the Nuer territorial section has in conjunction with its dominant lineage. The privileged intervention in a feud by an Ifugao neutral (Barton 1919) and the reconciliation of hostile communities by a Kalinga pact holder (Barton 1949) must be put in the same class of events as the reconciliation of warring Nuer segments by a leopard-skin chief. Furthermore, it is hard to see how any dualistic cleavage, social or ideological, such as that of the Winnebago expounded by Radin (1923) and explored by Lévi Strauss ( 1963, p. 133ff.), can fail to have the quality of complementary opposition.
The order, balance, and equilibrium that obtain throughout a society as perceived by its members depend largely on the complementary opposition of groups and categories of varying permanence. This mechanism is, of course, not peculiar to stateless societies, but it is of particular importance in them (Smith 1956). It is the self-regulating mechanism whereby, without any titular office or institution that represents or embodies sovereign authority within a bounded society, individuals who interact with one another, directly or indirectly, are ordered and positioned with respect to one another by their own reciprocal activities in dyadic relations and in joint action. This is achieved either through groups or situational categories, which are formed by and give expression to generally accepted principles, whereby the rivalry and hostility between parties in one context are mitigated by the cross-cutting obligations of some of them in another context (cf. Eliot 1948, pp. 59-60). Aggression heightens solidarity, while it is balanced by counteraggression. Despite the ever-present threat and the occasional actuality of violence, suffering, and death for the individual, the social structure is maintained in dynamic equilibrium according to relatively constant principles. The technological means of destruction are also relatively limited, and membership might fluctuate, clan turn into tribe, lineage turn into clan or vice versa, villages split and re-form, kindreds appear and disappear, without appreciable alteration in the over-all structure.
Distributive legitimacy . The characteristic of distributive legitimacy follows from the other four characteristics. As there are many polities within a single society, however differently defined and bounded from within, tensions and conflicts are worked out through the complementary opposition of groups arid categories formed on various bases and at several levels. Legitimate political action also belongs to multiple points and levels, without any overriding monopoly or delegation from above, although different types of action are followed and appropriately sanctioned in each context.
States cannot be multipolities, and although they have often been connected by ritual as well as contractual links, the locus and supremacy of the state have usually remained unambiguous. For states, legitimacy is correspondingly indivisible and undistributed. Although complementary opposition is an important and neglected phenomenon, possibly recognized under other names, it does not itself define the state. While intersecting kinship is undoubtedly ubiquitous, it makes a relatively minor contribution to the constitution of the state.
Stateless societies as working systems. In the course of history stateless societies have been unevenly distributed in terms of the types of ecology and world areas in which they are found. Oceania, sub-Saharan Africa, and the Americas all lay largely outside the sphere of the large-scale literate civilizations—with obvious exceptions such as the Maya, Aztec, and Inca in America and the Muslim peoples in Africa—until these vast areas and myriad peoples were overrun by Europeans during the last three centuries. Until then, peoples with state organization were exceptional, although some important cases and many incipient instances existed in Africa (see Lowie 1948, p. 11) and Oceania.
The five cumulative characteristics discussed above will be documented by as representative a selection of stateless societies as space and the uneven quality of ethnographies permit. These societies are the Kung Bushmen of southern Africa, the Siriono and the Nambikuara Indians of South America, the Murngin and the Tiwi of Australia, the Kwakiutl and other North American Indians, the Nuer of Sudan and the Turkana of Kenya, the Kalinga and the Ifugao of the Philippines, the Plateau Tonga of Zambia, and the Siane and the Star Mountain peoples of New Guinea. These include hunters and gatherers, and agricultural and pastoral peoples; they are patrilineal, matrilineal, and bilateral in descent; they are grouped in nomadic bands, settled villages, localized lineages, kindreds, and age organizations.
It may give a misleading impression to describe any one level of grouping in a stateless society as the political community. But among hunters and gatherers the band is usually taken as such, in the sense of the largest continuously coactive group. It consists of only a few dozen or at the very most a few hundred individuals. But even in the band there often seem to be several alternative levels rather than a single unique level of grouping. Bands may combine or split seasonally; some may become extinct while new ones are formed; and it may be relatively easy for members to switch from one band to another. Bands may be exogamous or not, depending upon the relation between kinship structure and residence rules.
There is considerable ambiguity as to the existence of tribal groupings in which bands are found. Rather than accept the idea that there is an unequivocal tribal identity, I would insist that it is of the essence of the stateless society that the widest sense of common identity should positively attach to different levels and differently based clusterings for different purposes and occasions. Tribal names may be the names of languages rather than of social groups, or they may have some general connotation such as “people” without precise limitation of reference. Frequently they are derived from popular misconceptions, geographical directions, or terms of abuse applied by hostile neighbors (Elkin  1954, pp. 25, 37). In fact, the dominant sense of identity and community properly pertains to different levels for different purposes and occasions because of the different spread and distribution of such relevant features as language, kinship systems, clan or age group names, myths of origin, and similar economies and ways of life. In the case of agricultural and pastoral peoples with large, dense populations, the potential number of relevant levels of identity is greater than for groups with small, sparse populations. There may be seemingly distinctive identities attaching to such names as Kung, Tiv, Gisu, Nath (Nuer), Kwakiutl, and so forth, but very often there are more and less inclusive categories that, on occasion, are more meaningful. Such meanings attach most strongly to nuclear zones, while peripheral to them there may be more significant cross-cutting classifications. A Tiv can prove he is a Tiv by genealogy although his genealogy includes non-Tiv peoples (Bohannan 1958, pp. 35-36). Northern and southern Aranda find one another’s systems incomprehensible (Lévi-Strauss  1963, p. 52).
The Kung bands recently studied by Marshall (1960) averaged twenty-five members, although earlier reports mentioned fifty to one hundred members. They are fluctuating groups of cognatic kin, displaying deep solidarity in joint exploitation of a harsh environment. There are no larger continuing or corporate groupings, but these bands are obviously not complete societies, for on the average each member has almost one-third of his nuclear kin in other bands. Affinal, agnatic, and cognatic links form multiple chains of interconnection from band to band. The number of personal names is limited, and all who possess the same one regard each other as kin and recognize associated obligations. Such categories of fictive kin run from band to band far beyond the range of any one individual’s acquaintance and form the bases of interband feasts and rituals when the opportunity arises.
Siriono bands (Holmberg 1950) are more isolated and self-sufficient, rarely meeting one another. However, marriage and the transfer of membership between bands do occur. There is no evident authority over the band beyond diffuse disapproval and spontaneous and collective criticism and upbraiding that may be evoked by particularly callous and disruptive behavior. Men exercise authority over wives and children, but beyond this level political aspects of activity are hardly discernible. However, the band is closely knit by intersecting kinship ties, and a limited number of general kinship categories clearly define the relationship of all band members to one another. The harshness of the environment in relation to the Siriono’s simple technology reinforces the solidarity of the band’s common residence in a single forest shelter, although individual members can go hunting for days and weeks at a time without others being concerned. Here human society seems reduced to its smallest proportions.
Murngin bands (Warner 1937) have a local patrician core with other attached relatives, but membership in the band fluctuates with the seasonal requirements of the food quest and the performance of numerous rituals. The band is the war-making group and the largest group with any continuity and solidarity. To this extent it may be seen as a political community. There is no violent conflict within it, or with the patricians of its members’ wives and mothers, so that there are changing networks of interclan obligations. Kinship and ritual are elaborated in and beyond the band through a number of interlocking systems involving clans, phratries, and special associations. The social world of the individual is classified into seven lines of descent extending over five generations, 71 kin categories forming eight intermarrying subsections. It is obvious that for any individual, family, or band, the effective field of society includes a much wider complex of interacting and interlocking groups and categories.
The Kwakiutl (Boas 1897; Codere 1950) depend more on fishing than on hunting and are exceptional in the richness of their natural resources and the relative density of population. There is a system of social stratification into nobles, commoners, and slaves, although they lack cultivation and domestic animals. The details of the Kwakiutl traditional social structure will never be altogether certain, but it is known that they lived in seashore villages composed of large houses, each probably occupied by a different descent group. One or several such villages made up the basic political community, which was also the usual fighting unit. There was intense rivalry for prestige between individuals, between brothers, between descent groups, and between political communities. Food, blankets, canoes, houses, coppers, songs and rituals, marriage, murder, and warfare were all exploited for display, consumption, acquisition, and destruction in the endless battle for prestige and rank. Many political communities were tied together in this circulatory process, their chiefs intermarrying, interfeasting, ambushing, and exterminating one another, although close kin ties between polities imposed certain restrictions on their fighting. The elaborate winter rituals also combined several neighboring polities. Order depended on kinship and ritual ramifications combined with the overriding prestige motivation. There were no specific political authorities or sanctions. Chiefs were ranked descent-group heads. They fought for rank, and their people fought and paid to share in it. When the burden was too great they refused, and the chief was powerless. With imperceptible changes of emphasis the same stateless complex transcended many tribal and language barriers throughout the Northwest Pacific Coast.
There are very few accounts of ethnic groups in aboriginal North America that have not confused linguistic, societal, and political units. Cumulative historical scholarship shows, for example, that neither the League of the Iroquois nor its component parts were states and that relations between the tribes were constantly fluctuating and even hostile; however, like other stateless societies they were characterized by the possibility of extending ritual relationships beyond the limits of political communities.
There is no good reason to suppose that the North American Indians fall outside the range of stateless societies, although individual visionary experiences and ritual associations, whose contribution to political order still cannot be precisely defined, were especially common, and there was considerable coherence and continuity of cultures and societies. It seems likely that the appearance and disappearance, splitting and amalgamation, and conquest and defeat of groups and subgroups took place within a fairly consistent structure of kin groups, settlements, and elaborate symbolic activities, each aspect of which might metamorphose gradually at different points from one type of system to another.
The Siane (Salisbury 1962) have 17 “tribal” groups, each composed of from two to nine patrician villages containing several wards focused on large communal men’s houses with separate quarters for women and children. The clan village is the feud unit and is described as “sovereign,” but the whole tribe (or sometimes phratry) is exogamous and constitutes a quasi kin group, to every member of which a kin term can be applied and chartered by myth. When one clan fights, the others in its tribe or phratry remain neutral. They celebrate ancestor festivals together. Violence within the phratry is limited to club fighting.
In the Star Mountains of New Guinea (Pouwer 1964) linguistically defined “tribes” are composed of parish polities divided into hamlets, which are aggregates of nuclear families. The parish is the war unit, yet certain parishes have sacred meeting houses that are the ritual centers for several other parishes. Hamlets are ephemeral and within them fighting is rare. There are 45 patricians common to several tribes, and a common culture extends beyond language boundaries.
The Ifugao (Barton 1919) have no continuously cohesive group larger than the extended family, although bilateral kinship is traced in concentric circles, as it were, as far as seventh cousins, exogamy extending to second or third cousins. Larger named districts are recognized, but they never meet or act together. There are no political authorities; killing and the acquisition of heads for supernatural power and prestige are highly valued, violence being restricted only by the inherent conflict of ramifying bilateral kinship obligations. Members of kindreds do assist one another both in offense and defense, but since each person belongs to many different ego-centered kindreds, he is always faced with a complicated balancing of opposing interests and obligations. The nearer his kinship to a plaintiff and the higher his rank, the stronger is his obligation and self-interest to back and defend the plaintiff and the larger his share of compensation received. The nearer his kinship to a defendant, the heavier is his contribution to compensation and the stronger his motive to enforce good behavior. A’s share of compensation received by B from C may even be canceled out by what he had to contribute to C to enable him to pay it. Moreover, the closer the kin relationship, the more adequate one is as a substitute victim in a feud. Some Ifugao traders from distant regions made personal pacts of mutual obligation to protect and avenge, which tended to escalate into territorial treaties.
The Kalinga (Barton 1949) live in larger, concentrated settlements that form the nuclei of more cohesive regions, whose most prestigeful leaders may enter into pacts that further enhance their prestige and enable all feuds to be settled between them. All the obligations of reciprocal kinship, including exogamy, obtain between pact holders. Pacts are initiated and reaffirmed by exchange feasting, drinking, dancing, flirting, and gift giving.
This striking type of social system, still not well integrated in the mainstream of anthropological thought, is common in most parts of Indonesia, but more isolated examples are found in all major regions of the world.
The Plateau Tonga (Colson 1951) display a singular lack of distinctively political activities in contrast to other African cultivators. They live in small villages having a vague matrilineal core to which other kin and nonkin are attached. There is no larger continuous group than the village, but the feuding organization is based on matrilineages scattered through several villages. The named matriclans are common not only to the Tonga but also to similar neighboring peoples, each associated with an animal and linked together by joking relationships. Although matrilineages rather than villages organize raids, all fellow villagers can be held responsible for one another’s misdemeanors in the context of feuding. Ancestor worship also draws lineage members from beyond the confines of the village. Varying clusters of villages are linked in allegiance to rain shrines of fluctuating repute, whose owners are regarded as representing first settlers and nominal owners of the country. Through them peace is ritually sanctioned when the first fruits are annually presented and the local spirits express their demands. The claims of men on the bridewealth of their classificatory sisters’ daughters or great-granddaughters give rise to intricate links between matrilineages that bind neighborhoods together.
Nuer society (Evans-Pritchard 1940a; 1940b) is arranged in an articulated segmentary series of territorial units, one inside the other, whose common action and coherent interrelation derive from the agnatic core that runs through each, although these agnates may be a minority of the whole. It is the dialectic between the territorially based communities and the agnatic cores on which they are structured and the networks of personal kinship ties that criss-cross between them that provide the operational mode of Nuer society. The political community is the largest segment capable of joint action and of settling internal disputes by standardized processes, and there are some dozens of these among the Nuer. Many of the clans are dispersed all over the country, creating personal and corporate obligations between members and sections of different polities. Initiation into the age organization is coordinated within three or four divisions of the country and held jointly by some smaller polities, while the social age equivalence established by it is recognized by all Nuer. Certain prophets acquired a reputation extending over many polities and could even induce occasional joint political action in warfare. The definition of a Nuer polity is contingent purely upon which level of segmentary grouping empirically acts as such and does so with sufficient consistency to permit its actions to be regarded as morally obligatory. This recognition is strongly influenced by the kinship and ritual relations that extend far beyond the usual field of political action.
Turkana society (Gulliver 1955; 1958) has no permanent residential grouping or authority beyond the extended family, but the orderly activity of much larger temporary collections of individuals is coordinated primarily by the principles that cluster around the age organization. Father and son belong to opposite-named alternations, so that every temporary gathering of men is precipitated into momentary moieties. Youths are initiated into new named age groups about every five years at some 25 centers. Through informal canvassing one name comes to be adopted for the same age set all over the country. Age organization both defines and emphasizes social age. The oldest men available at any time or place are the most potent ritually and are nearest the deity; they pray for rain and stop fights by walking between combatants. Homicide is almost unknown among the Turkana, since it is believed to prevent rain, the greatest disaster to nomadic pastoralists in semi-desert country. The Turkana secure their general social needs and protection further by developing personal kin networks, to which they add stock associates through the exchange of cattle. Large exogamous patricians are dispersed throughout Turkana society and among neighboring related peoples. Age organization gives a common identity to all Turkana and a spatial framework that can precipitate any chance assembly into agreed relationships and orderly common action. The Turkana never fought their related Jie neighbors, but they fought the equally related Karamojong and Dodoth, as well as the unrelated Samburu Masai and Marile, with whom they also intermarried.
Discussion. Having presented these few highly compressed examples, so diverse geographically and structurally, yet all maintaining social life without state forms through certain basic common principles, we may introduce some further recurrent themes. Few of the names by which these peoples are ethnographically known are their real names, and usually there is a cultural continuum along which social relations and cultural characteristics are clustered and concentrated at different times and places and for different purposes in a variable manner, yet according to coherent principles. Either cultural transformations are essentially gradual and phase into each other without discernible points of confrontation or clear lines of divergence, or even if there are marked boundaries, say, in language, they do not coincide with those of other cultural phenomena, as when zones of warfare, intermarriage, clanship, and ritual observance fall at different points so that again the net outcome is one of relativity. This situation can only obtain where large continuous areas are occupied by equally stateless societies. Clearer distinctions naturally appear where there is confrontation with more politically specialized groups or with recent migrants of a radically different culture that is not yet integrated.
None of the peoples discussed above have an unambiguous, categorical tribal designation, with the possible exception of the Turkana. The case of the Turkana is particularly interesting and in some respects represents the polar opposite to that of the Murngin. The latter have minute and elaborate local differentiations of many kinds, to the point that they insist that even clans have separate named languages although in fact they may not significantly differ. The result is that Murngin society consists of a large though fluctuating number of highly diverse parts. On the other hand, since the Turkana have no permanent groups beyond the family, they require a rather precise framework of principles to order their wider relationships. The spatial recognition of this framework provides an unusually clear group definition, although it is far larger than the effective social field for any Turkana individual or group. Thus, paradoxically, Turkana homogeneity differentiates the people and Murngin diversity does not; both arrangements serve the same general end in the maintenance of stateless society. These societies are not only stateless, but most of them are essentially nameless. They have designations, but these are specific to purpose, time, and place and must not be used to define the unique limits of the society. Distinctions that make unique names acceptable, recognized, meaningful, and necessary are part of the process of imperative over-all political coordination and specialization, out of which state structures arise from stateless situations that are logically incompatible with them.
Leadership . Certain characteristics of leadership and of behavior in situations of serious conflict are noticeably recurrent in these examples. Leadership depends to a surprising extent on personal qualities of achievement within ascribed limits of eligibility. Thus it seems that any Kung who belongs to a group that has general rights in an area can form a band if he can get people to follow him. Once established, his position becomes hereditary among his cognatic descendents. However, if these descendents fail, the position lapses although the band may continue. Leadership must be valued, although it is not indispensable. The leader has no imperative authority, and apart from his undoubted prestige he seems to have far more obligations than rights. He supplies initiative in leading the band from camp to camp and in planning hunting and other activities. The more successful he is, the greater is the dependence upon him. Lévi-Strauss strengthens the idea of recurrent characteristics of leadership in his account (1944) of the Nambikuara, who are formally rather similar to the Kung but culturally and geographically close to the Siriono. There is free choice of membership between bands, which unite in a temporary village for cultivation during the rains and then disperse again. Multiple kin ties run through many bands, so that a family may be more closely tied to another in a different band than it is to any in the same one. Chiefs are the relatively stable nuclei round which bands fluctuate and cluster. Chiefs have no political authority or sanctions; their efficacy depends fundamentally on doing better than others in the food quest and all essential social activities and on supplying initiative and leadership by consent of the group, which is their only basis of legitimacy. They have to work harder and be more generous than others in food-giving and entertainment. It would be impossible for them to satisfy these demands were it not for their effective monopoly of polygyny, through the practice of secondary unions, which the band grants as long as they succeed, and which in turn permits them success through the extra womanpower of subsidiary wives who supply labor, hunting assistance, sexual gratification, and general support rather than conventional child-bearing and domestic services. Yet the chiefs’ motivation is for prestige more than for women or food. Although they nominate their successors formally, this act can only ratify informal support by popular consent. The chiefs’ ambivalence, reluctance, and even refusal to assume the heavy burdens of leadership, also found in many far more specialized institutions of chiefship, are highly intelligible in this small-scale example. Leadership at its most elemental naturally depends on being judged better than the led, but even here ascribed aspects of the leader himself or of his family and descent group are frequently taken into account.
The Nuer who is the “bull,” or leader of his lineage segment and of the territorial group focused upon it, acquires this position through a similar combination of descent, ability, and good fortune in establishing a successful family. The leopard-skin chiefs, through whom armed conflicts may be resolved, must come from particular lineages, but within this wide area of eligibility their position as ritual conciliators depends solely upon the consensual insistence of the people. Among the Siane, the big man’s pre-eminence depends upon his age and his success in oratory, in founding a large family, and on his example of industry, wealth, generosity, and bravery. Any man can aspire to this position, and the status of figurehead so achieved depends upon general consent, although it is easier for one who stands in the structural category of “eldest brother.”
In principle, any Tonga man (and even woman in some cases) can form a village and become a headman if he has the necessary qualities for attracting and holding such a following. But as soon as such a position is achieved it acquires hereditary qualities, although there is a wide group of possible successors. As with the Nambikuara and Kung leaders, the Tonga headman has additional responsibilities but very few privileges and no political authority or sanctions.
The most general level of Turkana leadership is determined by the age organization, which designates the most senior person present as nearest the divinity. The Kwakiutl chiefs were hereditary heads of kin groups occupying clearly defined relative positions in a hierarchy of rank, but where each one came into this hierarchy at any time depended upon his personal achievements and the backing of his followers in the fluctuating hazards of the potlatch. Shamanistic divination offered those of lower status an arduous path to the personal achievement of another form of pre-eminence and prestige.
Head-hunting was the chief personal path to high status for Kalinga men, but birth into a family whose head already enjoyed wealth and high status was such an advantage that it was usually from this combination that pre-eminent positions of ritual influence and superintegration as pangats and pact holders were achieved and frequently passed on within the family. So great were the prestige at stake and the supernatural sanction and loss of face involved that a Kalinga pact holder would even kill a transgressing kinsman himself rather than fail in fulfilling a pact obligation.
Conflict resolution . For the Kung, the Siriono, and many other hunting and gathering peoples, the sheer quest for food and physical and social survival are all-engrossing, and physical violence or armed conflict between groups hardly ever occurs. But in other cases, especially among the larger-scale stateless societies of agricultural and pastoral peoples, ingenious ritual mechanisms of conflict resolution are important. This is true also of many Australian peoples, despite the fact that they do not practice agriculture.
It is certainly a common characteristic of stateless societies, in all major regions where they are found, that the highest levels of leadership and the widest levels of integration are secured and set in a ritual context. Legitimacy is distributive or pluralistic in that it belongs to several groups and categories, which are interlocked and functionally differentiated within the same social system yet not hierarchically ordered, and which are integrated yet not coordinated by any over-all political authority. There are frequently several distinct sources of leadership—typically, some arising from the structure of groups and interlocking categories, based on personal achievement within ascribed limits of eligibility, and some arising from more purely personal factors of a supernatural character, related to divination and personal experiences of revelation often developing outside the highest status levels. Examples are the shaman diviners in Kwakiutl and many stateless societies in Asia, or the prophets and diviners in Nuer, Turkana, Tonga, and many other African societies. There are also many instances from Africa, Asia, Oceania, and the Americas in which different levels of conflict were delimited by methodical restrictions, as in the progressive limitation of open warfare to adult males, to fighting only with clubs or sticks instead of with more harmful weapons, to wrestling matches and other formalized types of conflict that provide cathartic relief with little danger of serious injury.
In many of these cases it is difficult to resist the conclusion that in the very foundations of these simple social systems there are various mechanisms biased in the direction of maintenance, collective survival, and equilibrium. The techniques of conflict resolution would so often remain inoperative were it not for a priori sanctions of supernatural beliefs, reinforced by conveniently built-in motivations of prestige, greed, and even sexual gratification. Not only in the very small-scale societies, but also in the relatively small local groups of even the larger ones, there is constant reference to the fact that violent and continued conflict does not commonly occur in primary local groups because it is mutually recognized by all as too inconvenient and uncomfortable, so that those inclined to cause it are restrained by the spontaneous pressure of opinion against them or are ultimately forced to leave the group. It is important to note that this is characteristic of societies in any part of the world (such as the Nuer and the Siane) in which the expression of violence is approved and common in other contexts.
The stateless society is both an empirical, historical fact and a comment on political theory. From the time of Plato and Aristotle to the twentieth century political theory has been Western-oriented and ethnocentric, normative rather than empirical. Information about the stateless societies, which existed throughout the greater part of the world until recent centuries, was for practical reasons lacking and was in any case used as a negative foil for Western experience rather than valued for its own sake. Since the major interest was in the problems of political philosophy arising out of the cumulative Western experience, the incentive to understand the internal workings of stateless societies as empirical phenomena was doubly lacking. Some exceptions may be found, as in the case of the great historical jurisprudents, such as Maine, Maitland, Pollock, and Seebohm. Despite the excellence of their historical researches, their accounts of earlier phases of Western societies inevitably left crucial gaps, while effective data on contemporary stateless societies were still not available to them. Indeed, no adequate accounts of the political structure and workings of Oriental states have been available to the West until very recently, let alone of stateless societies outside the Western world, which, although smaller and in a sense simpler, are more difficult to penetrate and understand. Most works on political philosophy make brief initial reference to a largely hypothetical situation in which the institution of the state as generally understood is lacking, but this is a negative construct with about as much reality as the noble savage. There are two main attitudes: one in which the emergence of the state, so fundamental to society as we know it, is regarded as a boundary on the other side of which society hardly exists as a recognizable part of our common humanity; and another in which it is maintained that the state, as the fundamental institution of society, is already implicit in a herd of buffalo or a troop of monkeys and to that extent necessarily also attributable to early man.
Positive emphasis on stateless societies as political entities worthy of empirical study in their own right came only with twentieth-century anthropology, and ironically (one wonders whether also necessarily), at a stage of world history when they were hardly any longer to be found as going concerns. It is difficult to get out of the habit of looking at these small stateless societies in negative terms. They do lack the definitive territorial boundaries, exclusive sovereignty, and central monopoly of legitimate force that characterize states. We must regard states as a further evolved form of human polity than stateless society. But the latter have very positive characteristics. The most general and important is the fact that fundamental responsibility for the maintenance of society itself is much more widely dispersed throughout its varied institutions and its whole population, at least, usually, all its adult males. The remarkable spectacle of societies positively maintaining themselves at a high level of integration without any obvious specialized means of enforcement has undoubtedly led to new insight and attention to the fundamental responsibilities of all citizens, which for most people are obscured by the ubiquity of specialized political institutions. In stateless societies every man grows up with a practical and intuitive sense of his responsibility to maintain constantly throughout his life that part of the fabric of society in which at any time he is involved.. Stateless societies are so constituted that the kaleidoscopic succession of concrete social situations provides the stimulus that motivates each individual to act for his own interest or for that of close kin and neighbors with whom he is so totally involved, in a manner which maintains the fabric of society. It is a little like the classical model of laissez-faire economics translated into the political field. But if every man is thus for himself, he is so only within a very tight framework of reciprocal obligation that he cannot avoid absorbing. The lack of specialized roles and the resulting multiplex quality of social networks mean that neither economic nor political ends can be exclusively pursued by anyone to the detriment of society, because these ends are intertwined with each other and further channeled by ritual and controlled by the beliefs which ritual expresses.
Barton, Roy F. 1919 Ifugao Law. University of California Publications in American Archaeology and Ethnology, Vol. 15, No. 1. Berkeley: Univ. of California Press.
Barton, Roy F. 1949 The Kalingas: Their Institutions and Custom Law. Univ. of Chicago Press. → Published posthumously.
Berndt, Ronald M. 1964 Warfare in the New Guinea Highlands. American Anthropologist New Series 66 (Special Publication) no. 4, part 2:183-203.
Boas, Franz 1897 The Social Organization and the Secret Societies of the Kwakiutl Indians. Based on personal observations and on notes made by George Hunt. Pages 311-738 in U.S. National Museum, Annual Report, 1895. Washington: The Museum.
Bohannan, L. (1958) 1964 Political Aspects of Tiv Social Organization. Pages 33-66 in John Middleton and David Tait (editors), Tribes Without Rulers: Studies in African Segmentary Systems. London: Routledge.
Codere, Helen F. 1950 Fighting With Property: A Study of Kwakiutl Potlatching and Warfare, 1792-1930. American Ethnological Society, Monograph No. 18. New York: Augustin.
Colson, Elizabeth 1951 The Plateau Tonga of Northern Rhodesia. Pages 94-168 in Elizabeth Colson and Max Gluckman (editors), Seven Tribes of British Central Africa. Oxford Univ. Press.
Easton, David 1959 Political Anthropology. Biennial Review of Anthropology :210-262.
Eliot, T. S. (1948) 1949 Notes Towards the Definition of Culture. New York: Harcourt.
Elkin, A. P. (1938) 1954 The Australian Aborigines. 3d ed. Melbourne: Angus & Robertson. → A paperback edition was published in 1964 by Doubleday.
Evans-Pritchard, E. E. (1940a) 1963 The Nuer: A Description of the Modes of Livelihood and Political Institutions of a Nilotic People. Oxford: Clarendon.
Evans-Pritchard, E. E. (1940b) 1961 The Nuer of the Southern Sudan. Pages 272-296 in Meyer Fortes and E. E. Evans-Pritchard (editors), African Political Systems. Oxford Univ. Press.
Evans-Pritchard, E. E. 1949 Luo Tribes and Clans. Rhodes—Livingstone Journal 7:24-40.
Fortes, Meyer (1940) 1961 The Political System of the Tallensi of the Northern Territories of the Gold Coast. Pages 239-271 in Meyer Fortes and E. E. Evans-Pritchard (editors), African Political Systems. Oxford Univ. Press.
Fortes, Meyer 1944 The Significance of Descent in Tale Social Structure. Africa 14:362-384.
Fortes, Meyer 1945 The Dynamics of Clanship Among the Tallensi: Being the First Part of an Analysis of the Social Structure of a Trans-Volta Tribe. Published for the International African Institute. Oxford Univ. Press.
Fortes, Meyer 1953 The Structure of Unilineal Descent Groups. American Anthropologist New Series 55:17-41.
Gulliver, Phillip H. 1955 The Family Herds: A Study of Two Pastoral Tribes in East Africa, the Jie and the Turkana. London: Routledge.
Gulliver, P. H. 1958 The Turkana Age Organization. American Anthropologist New Series 60:900-922.
Gulliver, Pamela; and Gulliver, P. H. 1953 The Central Nilo-Hamites. London: International African Institute.
Hart, Charles W. M.; and Pilling, Arnold R. 1960 The Tiwi of North Australia. New York: Holt.
Hodge, Frederick W. (1907-1910) 1959 Handbook of American Indians North of Mexico. 2 vols. U.S. Bureau of American Ethnology, Bulletin No. 30. Parts 1-2. New York: Pageant.
Holmberg, Allan R. 1950 Nomads of the Long Bow: The Siriono of Eastern Bolivia. Smithsonian Institution, Institute of Social Anthropology, Publication No. 10. Washington: Government Printing Office.
La Fontaine, J. S. 1959 The Gisu of Uganda. London: International African Institute.
Levine, Robert A. 1960 The Internalization of Political Values in Stateless Societies. Human Organization 19:51-58.
LÉvi-Strauss, Claude 1944 The Social and Psychological Aspects of Chieftainship in a Primitive Tribe: The Nambikuara of Northwestern Mato Grosso. N.Y. Academy of Sciences, Transactions 2d Series 7:16-32.
LÉvi-Strauss, Claude (1958) 1963 Structural Anthropology. New York: Basic Books. → First published in French.
LÉvi-Strauss, Claude (1962) 1963 Totemism. Boston: Beacon. → First published as Le totémisme aujourd’ hui.
Lienhardt, Godfrey (1958) 1964 The Western Dinka. Pages 97-135 in John Middleton and David Tait (editors), Tribes Without Rulers: Studies in African Seg-mentary Systems. London: Routledge.
Lowie, Robert H. (1922)1962 The Origin of the State. Rev. & enl. New York: Russell. → An enlarged and revised version of articles first published in the Freeman in 1922.
Lowie, Robert H. (1935) 1956 The Crow Indians. New York: Rinehart.
Lowie, Robert H. (1948)1960 Some Aspects of Political Organization Among the American Aborigines. Pages 262-290 in Robert H. Lowie, Selected Papers in Anthropology. Berkeley: Univ. of California Press. → First published in Volume 78 of the Journal of the Royal Anthropological Institute of Great Britain and Ireland.
Mair, Lucy P. (1962) 1964 Primitive Government. Baltimore: Penguin.
Marshall, Lorna 1960 !Kung Bushman Bands. Africa 30:325-355.
Mayer, Philip 1949 The Lineage Principle in Gusii Society. International African Institute, Memorandum No. 24. Oxford Univ. Press.
Meyer, Eduard (1884-1902) 1953-1958 Geschichte des Altertums. 5 vols. Basel: Schwabe.
Middleton, John 1963 The Yakan or Allah Water Cult Among the Lugbara. Journal of the Royal Anthropological Institute of Great Britain and Ireland 93:80-108.
Middleton, John 1965 The Lugbara of Uganda. New York: Holt.
Middleton, John; and Tait, David (editors) (1958) 1964 Tribes Without Rulers: Studies in African Segmentary Systems. London: Routledge.
Moss, Claude R. 1920 Nabaloi Law and Ritual. University of California Publications in American Archaeology and Ethnology, Vol. 15, No. 3. Berkeley: Univ. of California Press.
Pouwer, Jan 1964 A Social System in the Star Mountains: Toward a Reorientation of the Study of Social Systems. American Anthropologist New Series 66 (Special Publication) no. 4, part 2:133-161.
Radin, Paul 1923 The Winnebago Tribe. U.S. Bureau of American Ethnology, Annual Report 37:35-560.
Salisbury, Richard F. 1962 From Stone to Steel: Economic Consequences of a Technological Change in New Guinea. London and New York: Melbourne Univ. Press; Cambridge Univ. Press.
Smith, M. G. 1956 On Segmentary Lineage Systems. Journal of the Royal Anthropological Institute of Great Britain and Ireland 86, part 2:39-80.
Southall, A. W. 1952 Lineage Formation Among the Luo. International African Institute, Memorandum No. 26. Oxford Univ. Press.
Swadesh, Morris 1948 Motivations in Wootka Warfare. Southwestern Journal of Anthropology 4:76-93.
Vinogradoff, Paul 1920-1922 Outlines of Historical Jurisprudence. 2 vols. Oxford Univ. Press. → Volume 1: Introduction. Tribal Law. Volume 2: The Jurisprudence of the Greek City.
Warner, W. Lloyd (1937) 1958 A Black Civilization: A Social Study of an Australian Tribe. Rev. ed. New York: Harper.
Watson, James B. (editor) 1964 New Guinea: The Central Highlands. American Anthropologist New Series 66 (Special Publication) no. 4, part 2.
Winter, E. H. 1958 The Aboriginal Political Structure of Bwamba. Pages 136-166 in John Middleton and David Tait (editors), Tribes Without Rulers: Studies in African Segmentary Systems. London: Routledge.