Rhythmic Anarchy in the Nuer (Evans-Pritchard – 1940)
A few years after his French counterpart but strangely without referring to him, Evans-Pritchard proposed a conception of archaic society politics which considered it not merely as a structure or a set of differential relations, but primarily as a web of relations which were expressed and brought about rhythmically.
Here again we have first to overcome a biased reception. In his introduction to the French translation of the Nuer (1969), Louis Dumont has operated a double move similar to that which Lévi-Strauss carried out in 1950 with respect to Mauss [1]: he sought to show, on the one hand, that Evans-Pritchard had anticipated Structuralism (thereby proving him right for championing it), but, on the other hand, that this anticipation had not been carried out adequately (thereby justifying the corrections that he felt obliged to bring to his work).
Evans-Pritchard would have laid the foundations of the “structural analysis” of “social systems,” but his concern for the “political” dimension—which would be, according to Dumont, an undesirable effect of our modern individualist ideology—would unfortunately have introduced into his work a certain “ambiguity,” which would have allowed some of the reductive interpretations developed eventually. Quite the contrary, I think that Evans-Pritchard, in The Nuer, moved in a direction totally alien to that of Structuralism, and that while using sometimes the term “structure,” he reached conclusions that were close to those Mauss and Granet had just arrived at a few years before.
Let’s go back to The Nuer. Evans-Pritchard recognized among the Nuer a list of hierarchized collective forms close to that provided by Mauss. At the top was the Nuer people. It was composed of seven tribes, divided in turn into segments of successive smaller size, nested in each other, called by the anthropologist primary, secondary and tertiary sections. Each tertiary tribal section eventually included a number of village communities, consisting of kinship groups and domestic groups. These political divisions corresponded largely with clan divisions (defined as the most extensive groups of agnates who attributed their origin to a common ancestor and between which sexual intercourse was forbidden) and lineages (subdivisions of the clan), but they could also be crossed with divisions by tribes. Finally, all these divisions intersected with divisions by sex, age, and generation.
Like Mauss, Evans-Pritchard noted that political authority did not depend on a superior state organization common to all the Nuer people, but that it belonged to their social system [2]. The Nuer people formed an extremely loose community, linked with a territory, a language, a few religious forms, matrimonial rules and customs such as the removal of the lower incisors or the six notches worn at the front by men from the day of their initiation. Its political unity was not ensured by any central administration, nor by any common law and only materialized by the fleeting alliance of the tribes which composed it (apparently especially the adjacent tribes) in the recurrent wars opposing them with their neighbors Dinka or to the Egyptian and English invaders (p. 123). The tribes that formed the lower level might seem more consistent, but the political logic that prevailed there was in fact the same. Each tribe bore a name, implied military solidarity (p.120) and a moral obligation to settle vendetta and other quarrels by arbitration: “We may therefore say that there is law, in the limited and relative sense defined in Chapter IV, between tribesmen, but no law between tribes” (p. 121). But these were in turn composed of interlocking sections which in fact rarely acted in concert and were in a permanent state of tension with each other. The tribe had no more common authority than the people.
Thus, at each level of the pyramid, there existed a form of solidarity, community or grouping that had a certain reality while simultaneously being undermined by internal divisions, which prevented any formation of a differentiated political authority. Evans-Pritchard described this paradoxical situation with the oxymoric expression “ordered anarchy” (p. 6). Now, like Mauss, he put the accent, to explain this phenomenon, both on the differential aspect of groupings and on the crossing of divisions. The Nuer people were part of the system of Nilotic peoples in the region and defined themselves in opposition to other peoples (Dinka, Shilluk, Anuak). The tribes that constituted it were opposing communities: “The internal organization of each tribe can only be fully understood in terms of their mutual opposition, and their common opposition to the Dinka who border them” (p. 123). And the sections of each tribe were in the same way in opposition to each other. However, since in these polysegmentary systems each particular individual was simultaneously part in a web of heterogeneous, even competing communities, someone could be a member of a tribal section but part of a clan that extended to another one and thus felt solidary with members of a section different from his own. This explained the peculiar “individualism” of the Nuer. They should, of course, count on their kinship or the solidarity that bound their section to the neighboring tribal sections, but these links were not permanent and were periodically changed into animosity. Their sentiment of belonging to “communities” was in no way contradictory to a fundamental “individualism.” Contrary to what many sociologists and anthropologists have been repeating since Spencer, Tönnies and Durkheim, without ever really thinking about it, there was no opposition between Gemeinschaft and Gesellschaft, nor between holism and individualism. These apparent oppositions result from the mere fact that the relations between individual and society have been deduced from a picture that has been artificially immobilized. Once observed in its movements, that is to say in its rhythms, singular and collective individuation is only one single phenomenon of variable nature which simultaneously transforms singular and collective individuals.
It is remarkable that, like Mauss, Evans-Pritchard did not separate this “structure” from the temporality during which it was actualized. On the contrary, it had no reality apart from the rhythm of the exchange and conflict life which actualized it. A simplification made by Dumont is revealing of what was at stake in Evans-Pritchard’s work and, at the same time, of how Structuralism dehistoricized his real perspective. Evans-Pritchard noted that the words (in particular those designating the various groups) took on different meanings according to the situation in which they were pronounced.
If one meets an Englishman in Germany and asks him where his home is, he may reply that it is England. If one meets the same man in London and asks him the same question he will tell one that his home is in Oxfordshire, whereas if one meets him in that county he will tell one the name of the town or village in which he lives. If questioned in his town or village he will mention his particular street, and if questioned in his street he will indicate his house. So it is with the Nuer. (The Nuer, 1940, p. 136)
This intuition, which the linguistics of discourse founded by Benveniste has proved entirely right and which implied that the system of language does not precede the enunciation but instantiates it each time in a particular way according to the particular situation of enunciation, is interpreted in the following manner by Dumont (who quotes Pocock, a commentator on the Nuer).
“The words and the objects or attitudes to which they refer must be understood in their relations as constituting a system endowed with meaning.” By passing from Radcliffe-Brown to Evans-Pritchard, we thus find “a movement from function to signification,” and the exegete here must emphasize the connection with Levi-Strauss. (L. Dumont, “Préface,” op. cit., p. VII.)
The radical historicity of the system and its oppositions, their essentially dynamic aspect, are thus transformed into a relative historicity derived only from the non-naturalness of the cultural signs. As in Structural linguistics, which postulated the primacy of the language – la langue over the discourse – le discours, the system actually pre-existed its instantiation, which was at best only a “putting-into-action” or at worst an “expression.” Discourse and action were considered merely as “uses” or “manifestations” of the structures that pre-existed them.
Now, if we pay careful heed as to how Evans-Pritchard accounted for these variations in allegiance to different communities, we will see that he never presupposed the existence of a “structure,” but that he always started—just as the most modern linguistics started from discourse—from the unfolding conflicts and exchanges themselves. The latter formed the two inseparable sides of sociality and generally varied in the same way.
Greater hostility appears to be felt between villages, groups of villages, and tertiary tribal sections than between larger tribal sections and between tribes. Probably the raids conducted tribally and in tribal federation against the Dinka had an integrating action, but the Dinka were not aggressive against the Nuer and it seems that the maintenance of tribal structure must rather be attributed to opposition between its minor segments than to any outside pressure. If this be so, and a consideration of the institution of the feud suggests that it is so, we arrive at the conclusion that the more multiple and frequent the contacts between members of a segment the more intense the opposition between its parts. (The Nuer, 1940, p. 150)
But the observation of trade and conflict showed that the hierarchical structure of interlocking sections described above did not really exist and could anyhow never be observed in its entirety. It was a web of divisions that could never be actualized together or that could be actualized only successively. In reality, the following happened.
Each segment is itself segmented and there is opposition between its parts. The members of any segment unite for war against adjacent segments of the same order and unite with these adjacent segments against larger sections. (The Nuer, 1940, p. 142)
In other words, every individual considered himself a member of his tertiary section in his conflicting relationship with the members of the other tertiary sections, but a member of his secondary section (and in a co-operative relationship with the members of the other tertiary sections) in its relation to the members of the other secondary sections, and so on going up the pyramid. Thus, as suggested by the semantic variations of the group denominations, the sentiment of belonging was differential, but—and this is essential—these sentiments were changing according to the situation.
One value attaches a man to his group and another to a segment of it in opposition to other segments of it, and the value which controls his action is a function of the social situation in which he finds himself. (The Nuer, 1940, p. 137)
Thus, observation did not deliver a structure but, again, a rhythm. The tribal groups had a variable geometry and were defined through a succession of fusions and fissions. The rhythmic alternation of conflict and alliance between the different lineage and territorial segments was the basis of the Nuer singular and collective individuation.
Between tribes there can only be war, and through war, the memory of war, and the potentiality of war the relations between tribes are defined and expressed. Within a tribe fighting always produces feuds, and a relation of feud is characteristic of tribal segments and gives to the tribal structure a movement of expansion and contraction. (The Nuer, 1940, p. 161)
To describe this hitherto unnoticed reality, Evans-Pritchard used the image of rubber bands that alternately stretch and tighten.
We would emphasize further that blood-feuds only directly involve a few persons and that though they sometimes cause violence between whole local communities—a feud in a wider sense—ordinary social contacts continue in spite of them. The strands of kinship and affinity, of age-set affiliations, and of military and even economic interests remain unbroken; and these strands act like elastic between the sections, being capable of considerable expansion by disturbed political relations, but always pulling the communities together and keeping them as a single group in relation to other groups of the same kind. (The Nuer, 1940, p. 162)
It is true that when he tried to formalize his observations, Evans-Pritchard declared that the “structure remains fairly constant” (p. 107), but the Structuralist reading has rendered this assertion opaque, because it denoted less the fixity of a structure than the constancy of a rhythm, i.e. of the oscillations by which groups and individuals were brought about. And that is why he coined this seemingly strange and so un-Structuralist expression of “structural movement.”
The blood-feud may be viewed as a structural movement between political segments by which the form of the Nuer political system, as we know it, is maintained. (The Nuer, 1940, p. 158)
Far from Lévi-Strauss, Evans-Pritchard underlined, like Mauss, “the dynamic quality of the political structure” (p. 148) and noted that “it cannot very easily be pictured diagrammatically, for political relations are relative and dynamic” (p. 137). And to make himself clear, he submitted any structural analysis to the primacy of dynamic and rhythmic description.
[Political actualities] are conflicting because the values that determine them are, owing to the relativity of political structure, themselves in conflict. Consistency of political actualities can only be seen when the dynamism and relativity of political structure is understood and the relation of political structure to other social systems is taken into consideration. (The Nuer, 1940, p. 138)
Evans-Pritchard’s description provided a picture of the rhythm of individuation that basically corroborated Mauss’ analysis. Singular and collective individuation were rhythmic processes. At the same time, it showed that the anthropological concept of rhythm should be somehow complexified. First, conflict rhythms were not exactly parallel to morphological rhythms. Feuds in the ternary section or between ternary sections were much more frequent during the rainy season than during the drought in which agreement or compensation were more easily reached. But other Nuer tribes were attacked, or the Dinka raided, mainly during the dry season. In other words, the conflict followed different rhythms depending on the level in the social pyramid. Second, solidarity rhythms varied accordingly. The smaller the tribal section, the more vigorous “the sense of community, close lineage ties, and some economic interdependence” (p. 157). Conversely, “the larger the segment involved the greater the anarchy that prevails” (p. 157). Both factors, the periodic transformation of the differential oppositions during the year, and the decreasing solidarity when climbing up into the social system, could not be accounted for from a plain structural perspective. Not only did the Nuer social system not really exist per se, but it did not have a homogeneous “tension” because the “differentials” that defined its “values” were not identical. Both solidarity and tension resulting from exchanges and conflicts were the strongest at the base of the social system and diminished when going up through it. They were next to nil at the level of the Nuer people. But there were strong reasons to believe that these variations in tension and solidarity reflected in turn a change in the rhythms of sociality (trade and conflict) by which the various singular and collective individuals were formed. The rapid alternation of the fission/fusion cycles that characterized the smaller sections was included into larger and perhaps more erratic ones (apparently annual or biennial) of the wars against the Dinka.
