Archaeology and the Origins of Social Stratification in Southern Bougainville
John Terrell p. 23-43


One of the Lessons of modern economic geography, abstract theory of graphs, and contemporary thinking in theoretical biology appears to be that hierarchical control networks are one solution to the problem of what to do about systems that are so complex, they may be unstable, unworkable, uneconomical, or any of these in combination. The issue explored in this essay is : Had the dynamics of life in southern Bougainville Led to the evolution of a hierarchical (stratified) system of social controls in Buin territory before the arrival of the Europeans ? Reviewing the controversy in the ethnographic literature between Douglas Oliver and Richard Thurnwald over the origins of the allegedly "feudal" society of the Buin-speakers, as well as recent archaeological finds from the Buin Plain, it is suggested that the need for social controls in southern Bougainville was probably not sufficiently complex enough that a system of ascribed leadership roles or "hereditary statuses" might have been an optimal solution to systems complexity in social, economic, political, etc. relationships. This observation has direct bearing on theoretical arguments concerning the origins of stratification and the state system of governance : notably the work of Service, Sahlins, and others in America. While the question of origins may be historically fascinating, it may be trivial in theoretical importance.

Une des leçons tirée de la géographie économique moderne, la théorie abstraite des graphiques et les pensées contemporaines en biologie théorique paraît être que les réseaux de contrôle hiérarchique sont une solution du problème de ce qu'il faut faire concernant les systèmes qui sont tellement complexes qu'ils risquent d'être déséquilibrés, impraticables, peu économiques, ou n'importe quelle combinaison de ceux-ci. La question étudiée dans cet article est : Est-ce que la dynamique de vie en Bougainville du Sud a mené à l'évolution d'un système hiérarchique (stratifié) de contrôles sociaux dans le territoire de Buin avant l'arrivée des Européens ? Examinant la controverse dans la littérature ethnographique entre Douglas Oliver et Richard Thurnwald sur les origines de la société dite "féodale" des parieurs de Buin, ainsi que les découvertes archéologiques récentes dans la Plaine Buin, il est suggéré que le besoin pour les contrôles sociaux en Bougainville du Sud n'étaient probablement pas suffisamment complexes pour qu'un système de rôles attribués au chef ou les "statuts héréditaires" ait pu être une solution optimum à la complexité des systèmes des rapports sociaux, économiques, politiques, etc. Cette observation a un rapport direct avec les arguments théoriques touchant les origines de stratification et le système d'état de gouvernement : notamment le travail de Service, Sahlins et d'autres en Amérique. Tandis que la question des origines peut être fascinante historiquement, elle peut être insignifiante du point de vue de l'importance théorique.


3The origins of the state and of civilization are disputed mysteries that have always fascinated philosophers and social scientists. In truth, it may well be that we have always known the right answers to the questions asked about the reasons, or causes, for state organization and for the elaborations of cultural life usually thought concomitant with the appearance of class differences, inequitable distribution of goods and services, differential exercise of power, and the like. As reviewed by Adams (1966), Carneiro(1970), Flannery (1972), Fried (1967), Krader (1968), Mair (1964), Service (1962, 1975) and other anthropologists, favorite solutions to these mysteries, nonetheless, have often been explanations that go beyond commonsense notions about the usefulness or "functions" of leadership personnel, about the troubles apparently inherent in dominance situations, etc. Scholars have been led to compare historical states and civilizations in hopes both that special intrinsic qualities or external circumstances possessed in common might be distilled from the particulars of time and space, and also that those special qualities or conditions would prove to be either the actual causes behind the origins of these mysteries, or at least the best clues to what the causes might ultimately be.

4Such an approach to the origins of the state and of civilization has an obvious weakness : it is a methodology of comparative assessment easily without controls. It can lead to one-sided solutions that mistake some particular circumstance--such as the presence of irrigation agriculture--for a cause sine qua non. It is better to adopt the position that we must ask not only "What is the origin of the state or of civilization ?" but also "Why are states not universal ?" In other words, what are the threshold conditions favoring the establishment of states and of civilizations by any means however simple or however complex--i.e., by conquest, by diffusion, by evolution, etc. Why is it that these threshold conditions have been historically so rare ?

5Kent Flannery, in arguing for the usefulness of computer simulation models for analyzing the cultural evolution of civilizations, notes what may be one of the most important issues involved :

6One of the thorniest problems in cultural evolution is the origins of hereditary inequality--the leap to a stage where lineages are "ranked" with regard to each other, and men from birth are of "chiefly" or "commoner" descent, regardless of their own individual capabilities (Flannery 1972 : 402). Flannery's phrasing of this "thorniest" of problems implies certain preconceptions about how societies evolve toward statehood which may or may not be historically correct, but which are not essential. What is important is the problem of ranking per se, not whether or not all societies must pass through a "stage of hereditary inequality" if they are to become states. This thorniest of problems has a solution which is, in fact, quite commonsensible. In order to review how this might be true, I would like to examine the functions of control networks by taking a specific ethnographic situation as a case study : the problem posed by the origins of social stratification among the Buin-speakers of southern Bougainville.


7Hierarchical control networks are one solution to the problem of what to do about systems that are becoming so complex, they are threatening to become unstable, unworkable, uneconomical, or any or all of these in combination. This premise seems to be one of the lessons taught by modern economic geography, the abstract theory of graphs, and contemporary thinking in evolutionary biology.

8Complexity, like ecosystem diversity, can be defined "as a function of the number of possible interactions in a system and the degree to which they are structured" (Johnson and Raven 1970 : 129). By "control" I mean the effect or interaction of one element within a system upon another. By a "hierarchical control network" I imply something more general than the ideal pyramidic chain-of-command within a modern bureaucracy. Along with Herbert Simon (1969 : 36), I refer to that sort of complexity seen in systems which appear to be composed of subsystems that are themselves composed of subsystems, and so on. I intend the term "network" to be taken simply, rather than in its technical sense as a directed graph which is connected and has no loops (Busacker and Saaty 1965). I mean only the vernacular definition, i.e., something having a complex arrangement or structure.

9The specific proposition lying behind the discussion to follow in this paper is inspired by Richard Levins' argument that "the dynamics of an arbitrary complex system will result in a simplified structuring of that complexity" (Levins 1973 : 113). The question I will explore is this one : Was the dynamics of political life in southern Bougainville leading to the evolution of hierarchival systems of social control prior to the arrival of the Europeans ? The specific proposition will be this : The need for social controls was not sufficiently complex in southern Bougainville before the coming of the white man that a system of ascribed leadership roles or "hereditary statuses" might have been an optimal solution to systems complexity in social, economic, political, etc. relationships. The evidence used to examine this premise will be taken both from social anthropology and from archaeology.

10Was political life in southern Bougainville leading to the evolution of hereditary social stratification ? This question takes us first of all to a controversy some years ago between two well-known social anthropologists : the late Richard Thurnwald, one of the fathers of economic anthropology, and Douglas Oliver, whose famous book, A Solomon Island Society : kinship and leadership among the Siuai of Bougainville (1955), is often considered to be one of the classics of modern social anthropology.


11Richard Thurnwald first visited Bougainville in 1908-1909 (Thurnwald 1909, 1910, 1912). He returned in 1933-1934 with his wife, Hilde (Thurnwald 1934a, 1934b ; Hilde Thurnwald 1934, 1937). They spent ten months living again among the Buin (also called the Telei or Rugara) people who inhabit the eastern side of the inland plain which fans out from the central volcanoes at the southern end of Bougainville (Fig. 1). In 1938 Douglas Oliver and his wife, Eleanor, arrived on the island : they carried out anthropological investigations for more than a year and a half also throughout southern Bougainville (Oliver 1943, 1949, 1955, 1968, 1973). They spent most of that time living with the Siwai or Siuai (also called the Motuna) who dwell on the southern plain just to the west of the Buin people.

12During their field work, both Thurnwald and Oliver were impressed by the nature of political life among these peoples and especially by the activities of the local "big-men" or leaders, called mumi by the Siwai and mumira by the Buin. Most of Thurnwald's accounts are in German, while Oliver's, of course, are in English. Oliver's fascinating reports have greatly influenced anthropological thinking about the status and role of Melanesian "big-men". Elman R. Service's recent attempt to account for the origins of the state and of civilization, for example, cites Oliver's Siwai as exemplars of what he judges to have been perhaps the most important single state in the evolution of the state : the institutionalization of power, i.e., the beginnings of hierarchical society (Service 1975 : 71-80, 291-294).

13As Service has himself phrased the argument, writing, mathematics, scribes, great specialized art, metallurgy, elaborate ceremonial religion, grand public monuments and other "specialized appurtenances usually attributed to the archaic civilizations" may be considered "the final benefits of a form of centralized and expanding political organization that began in the simple attempts of a "big-man" to perpetuate his social dominance by services to his fellows" (1975 : 308). More specifically :

14The Watershed in the evolution of human culture occurred when primitive society became civilized society. As we know from modern anthropological studies, primitive societies were segmented into kin groups that were egalitarian in their relations to each other. Eventually some of them became hierarchical, controlled and directed by a central authoritative power--a power instituted as a government (1975 : 3-4). ... it is a fact that segmental societies, however equal their parts, do exalt individuals. They follow war chiefs, accept advice from wise men, and believe in the unequal access of persons to supernatural power. And this proclivity sets the stage for more permanent hierarchies of differential power (1975 : 291). How does an influential person come to occupy an office, so that as his charisma wanes the office can be filled by someone else ? In other words, how does personal power become depersonalized power, corporate and institutionalized ? How does a high achieved statues become an ascribed status ? In more societal terms, the question is : How does an egalitarian, segmental society become an hierarchical society with permanently ascribed differential ranks of high and low statuses (1975 : 71-72) ?

15Elman Service and others recommend the Siwai of Bougainville as a good instance of an "embryonic chiefdom" capable of crossing this "Great Divide" between the primitive and civilized worlds by achieving a hierarchical form of governance (1975 : 74). This suggestion is interesting, because Richard Thurnwald has inferred that the Buin people at one time in their past has already made that great leap. Thurnwald's German writing style can be difficult to translate, yet it is unfortunate that Service does not refer to Thurnwald's accounts of "big-men" in Buin : unfortunate, because the process that may have led to social stratification among the Buin suggested by Thurnwald is different from that presumed by Service. Since the controversy between Oliver and Thurnwald centered precisely on this question of process, Service's elevation of the Siwai to the rank of an "embryonic chiefdom" potentially on the road to state organization and archaic civilization gives all the more reason to re-examine that controversy.

16Thurnwald and Oliver have both expressed the opinion that the status of "big-men" in Siwai differed from that of prominent leaders in Buin at the turn of the century. Before Thurnwald's death in 1954, however, they disagreed on the reason why. "Political organization among these Terei (Buin) is, indeed, more complex than in the neighboring Siuai area, but Thurnwald's explanation for it is, I firmly believe, quite overdrawn. His preoccupation with migrations of conquerors is in line with similar preoccupations of W.H.R. Rivers, but one need invoke no actual conquest by a 'superior race' to account for the development of chieftainship here ; trade contacts with Alu Islanders (i.e., people from the Shortland Islands in the strait south of Bougainville) could just as well have served to introduce the material objects and standards of value favorable for such an evolution" (Oliver 1943 : 61). "Dr. Oliver shows that the mumi of Siuai and vicinity acquired their status by a process that does not presuppose invasion and conquest by outsiders. This does not 'prove conclusively that this hypothesis (of mine) has no foundation' (to quote a remark by Ian Hogbin). It proves exactly nothing concerning Buin, a people living in a position exposed to foreign attacks..." (Thurnwald 1951 : 137).

  • 1 Thurnwald gives further details, also (1937 : 4-5). Frau Thurnwald has reconstructed the process o (...)
  • 2 As this paper was being completed, and after Kothleen Fine had typed the final draft of all but th (...)

17I will not attempt to do justice here to both sides in the Oliver-Thurnwald controversy. Indeed, arbitration is not apropos to what I would like to discuss. The dispute is taken as a case study not to determine who was right about the origins of social stratification in Buin, but instead :1 What are the alternative hypotheses about the past that might be extracted from what each of these scholars has written ?2 What can archeology contribute to the evaluation of these contrary hypotheses ? (3) What effect might the verification of one hypothesis rather than the other have on the broader issue raised by Service and others concerning the primitive origins of the governmental bureaucracies and the state ? What follows, therefore, will refer only to what shall be called the "Oliver hypothesis" and the "Thurnwald hypothesis", and will not attempt a detailed critique of all that has been said a-bout social stratification among the Buin and Siwai.

18In his recent book, Bougainville. A Personal History, Douglas Oliver writes : "Generally speaking, the more cohesive tribes (politically active neighborhoods) were to be found among the coastal Austronesian-speaking peoples, and in the Buin non-Austronesian area of southeastern Bougainville. Elsewhere tribes appear to have been smaller and more loosely organized. Thus regional differences in tribal size and cohesiveness may have been associated with differences in the ways men became leaders" (1973 : 71). According to his analysis, tribal leadership on Bougainville ranged between the two extremes of ascribed status, on the one hand, and achieved status, on the other. To quote him directly :

19At one extreme were those tribal neighbourhoods dominated numerically, or in terms of land-holdings, by one particular matrilineage. In such cases the members of the principal matrilineage constituted an aristocracy, and their senior member a hereditary chief, to be succeeded in time by the eldest son of his eldest sister (and not by his own son, who would of course have been a member of a different matrilineage). So far had this process gone in some coastal tribes, and in the southern part of Buin, that these societies reached the point of clear-cut stratification, having been divided into two or even three hereditary classes : aristocrats, commoners and intermediates (ibid.).

20In broad outline, this is the pattern of ascribed leadership encountered by Beatrice Blackwood on Buka and in northern Bougainville during her field work in the area in 1929-1930 (Blackwood 1935). It is also the kind of pattern drawn by Thurnwald for the Buin :

21The natives are an essentially sedentary people, although dwellings were sometimes moved as a result of warfare and personal quarrels. The communities are kept together by feudal chiefs, each of whom is the representative senior of his family in the district (1934a : 140).

22The other extreme-leadership by achieved status-was encountered among those tribes "whose leaders earned their positions of authority by exercising military or political skill. Usually however, actual fighting prowess was less important than the ability to gain and inspire followers, which was exemplified by forcefulness of personality and by shrewd distribution of favours and hospitality" (1973 : 71-72). Here Oliver seems to be referring to tribes such as those of the Siwai (Oliver 1955).

23Social evolutionists, following Elman Service, might classify Bougainville tribes in which leadership was an hereditary right as more "evolved" than tribes among which leadership had to be gained individually from generation to generation through good public service and skillful social, economic and political maneuvering. The facts about social organization on Bougainville do not fit such a neat ordering of political forms (Terrell n.d.). Among the Teop of northeastern Bougainville, for example, leadership roles are customarily filled by eldest surviving sons and daughters of particular "lordly" matrilineages. Yet it is freely admitted that an office can be filled by someone other than an heir-designate if the latter should be found to be incompetent, apathetic, or plainly inferior to someone else. And that someone else need not even be a kinsman, because it is easy enough to adopt a person so that he may fulfill the needed role.

24Enough has been said to warn the ingenuous social evolutionist that political life on Bougainville is not a matter of extremes. Oliver's characterization of the situation on the island as such is only a convenience. Elsewhere he has shown how truly difficult it is to classify the local tribes into clear-cut types. Writing, for instance, about only the Nagovisi, Siwai and Buin, all of whom dwell on the southern plain, he has reported that 'with the exception of certain dyadic groups--for example, trade partners, sibling pairs, friends--every group we observed in southern Bougainville was structured hierarchically : one or two persons exercised authority more frequently than other members" (Oliver 1968 : 163). Among the Nagovisi, who live to the northwest of the Siwai, hamlet leaders are (or were, traditionally) "Old-ones", male and female, who are normally also senior figures in the local matrilineages. "Kinship and age are the crucial qualifications for these positions, but a commanding or persuasive personality is also a factor" (ibid). Among the Siwai and Buin, on the other hand, where matrilineages are less obvious to the observer and less important in traditional life, "factors of personality and of 'renown' sometimes override age and even kinship affiliation as qualifications for hamlet leadership" (ibid).

25Oliver has suggested that variation such as this in the specifications for leadership roles may follow what biologists call a "cline", or gradient, running across the plain from the northwest where the Nagovisi live down to the southeast where the Buin people dwell. In western Siwai a man of renown, a mumi, is generally also a matrilineage leader, as in Nagovisi. In northeastern Siwai, however, whether or not a man rises to a position of influence may depend, in part, on whether or not his father was a mumi : in short, at this point on the southern plain one can begin to detect an incipient bias favoring patrilineal succession to mumiship(1968 : 163-164, 166). Finally, in the territory of the Buin people, the institution of leadership built upon renown exists "side by side with one of inherited rank" (1968 : 166).

26Lest the reader be confused about how leadership in Buin differs from leadership, say, among the tribes of western Siwai, it should be emphasized before going any further that one of the differences being drawn seems to be that between patrilineal versus matrilineal ascription, and not, as one might at first think, between acquired or ascribed status per se. An added difference may be also whether or not it is useful to regard those of high status as comprising a "chiefly stratum". This possible distinction, however, is an especially elusive one, because even Thurnwald was willing to say that a chief in Buin "is housed, dressed and fed exactly like his bondsman... The stratification, therefore, can only be discovered by close observation of the behaviour and customs, and by obtaining confidential information" (Thurnwald 1934a : 125).

27Without attempting to delve further into the obvious complexities of political organization on the Buin Plain, let it be said in summary that Oliver and Thurnwald seem to have agreed with each other that the institution of inherited rank was stronger among the Buin than the Siwai. In Oliver's words : "there is a clearly discernible cline, from northwest to southeast, in the devaluation of sibship. There is a parallel trend in emphasis upon 'renown' as a factor in determining social hierarchies, with the additional factor of inherited class-status in the southeast" (Oliver 1968 : 167). He goes on to ask, "What accounts for these regional differences ?" Since Oliver and Thurnwald have suggested two somewhat differing explanations, let us turn then to the hypotheses they have advanced.

28Thurnwald's hypothesis : Lawrence Krader in his book, Formation of the State, writes that Thurnwald "conceived the state to be formed by the conquest of one people by another" (Krader 1968 : 3). The Thurnwald hypothesis is, essentially, a conquest theory of state formation writ small. Phrased succinctly, Thurnwald asserted on a number of occasions that the "chief's stratum" in Buin had resulted from "an invasion of a tribe of black navigating conquerors" (Thurnwald 1951 : 138). As far as I have been able to discover, he never described the process or sequence of events in historical detail. On several occasions, however, he did speculate on what might have happened :

29Originally the mumira belonged to a stock that swept over Buin from the Alu and Mono Islands. Probably enterprising persons by head-hunting, successful fighting and feasts attracted followers among the "aborigines". As they took wives and settled among the indigenous population, racial and cultural fusion gradually advanced, although the progeny of the invaders reserved privileges for their kinfolk, thus establishing a kind of feudal regime (Thurnwald 1934a : 133).

30He suggested the original invasion had been carried out in a fashion comparable to the headhunting raids for which the islanders in the western Solomons were famous in the nineteenth century (Guppy 1887 : 16-17, 27).

  • 3 Thurnwald gives further details, also (1937 : 4-5). Frau Thurnwald has reconstructed the process o (...)

31"Derartige Raubzüge, die mit Schädeljagden verbunden waren, wurden noch bis vor etwa 50 Jahren, z. B. nach der Küste von Bambatana (Insel Choiseul) unternommen und wurden bei meinem Besuch dort 1908 noch gut erinnert" (Thurnwald 1937 : 4)3.

32Perhaps the most important feature of Thurnwald's hypothesis is the allowance it makes for the disintegration of the "traditional ethnic stratification" in Buin with the passage of time, due to intermarriage, the acquisition of wealth and renown by commoners or those of mixed-blood, etc. He wrote :

33It should be borne in mind that people (in Buin) promote their acquisition of wealth by using rational calculations of an economic nature. They do it consciously and intentionally, incited thereto by the desire to improve their social position. This process, by which the influence of the hereditary aristocracy was countered by the influence of wealth, began long ago... In this way the established stratification of society has, in the lapse of time, become disturbed and the principle upon which it was founded has shifted (1934a : 132).

34In other words, Thurnwald argued that the aristocratic Buin social system, built upon ethnic stratification and ascribed status, was in the process of devolving into one based upon the principle of acquired status. In this quotation he seems to imply, as well, that this gradual disintegration of the old order was not leading back to the kind of egalitarian matrilineal system that both he and Oliver (1943, 1955 : 470 ; 1968) have proposed may have once been universal on the plain. Instead, stratification remained, in weaker form perhaps ; only "the principle upon which it was founded has shifted".

35Oliver's hypothesis : Lawrence Krader has further observed that theories of state formation by conquest usually fail as general theories, because such interpretations rely upon factors external to a society and they do not necessarily take into account the internal processes leading to the formation of a given state. He argues :

36Migration of a bellicose people to the vicinity of a peaceable one or the converse, and subsequent conquest by the former of the latter, does not in itself lead to class stratification and state formation. There must also have been beforehand at least the germ of social stratification, of an administrative system, of an ideology of superiority and of rulership, and of a burgeoning economy with some differentiations of economic functions. The Eskimo and neighboring Chukchi, for instance, made war upon each other, with occasional conquests, but we do not speak of a Chukchi or Eskimo state (Krader 1968 : 45).

37Krader goes on to comment that Thurnwald's ideas about the formation of states through conquest in east Africa did not, in fact, suffer from such obvious deficiency (1968 : 49-50). Douglas Oliver has, however, indicated such a weakness in Thurnwald's explanation for the alleged stratification of Buin society :

38...Thurnwald's explanation for the class-stratified society of the Rugara-speakers (Buin) is in terms of alien conquerors, Melanesian-speaking warriors from the islands to the south, who imposed their regime upon the more primitive and less organized aborigines. This may be so, but it is not the only possible explanation. An alternative hypothesis is that the Rugara institution may be viewed as an extreme but logical variation on the prestige-ranking theme, a crystalization in dynastic form of beliefs and practices present, incipiently, elsewhere. Frequent contact by the Rugara-speakers with aristocratically organized outsiders may have provided a reference-model for an otherwise local development made possible by larger surpluses and more shell money, traded directly from the southern islanders (Oliver 1968 : 167-168). It can be seen, I believe, that these two hypotheses are not altogether dissimilar. Both Thurnwald and Oliver have written that the Siwai were less exposed to outside contact with Strait islanders than the Buin. In 1951 Thurnwald remarked : "In earlier writings I have commented on the society of Siuai, which differs decidedly from that of Buin. I assume that the Melanesians did not enter the hill country, but kept near the coast, although they maintained indirect relations with Siuai" (1951 : 138). "Cultural differences between the two areas do, however, exist. Many of these I believe can be attributed to the proximity of the Terei (Buin) people to Alu Island" (Oliver 1943 : 61). "There may have been some coastal trade with Alu and Mono islanders, but this cannot have been very lively inasmuch as the Siuai had few material surpluses to offer in exchange and inasmuch as the islanders did not to my knowledge offer a large market for slaves" (Oliver 1955 : 470). Similarly, they have both discussed the process leading to stratification in Buin as one involving a give-and-take between foreign and indigenous elements.

39The major difference between the Thurnwald hypothesis and the Oliver hypothesis thus seems to lie in the immediate cause or events held responsible for the conversion of Buin society away from matrilocal, matrilineal institutions and toward "a strengthening of patrilineal political ties" (Oliver 1943 : 61). Oliver credits stimulus diffusion and suggests that class-stratification evolved in part because of contact with an already stratified society, that of the Mono-Alu or Shortland Islanders. Thurnwald also thought Buin society had changed from a formerly more egalitarian condition, not so much by evolution due to the impetus of external ideas and trade, but by force applied by external invaders from the same source. Stratification did not evolve locally, it was imposed.

  • 4 As this paper was being completed, and after Kothleen Fine had typed the final draft of all but th (...)

40In both hypotheses, chance plays a dominant role. If the Strait islanders had happened to be less warlike or less stratified, the situation in Buin might have developed along different lines. I must confess that I prefer Oliver's hypothesis, not because it is antagonistic to what seems a rather old-fashioned "invasion hypothesis", but instead because it seems a more convincing phrasing of very much the same set of circumstances. Both hypotheses have weaknesses. Both are speculative ; both presume a degree of class stratification in Buin which might be challenged in fact4. Thurnwald, lamely I think, had to argue against his own observations, in order to argue for an aristocratic tradition and class in Buin : i.e., he had to claim that the mumira class had "deteriorated", "disintegrated", suffered "fusion" with the aboriginal natives, etc. As Oliver seems to hint now and then in his writings (e.g., Oliver 1943 : 62, note 29), Thurnwald's eye-witness observations, indeed, could be read to suggest a structural picture of Buin political society little different from that painstakingly drawn by Oliver for the Siwai. On the other hand, Oliver's hypothesis presumes certain differences in human ecology between the territory of the Siwai and that of Buin which may not exist (Terrell 1976 : 154-155), and it further supposes an "aristocratic tradition" in the Bougainville Strait which cannot be assumed without qualification. Guppy reports that "the chiefs of the islands of Bougainville Straits possess far greater power over their peoples than that which is wielded by most of the chiefs we encountered at the St. Christoval end of the group" (Guppy 1887 : 20). Yet I have argued elsewhere (Terrell and Irwin 1972) that much of the power and economic wealth of "King" Gorai and other chiefs in the Strait in the latter half of the nineteenth century can be interpreted as a direct result of their entrepreneurial dealings with European traders. Moreover, even these remarkable big-men were described by visiting white men as little distinguishable by appearance from their "subjects". Gorai did have by repute a great number of wives, and Guppy relates that the "inhabitants of the Shortland Islands, Gorai's immediate rule, live in great awe of their chief". Yet he adds that "We were unable to see very much of the mode of exercising his power ; but I suspect that Gorai, like other chiefs, places but little value on the lives of his people (1887 : 22).

41I have ventured far into the arena of social anthropology. I have gone as far as I have to emphasize some of the complexities and uncertainties of the problem of social stratification in southern Bougainville. Let me suggest then a few of the things we might ask about the past, given the Thurnwald and Oliver hypotheses as I have presented them.

42If the Thurnwald hypothesis is correct, it seems reasonable to expect to find these characteristics in the genetic composition of the Buin villagers and in the archaeological record from this part of Bougainville :

431. According to one of Thurnwald's earliest papers (1910 : 101-110), the Buin population can be divided into two racial types which correlate more or less discretely with the chiefly class and the aboriginal lower class. "Natürlich kommen zwischen diesen beiden eine Menge Misch-Typen vor. Der erste Typ is dem anderen wohl an Intelligenz und Kulturgütern überlegen, doch hat der letztere es durchgesetzt, seine Sprache in Buin zu erhalten. Die Leute von Buin sprechen eine nicht melanesische Sprache, die mit der Bergvölker (i.e., the aborigines) verwandt ist" (1910 : 101).

442. Since the black, seafaring invaders are alleged to have introduced important "high" culture traits, the archaeological record may show the sudden appearance of new culture elements a few hundred years ago--Hilde Thurnwald (1938 : 214) has speculated that the initial marauding began about 200 years ago.

453. If it is possible to identify certain monuments, artifacts, settlement traits, etc. as introduced traits belonging to a new chiefly class, we may anticipate finding that such elements have become more general--i.e., common--with the passage of time, as the aristocracy has gradually disintegrated and commoners or men of mixed-blood have begun to achieve high status in competition with the old aristocrats.

46Alternatively, given the Oliver hypothesis, the following patterns may be expected :

471. If intermarriage has taken place between Strait islanders and the Bougainville tribes, we might find that the Buin people are genetically more similar to the off-shore islanders than either the Siwai or the Nagovisi are discovered to be. However, given Oliver's hypothesis, there is no reason to suspect extraordinary genetic variability within the Buin population, due to the occurence of two ethnic strata with only partial "racial mixture".

482. If similarities between the Buin tribes and the Strait islanders have come about merely through borrowing and trade, there is no reason to expect to find in the archaeological record a sudden influx of culture traits a few hundred years ago. Once contact between landsmen and off-islanders was established, traits might have passed back and forth at any time in the past. Even the sudden appearance of a complex of traits within the archaeological record, moreover, might be explained as wholesale borrowing.

493. New traits borrowed from Strait islanders from time to time may have become fashionable for awhile and then gone out of fashion. There is nothing in the Oliver hypothesis to lead us to anticipate a major "cultural decline" due to the disintegration of an old order.

50Shortly I shall turn to archaeological evidence bearing on these possibilities. Before doing so, however, it is appropriate to mention briefly something about the results of modern biological research on the genetics of Bougainville islanders undertaken by Jon S. Fried-laender and others from Harvard University. While Friedlaender has thus far not attempted biological surveys in the Buin area, he has completed systematic studies among the Siwai and other linguistic groups in southern Bougainville and, most recently, on Buka and selected areas of northern Bougainville (Friedlaender 1975 and pers. comm.). Following European contact, there was a marked decline in the native population of the Strait islands (Irwin, pers. comm ; Thurnwald 1910), and much of the present-day population is comprised of migrants from Bougainville. Thus it may be now more or less impossible to compare Strait islanders and plainsmen directly. Friedlaender has found that Torau-speaking Melanesians living on the east coast of Bougainville, after whom Thurnwald once named his black invaders (Thurnwald 1937 : 4), can not be distinguished readily from Bougainville islanders in general and, in fact, most closely resemble their present neighbors in the Kieta area, the non-Melanesian-speaking Nasioi. Although it is likely that the Torau have been residing on the east coast for hundreds of years (Black n. d. ; Terrell and Irwin 1972), no one doubts their affinities with the islanders in the Bougainville Strait. Such circumstantial evidence, therefore, seems to imply that Thurnwald's racial argument has no basis in fact (Friedlaender, pers. comm.).

51What, therefore, does archaeology have to say about the Oliver and Thurnwald controversy ?


52In 1969-1970 I carried out archaeological surveys on Bougainville in four selected areas, including the Paubake Survey Area located in Buin territory (Fig. 1). During the final three months of the expedition which were given over entirely to work in Paubake, I was joined by two New Zealand archaeologists, Dr. Geoffrey Irwin, presently Lecturer in Anthropology at the University of Auckland, and Kenneth C. Gorbey, now Director of the Waikato Museum in New Zealand. Since the monograph on these researches is to be published in the near future, I will only summarize points of direct relevance here (Terrell 1976).

53At various times in his narratives on his journey to Buin in 1908-1909 Richard Thurnwald mentions the existence of certain isolated stone boulders and stone groups which locally had supernatural associations (1909 : 518-519 ; 1910 : 135 ; 1912, Vol. 1 : 369-370). He speculates whether or not they might have served in the past as grave stones. In this regard he reports that a stone near the village of Morou was alleged to be the burial place of a "big-man" named Tsikinue (1912, Vol. 1 : 370). He notes that it seemed revealing that Tsikinue was said to have been buriedthere, because the historic Buin villagers practiced cremation (1912, Vol. 1 : 33). On his return to Buin in 1933-1934 he undertook further study of these mysterious stones (Thurnwald 1934b). He made rough field surveys of stones about the villages where he and his wife lived, and he even "ventured to dig underneath a few megaliths without encountering any objection from the natives".

54He discovered that most, if not in fact all, of the stones "are never hewn, or bear any other traces of human workmanship". He thought he could distinguish three types of monuments :

551. Megaliths which varied considerably in size ranging from roughly 30 x 30 inches up to 70 or 80 inches square, if cubic in shape. "The larger blocks are supported by six basic stones, comparatively very small in size. Little blocks are supported by three or four basic stones only".

562. Monoliths "standing erect and sometimes bearing traces of human sculpturing, either in in the shape of a prism or a rectangle ; in one or two cases the profile of a face could be guessed".

573. Stone circles or ovals : "Comparatively small stones of a few inches in diameter arranged in a circle or oval associated with traces of cremation. They are apparently of no great age".

58It is unclear how many of the so-called megaliths he may have excavated in 1933-1934. His account reads as if he were describing only a single example :

59A few inches beneath the surface I found remains of nutshells, round objects like throwing stones, and broken pottery. Somewhat deeper, a few inches more, human bones in an advanced stage of decomposition were discovered, together with broken stone implements, crude axe blades, and seemingly wooden objects which it is scarcely possible to reconstruct.

60He indicates that the finds were handed over to the Department of Anthropology at the University of Sydney, but they are no longer to be found there (Specht 1969 : 11).

61Thurnwald was intrigued by the human remains : "Perhaps the stones hide the remains of an invading population from which the present chief's families are late hybrid descendants". He was somewhat perplexed by the condition of the burial :

62The position of the human remains shows that the body had not been buried as a whole. The bones were lying in a jumble, not arrayed like a full skeleton. No skulls or teeth were found, though some of the pits were thoroughly examined, the basic stones and the megalith removed, and digging continued below the layer of bones. My impression is that a low pit was originally excavated and then filled with bones which were sometimes placed on one side of the megalith rather than directly under it. For some reason the skulls must have been withheld from the grave.

63He concluded that the transport of such large blocks must have taken a great deal of strenuous exertion on the part of a large number of men : "...certain persons must have been able to urge others to help in the performance of the work ; and these others must have been willing to concur with them. This suggests an enterprising people". Citing reports of similar stone monuments in Siwai, on the islands in Bougainville Strait and to the north on the island of Buka, he ends his report with the thought : "On other islands of the Pacific also, monuments are found, the investigation of which would throw some light upon the migrations and early history of this part of the world".

64The primary objective of most archaeological surveys is not to answer questions about the past, but rather to determine the most interesting questions to ask. Solutions usually come later from detailed, intensive research directed toward carefully phrased research problems. The objectives of the archaeological surveys on Bougainville in 1969-1970 were simply to define some of the outlines of prehistory on the island, the largest in the Solomon Islands, and gather evidence which might prove useful in accounting for the great ethnological and biological diversity of these islanders, an ethnic complexity so extraordinary that Bougainville is a microcosm of the diversity not only of the whole of the Solomons, but also of the whole of Melanesia. If the tentative nature of our findings is kept firmly in mind, the following observations bearing upon the prehistory of the Paubake area can be offered.

65Investigation of stone monuments in Buin, it now appears, has nothing to contribute to the study of so-called "Megalithic folk" alleged by some scholars to have migrated throughout the Pacific bearing advanced culture traits as their gifts to backward, short, stolid "aborigines" (e.g., Riesenfeld 1950). We could find only two "monoliths" in the Paubake Survey Area ; none was obviously sculptured. While the possibility cannot be ruled out that Thurnwald's monoliths on which "the profile of a face could be guessed" might somehow be related to the famous sculptured burial pillars of Choiseul (Bernatzik 1935 : Fig. 43), there is otherwise no reason to think that the stone monuments of Buin are anything more than local inventions. Evidence favoring this conclusion is afforded by the distributions of both "megaliths" and "stone circles".

66We discovered over 70 stone megaliths (PI. 1a), ranging in the length and width of the capstone between roughly 30-400 cm, and in height above ground level between 15-200 cm. The "basic stones" or underlying support stones varied in number from between none at all to 9 in total. It is clear that the function of the understones was merely to level the primary capstones and that the number employed in a specific instance is not terribly significant. In order to avoid the presumption that these table-like stone constructions reflect a Megalithic Culture, we have termed them "capstones".

67The distribution of such finds is revealing. While Thurnwald was surprised by the occurrence of hugh blocks of stone on an alluvial landscape (1910 : 135 ; 1934b), the distribution of capstones, singly, in groups of 2-3 and more, or in impressive long avenues (Fig. 2), appears to correlate precisely with the distribution of naturally-occuring stones of a similar nature (Fig. 3). As reported by the soil scientists at the Australian Division of Land Research (Scott et al. 1967), the northern region of the Paubake Survey Area depicted as unshaded on the maps in Fig. 3 & 7 is a portion of what they have called the "Buin Land System"--the alluvium of which contains eroded boulders of volcanic origin identical to those used in the construction of capstones. Boulders and smaller stones are to be found, in fact, on or partially within the ground surface throughout the northern part of the survey area. Where rivers have cut through the alluvium, it is often possible to find huge boulders buried deep in the ground or lying exposed in river channels (Pl. 1b). Frequently the upper surfaces of capstones and naturally-positioned boulders have a distinctive "pocked" appearance. Locally such stones may be identified as "nutting rocks" where generations of villagers have cracked open Canarium almonds. Perhaps some of these stones have been used in this fashion, but the pitted surfaces are the result of the weathering of phenocrysts which are part of the volcanic composition of the rock.

68Originally we assumed that these numerous capstones were a kind of burial monument, as Thurnwald had inferred they were. Excavation below two small capstones at the Loiai Site (BoP-3) near Luaguo Village, however, failed to discover pits or burial remains of any description. With rare exception, no one in Buin today can say what function these monuments may have served. Some reported, nonetheless, that the capstones were "pudding tables" used in food displays at feasts. The best hypothesis at the moment appears to be that Thurnwald was only partially right in his interpretation.



Fig. 2

69It is my suspicion that Thurnwald was correct to be impressed at the labor which must have gone into building especially the larger capstones. It seems a good guess that they are prestige symbols directly comparable to the great timber "slit gongs" that "big-men" in both Buin and Siwai commission to have made for them as monuments to their growing renown and which are transported by their followers to their club-houses with pomp and feasting (Oliver 1955 : 379-386 and Figs. 38-40 ; Thurnwald 1910 : 114-115). "Gong-carrying is one of the Siuai's most spectacular activities. One occasion I witnessed involved some two hundred men ; and a twenty-five-foot-wide trail had to be cut through the forest and grove to get the gong to the club-house" (Oliver 1955 : 385).

70The fact that Thurnwald may have found human remains below the capstones he excavated does not necessarily refute this sociological hypothesis. As far as we can tell at present, capstones and the types of burial structures called stone circles by Thurnwald were, in reality, contemporary culture traits during at least part of their periods of fashion in the prehistory of the Paubake area. In other words, it appears that Thurnwald was wrong to suppose that a practice of inhumation below megaliths came before cremation burial in stone circles. Nonetheless, it seems entirely plausible that the burials he found were the dedicatory remains of victims slain at the inauguration of new capstones : a romantic notion, perhaps, but a practice actually describe by Thurnwald as part of the inauguration of chieftain's halls (Thurnwald 1912, Vol. 3 : 51-53). It is at least intriguing that he reports skulls were removed from the hapless victims so that such trophies could be displayed in the new club-houses. In any case, it remains for future excavations to determine whether or not capstones are likely to be found in association with club-house remains : an association which would strengthen this particular hypothesis .

71Thurnwald's "stone-circles" were also investigated in 1969-1970. It is now possible to recognize three general types, which we have designated : (a) the rectangular type with prominent corner-stones or "wings" (Fig. 4) ; (b) the simple rectangular type (Fig. 5) ; and (c) the irregular type, apparently indentical to what Thurnwald called stone circles (Fig. 6).

72Once again, these stone monuments are also found in the north where there is natural stone readily available for their construction--including in this case, however, the stone outcrops at Malabita Hill near the coast (Fig. 7).

73Unlike Thurnwald, we did encounter some opposition to the excavation of these burial grounds, which today are locally called tsigoro . By agreement, we excavated only a single tsigoro at the Loiai Site (Fig. 8). As chance had it, however, we did come across burials in other excavations. It is thus possible to sketch tentatively the history of burial rites over a period of perhaps a millennium or more.

74Formerly, the word tsigoro seems to have meant only a place where a funeral pyre (tsigo) has burned. It has now replaced the traditional word for grave site today familiar only to a few old people, the term tiririno or tsiririno. Thurnwald tells us (1912, Vol. 3 : 22-23) that after a cremation, the remaining bones and ashes were collected in a taine or carrying-bag and then buried in a defined cemetery or tsiririno. A wooden framework called a bare was erected at the place where the cremation had occured, a place he says was called a tsigo. Food offerings to the dead were then burned at the bare each day for a month after the funeral.

75The tsigoro in Area 2 at Loiai was excavated because it seemed to be a well-preserved example of the winged-rectangular type of burial site which--like all other stone tsigoro--was said to be the burial place of a "big-man". First, two scattered human cremations were discovered in the topsoil within the tsigoro (Bone Scatters A and B in Fig. 8). Further excavation revealed a primary burial pit well below the zone of active soil formation. Within the pit we found an urn burial of (probably) a man estimated to be 20-40 years of age (Fig. 9 and Pl. 1c). Associated with his remains were drilled canine-teeth of an as-yet-unidentified mammal, presumably from a necklace, which had been burned along with the body. A stratigraphic section along the north wall of the tsigoro confirmed the added fact that the burial structure had been erected in an area previously the site of a poststructure (s), perhaps a house of some description. Of particular interest were, however, a number of small post-holes more or less around the burial pit, at least some of which post-dated the infilling of the pit.

76After we excavated this urn burial, local people did tell us that they had heard of such things. But this burial rite is, as far as I can tell, totally unreported in the ethnographic record for the southern Bougainville plain. At the time of excavation, we did not know of the Buin custom of constructing a bare or framework at the site of cremation. The small post-holes just mentioned which were found around the primary burial pit might, of course, be so interpreted. The fasccinating point, nonetheless, is that this kind of urn burial appears to be paralleled very closely by the type of burial given high-status individuals on the islands in the Bougainville Strait and in the Melanesian-speaking Torau villages of the east coast of Bougainville : localities where, apparently, cremation was a rite reserved for persons of rank, while lesser individuals were interred or buried at sea (Frizzi 1914 : 12-14 ; Guppy 1887 : 51-52 ; Parkinson 1898-99 : 9 ; Parkinson 1907 : 484 ; Wheeler 1914 : 64-78 ; Woodford 1890 : 37).

77This parallel is surprising because of the suspected age of the Loiai burial urn. My luck in getting radio-carbon dates that are easy to interpret is not very good. There are two for this obviously well-defined event : 1710 B.C. ± 180 (GX-2218, uncorrected) and 1140 A.D. ± 130 (uncorrected, GX-2219). I think the latter comes closer to the truth, particularly if seemingly related evidence from the Shortlands, including one carbon date of 1040 ± 95 B.P., is taken into consideration (Irwin 1972 : 103). Whatever interpretation is made of these carbon dates derived from two different charcoal samples, it does look as if we should think in terms of an age of at least 600-1, 000 years.

78If this one tsigoro is typical of the type we have called the winged-rectangular tsigoro, then it appears that this ancient burial pattern combines elements both of historic Buin tsiririno and bare, and also of the historic burial customs followed on the islands in the Strait but not found in Buin or Siwai (Oliver 1955 : 212). In sum, the Strait islanders during the early decades of European contact were still practicing burial customs seemingly little changed from the pattern suggested by the Loiai Area 2 tsigoro. The Buin (and Siwai ?) in historic times, on the other hand, were following customs that were only a dim reflection of their own ancestral pattern.

79I should add that there is one obvious difference between the old Buin pattern and the historic rite in the Bougainville Strait. Historic monuments seemingly "cognate" to stone tsigoro were wooden structures built at the site of cremation. Considering that tsigoro occur in Buin where there is stone readily at hand for their construction, this apparent difference seems insubstantial. Yet it is a unique characteristic of the Buin rite worth keeping in mind (note however : Ribbe 1903 : Fig. 15).

80Now if the historic Buin burial pattern was, in fact, a derivative of the ancient tsigoro pattern found at Loiai, then it may be possible to document the course of change or drift in the archaeological record. It appears that archaeology can be used to do just that. But it must not be forgotten that the evidence available is still limited. Perhaps more important, the changes we can infer may have occurred only in the burial tradition for high-status individuals.

81When the dimensions of the area enclosed within a tsigoro are plotted for every tsigoro found in 1969-1970 which could be measured (Fig. 10), it is apparent that the irregular type is, on the average, smaller than either the winged-rectangular type or the simple rectangular type. Moreover, rectangular tsigoro of either form are rare ; irregular tsigoro are more common. All three types, nevertheless, are uncommon enough to suggest they were not the sort of monument built for every man or woman in Buin : i.e., they were, as they are today said to be, burial sites of individuals of high-status. Further, irregular tsigoro are normally found clustered together, as in Figure 6. In those cases when two or three types occur at the same cemetery, as at Turiboiru (Fig. 11), it sometimes looks as if an original winged-rectangular or simple rectangular tsigoro was later added to through the construction of more irregular tsigoro walls, as at Nigeriai (Fig. 12). Our present interpretation of these observations is that winged-rectangular type is the oldest form, the simple rectangular type is a close derivative from the winged variety, and the irregular type is a further derivative from the simple rectangular type. It seems reasonable at this point to accept the local claim that the individuals buried in grouped tsigoro were all patrilineally related, or were at least kinsmen. This assumption then may explain the apparent "additive" configuration of many of the tsigoro found in Paubake.

82The excavation of three cremation burials quite by accident in 1970 at the Bekuinotu Site (BoP-34), seen in conjunction with the cremation burials found scattered in the topsoil at several locations at Loiai, permits a tentative extension of this proposed sequence of changes in tsigoro ritual down to historic times. The burials at Bekuinotu were all in basin-shaped pits, as at Loiai. But the cremated remains were not in pots or urns. They were not marked with stone tsigoro walls. While they have been dated as "less than 200 years" (GX-2216, GX-2217), stratigraphically the oldest burial is associated, nonetheless, with a tsigoro-like ring of small cobbles found actually within the burial pit. The two other cremations lacked even this feature. Now the stratigraphic position of the cremation scatters at Loiai certainly seems to imply that they may be even younger than the Bekuinotu burials. Thus it looks as if burial ritual in Paubake (assuming we are not mixing high-status and low-status burials) went through a course of "devolution" starting with changes in the construction of tsigoro(hypothetically), at least, a change from winged-rectangular to simple rectangular to irregular tsigoro), leading to the eventual cessation to tsigoro construction altogether (except perhaps in the form of a small ring of stones within the burial pit), and ending with the practice of simply scattering cremations over the ground in a place designated as a tsiririno, resulting in the incorporation of cremations into topsoil. At some point, too, the use of burial urns must have ended, but we cannot even guess when that occurred.

83A similar and far more reliable "devolutionary picture" could also be drawn tracing the history of pottery-making in the Paubake Survey Area (Terrell 1976). It seems likely that pottery appeared in Buin about the same time that we also find the development of tsigoro and capstones. There is no evidence to infer, however, that these three culture traits appeared in Buin at one and the same time, although our knowledge is still so fragmentary that we cannot rule out the possibility altogether.

84What does this all mean ? More to the point, having gotten this far from the topic which began this narrative--the origins of the state and of civilization--can we get back to that larger issue ?


85There is reason to be confident in the belief, based not only on the new archaeological knowledge of Bougainville but also on our rethinking of earlier theories and beliefs, that trade, travel, settlement and marauding among the islands in the Bougainville Strait (including Choiseul) and the tribes of southern Bougainville all have a most respectable antiquity behind them : at least 1,000-1,500 years by present estimate, if not even farther back in prehistory (Black n.d. ; Irwin 1972 ; Terrell and Irwin 1972). Thurnwald and others have been wrong, I think, to believe instead that contact between Buin and the Shortlands was established only recently and was necessarily hostile. It appears certain from Thurnwald's writings that he was led himself to speculate about a Mono-Alu invasion in part because of tales and traditions locally popular concerning events in the latter part of the nineteenth century when "King" Gorai and other Mono-Alu chieftains were expanding their hegemony over the southern shores of Bougainville. His error lay in reading too much into those events.

86Most of Thurnwald's "invasion hypothesis", therefore, is now entirely suspect. It is improbable that we will ever be able to define a "period of invasion". Even if an invasion of some kind did occur at some point in the past, it is unlikely that the conquerors were in fact racially distinct from the Buin "aborigines". On a priori grounds alone it seems most unlikely that we could now identify a foreign genetic component in Buin or that the invaders swarmed into Buin in numbers large enough to establish a chiefly stratum which was also an ethnic stratum.

87Yet there is now also enough archaeological evidence from the Paubake Survey Area to believe that contacts between Buin and the Strait islanders have indeed been influential in the course of local prehistory. I would argue, however, that in every case now known where borrowing or conquest might be surmised we see this distinctive pattern : Buin culture traits such as pottery, burial rites, prestige pudding-tables, and the like may owe some inspiration to external sources (or vice versa ?), but they are, nonetheless, local expressions or realizations. Culture traits appear in the prehistoric sequence, change and, in some instances, are forgotten : all in unique ways.

88I have already observed that one of the characteristics of culture change in Buin seems to be "devolution" or drift in isolation. As Oliver long ago remarked, the broad moat of swampland at the coast throughout most of southern Bougainville has probably been a barrier discouraging intensive contact with the outside world. Geography favors the conquest of settlers and conquistadors in this part of the world, regardless whether they arrive on friendly or unfriendly terms. As I have previously argued (Terrell 1972a), there is only a limited amount of habitable land directly at the southern shores. Once coastal settlers move inland behind the swamps, either because of inclination or "population pressure", they are faced with the same environment, regardless whether they be friend or foe. They become isolated by the broad swampland from ready contact with the coast. Given time, it seems inevitable that they will be assimilated by the "aborigines" and lose their separate identity. Any innovations they may bring about in Buin culture may also suffer absorption, re-interpretation and perhaps ultimate obliteration. Moreover, changes in Buin probably tend to be felt for only short distances away, because of the difficulties of travel across the southern plain, which is greatly dissected by streams and rivers flowing down from the central volcanoes.

89What about "big-men" ? If my interpretations of capstones and tsigoromonuments are correct, it does look as if what Oliver has called "renown" has been around in Buin for a very long time. It does not seem unreasonable to suspect that tiririno or cemeteries comprising a number of a nearby or contiguous tsigoro, such as those in Fig. 5, 6, 11 and 12 are just what they are today locally said to be : the burial places of "big-men" and their sons. It does not seem necessary to belador the likely "patrilineal" implication of such a cemetery configuration.

90Was the (apparently) hereditary leadership system of the Buin established by conquest, or by foreign inspiration and local potentiality ? It really does not seem to matter. In either Oliver's interpretation or in Thurnwald's interpretation the eventual result is the same : "the institution of feast-giving for renown exists side by side with one of inherited rank."

91But are the mumira "big-men" ? Or have they made the great leap in the evolutionary direction toward statehood and civilization ? I think Service and others are in error to emphasize the origination of ascribed offices as an important advance in the evolution of civilizations, i.e., as the origin of the state. Jay Callen argues that anthropologists writing about "primitive" political associations generally and Melanesian "big-men" in particular have usually laid emphasis on the personalities of politics. They have not given sufficient attention to what leaders do for those they lead. I believe much the same criticism could be made of Service's proposition.

92Using data drawn from Oliver's monumental study of the Siwai, Callen suggests that the spatial distribution of leader's settlements in Siwai before World War II was not random and, in fact, the distribution of politically significant villages there can be predicted using Christaller's Central Place Theory. He asks :

93What goes into the making of a "big-man" ? The traditional replies by Pacific anthropologists stress a leader's personal qualities (ambition, charisma, generosity, cunning, etc.) and the small core of followers, usually close kinsmen, who support his political career. "Big-men", however, are members of a political organization which displays a spatial as well as a sociological structure. It is this spatial patterning of political phenomena which suggests that, in Siwai, leaders were as much a function of the central places they inhabited as vice-versa. In a certain sense, potential political centers may be said to have "created" the "big-men" to occupy them (Callen 1976 : 23).

94This last thought brings me back full-circle to the beginning of this discussion and to the premise introduced at that time. I think Service is right to see the origins of the state exemplified by institutions such as that of competitive feast-giving among the Siwai. But what is portrayed by such social-climbing is not the evolutionary potential of getting hold of an office and keeping it in one's family. Instead, I think what is revealed is simply that that is all the use the Siwai have for leaders. Their political system is no more complex apparently than it need be. If "hierarchical control" among the Buin is, in truth, more elaborate, then the question to be asked is not : "What are the origins of social stratification in southern Bougainville" ? Instead, it is : "What purposes does it serve" ?

Pl. I. - a) Stone megalith, b) Huge boulders buried deep in the ground or lying exposed in river channel, c) Burial urn of probably a man estimated to 20-40 years of age.


DOI are automaticaly added to references by Bilbo, OpenEdition's Bibliographic Annotation Tool. Users of institutions which have subscribed to one of OpenEdition freemium programs can download references for which Bilbo found a DOI in standard formats using the buttons available on the right.


ADAMS, R. McC. (1966) : The Evolution of Urban Society : early Mesopotamia and prehispanic Mexico. London : Weidenfeld and Nicolson.

BERNATZIK, H.A. (1935) : South Seas. New York : Henry Holt and Company.

BLACK, S. (n.d.) : "The excavation at Teobebe, Teop Island". Reports of the Bougainville Archaeological Survey, 10. Chicago : Department of Anthropology, Field Museum of Natural History.

BLACKWOOD, B. (1935) : Both Sides of Buka Passage : an ethnographic study of social, sexual, and economic questions in the north-western Solomon Islands. Oxford : Clarendon Press.

BUSACKER, R.G. and SAATY, T.L. (1965) : Finite Graphs and Networks : an introduction with applications. New York : McGraw-Hill Book Company.

CALLEN, J. (1976) : "Settlement patterns in pre-war Siwai : an application of Central Place theory to a horticultural society". Solomon Island Studies in Human Biogeography, 5. Chicago : Department of Anthropology, Field Museum of Natural History.

CARNEIRO, R.L. (1970) : "A theory of the origin of the state". Science, 169 : 733-738.

FLANNERY, K.V. (1972) : "The cultural evolution of civilizations". Annual Review of Ecology & Systematics, 3 : 399-426. DOI : 10.1146/

FRIED, M.H. (1967) : The Evolution of Political Society : an essay in political anthropology. New York : Random House (Studies in Anthropology).

FRIEDLAENDER, J.S. (1975) : Patterns of Human Variation : the demography, genetics and phenetics of Bougainville islanders. Cambridge, Mass. : Harvard University Press.

FRIZZI, E. (1914) : "Ein Beitrag zur Ethnologie von Bougainville und Buka mit Spezieller Berücksichtigung der Nasioi". Baessler-Archiv, Beiheft 6. Leipzig and Berlin : B.G. Teubner.

GUPPY, H.B. (1887) : The Solomon Islands and Their Natives. London : Swan Sonnenschein, Lowrey & Co.

IRWIN, G.J. (1972) : An Archaeological Survey in the Shortland Islands : B . S . I. P. Unpublished Master's Thesis, University of Auckland.

JOHNSON, M.P. and RAVEN, P.H. (1970) : "Natural regulation of plant species diversity". Evolutionary Biology, 4 : 127-162.

KRADER, L. (1968) : Formation of the State. Englewood Cliffs, New Jersey : Prentice-Hall, Inc. (Foundations of Modern Anthropology Series).

LEVINS, R. (1973) : "The limits of complexity". Pp. 109-127 in : H.H. Pattee (ed.), Hierarchy Theory : the challenge of complex systems. New York : George Braziller (The Inter-national Library of Systems Theory and Philosophy).

MAIR, L. (1964) : Primitive Government. Baltimore, Md. : Penguin Books.

OLIVER, D.L. (1943) : "The Horomorun concepts of southern Bougainville : a study in comparative religion". Pp. 50-65 in : C.S. Coon and J.M. Andrews, IV (eds.), Studies in the Anthropology of Oceania and Asia. Cambridge, Mass. : Papers of the Peabody Museum of American Archaeology and Ethnology, Harvard University, 20.

OLIVER, D.L. (1949) : Studies in the Anthropology of Bougainville, Solomon Islands. Cambridge, Mass. : Papers of the Peabody Museum of American Archaeology and Ethnology, Harvard University, 29, Nos. 1-4.

OLIVER, D.L. (1955) : A Solomon Island Society : kinship and leadership among the Siuai of Bougainville. Cambridge, Mass. : Harvard University Press.

OLIVER, D.L. (1968) : "Southern Bougainville". Anthropological Forum, 2 : 158-179. DOI : 10.1080/00664677.1968.9967225

OLIVER, D.L. (1973) : Bougainville : a personal history. Honolulu, Hawaii : The University Press of Hawaii.

PARKINSON, R. (1898-99) : "Zur ethnographie der nordwestlichen Salomo Inseln". Abhandlungen und Berichte des Königlichen Zoologischen und Anthropologisch-Ethnographischen Museums zu Dresden, 7, No.

PARKINSON, R. (1907). Dreissig Jahre in der Südsee : Land und Leute, Sitten und Gebräuche im Bismarckarchipel und auf den deutschen Salomoinseln. Stuttgart : Strecker & Schröder.

RIBBE, C. (1903) : ZweiJahreunterdenKannibalenderSalomo-Inseln : ReiseerlebnisseundSchilderungenvonLandundLeuten. Dresden-Blasewitz : Hermann Beyer. DOI : 10.2307/198478

RIESENFELD, A. (1950) : The Megalithic Culture of Melanesia. Leiden : E . J. Brill.


McALPINE, J.R., SAUNDERS, J.C. and SPEIGHT, J.G. (1967) : LandsofBougainvilleandBukaIslands, TerritoryofPapuaandNewGuinea. Land Research Series No. 20. Melbourne, Australia : Commonwealth Scientific and Industrial Research Organization.

SERVICE, E.R. (1962) : PrimitiveSocialOrganization : anevolutionaryperspective.

New York : Random House (Studies in Anthropology).

SERVICE, E.R. (1975) : OriginsoftheStateandCivilization : theprocessofculturalevolution. New York : W.W. Norton and Company.

SIMON, H.A. (1969) : The Sciences of the Artificial. Cambridge, Mass. : Massachusetts Institute of Technology Press.

SPECHT, J.R. (1969) : Prehistoric and Modern Pottery Industries of Buka Island, T.P.N.G. Unpublished Doctoral Thesis, Australian National University (2 volumes).

TERRELL, J.E. (1972a) : "Geographic systems and human diversity in the northern Solomons" . Reports of the Bougainville Archaeological Survey, 5. Chicago : Department of Anthropology, Field Museum of Natural History.

TERRELL, J.E. (1972b) : "A tale from the south Pacific". Field Museum of Natural History Bulletin, 43, 3 :12-18.

TERRELL, J.E. (1976) : PerspectivesonthePrehistoryofBougainvilleIsland, PapuaNewGuinea : astudyinthehumanbiogeographyofsouthwesternPacific. Unpublished Doctoral Thesis, Harvard University.

TERRELL, J .E. (n.d.) : "The birth of civilization and the state in the south Pacific : Melanesia's legacy to the world". (Paper in preparation for the Annual Meetings of the American Anthropological Association, Washington DC, 1976).

TERRELL, J.E. and IRWIN, G.J. (1972) : "History and tradition in the northern Solomons : an analytical study of the Torau migration to southern Bougainville in the 1860's". Journal of the Polynesian Society, 81 : 317-349.

THURNWALD, H. (1934) : "Woman's status in Buin society". Oceania, 5 : 142-170. DOI : 10.1002/j.1834-4461.1934.tb00138.x

THURNWALD, H. (1937) : MenschenderSüdsee : CharaktereundSchicksale, ErmitteltbeieinerForschungsreiseinBuinaufBougainville, Salomo-Archipel, Stuttgart : Ferdinand Enke.

THURNWALD, H. (1938) : "Ehe und Mutterschaft in Buin (Bougainville, Salomo-Archipel)". Archiv für Anthropologie und Völkerforschung, (n.s.) 24 : 214-246.

THURNWALD, R. (1909) : "Reisebericht aus Buin und Kieta". Zeitschrift für Ethnologie, 41 : 512-532.

THURNWALD, R. (1910) : "Im Bismarckarchipel und auf den Salomo-inseln 1906-1909". Zeitschrift für Ethnologie, 42 : 98-147.

THURNWALD, R . (1912) : ForschungenaufdenSalomo-InselnunddemBismarck-Archipel. Vol. 1 : "Lieder und Sagen aus Buin". Vol. 3 :"Volk, Staat und Wirtschaft". Berlin : Dietrich Reimer (Ernst Vohsen).

THURNWALD, R. (1934a) : "Pigs and currency in Buin : observations about primitive standards of value and economics". Oceania, 5 : 214-217.

THURNWALD, R. (1934b) : "Stone monuments in Buin (Bougainville, Solomon islands)". Oceania, 5 : 214-217.

THURNWALD, R. (1937) : "Ein vorkapitalistisches Wirtschaftssystem in Buin : Ein Beitrag zur Kenntnis primitiver Wirtschaft und von Frühgeld". Archiv für Rechtsund Sozialphilosophie, 31 : 1-37.

THURNWALD, R. (1951). "Historical sequences on Bougainville". American Anthropologist, 53 : 137-139.

WHEELER, G.C. (1914) : "An account of the death rites and eschatology of the people of the Bougainville Strait (western Solomon Islands)". Archiv für Religionwissenschaft,

17 : 64-112.

WOODFORD, C.M. (1890) : ANaturalistAmongtheHead-Hunters. Melbourne and Sydney : Petherick.


1 Thurnwald gives further details, also (1937 : 4-5). Frau Thurnwald has reconstructed the process of conquest in detail : Von der Küste aus erblickt man die Shortland-Inseln Alu und Fauro, und das weiter westlich gelegene Mono. Die Bewohner dieser Insein haben entscheidend Ln das Schicksal des Buin-Volkes eingegriffen. Es waren Banden der kriegerischen seefahrenden Mönner aus Mono und Alu, wahrschein-Lich auch noch von Roviana auf der Insel New-Georgia (zentrale Solomonen), die vor ungefähr zwei Jahrhunderten anfingen in Buin einzudringen. Sie holten aus der dort ansässigen Bevölkerung zunächst Frauen und junge Männer als Sklaven herüber nach Alu, kamen wieder und wieder und ein Teil von ihnen siedelte sich schliesslich in Buin an. Mitteis ihrer überlegenen kriegerischen und zivilisatorischen Ausrüstung gelang es ihnen in Buin ein Herrschaftssystem aufzurichten, das die ansässige Bevölkerung zur abhängigen Unterschicht, ja zu Hörigen mochte und ihre alte Lebensordnung weitge-hend abwandelte (Hilde Thurnwald 1938 : 214-215).

2 As this paper was being completed, and after Kothleen Fine had typed the final draft of all but the concluding section, I received a copy of Professor Jared Keil's dissertation, "Local Group Composition and Leadership in Buin" (Harvard, August 1975), kindly sent me by Dr. Keil after a delay created by a Canadian postal strike. His study is based on field work in Buin from November 1971 to September 1973. He indicates that Buin society is, indeed, stratified into two classes, as Thurnwald reported. He believes the invasion hypothesis, however, is unwarranted and unnecessary. If it is rejected, Keil feels that the evidence recorded by the Thurnwalds, in fact, illustrates many of the processes he examines in his own study. Keil's discussion is elegant and his conclusions dealing with so-called "big-man politics" are of porticular interest. He finds the conduct of political life in Buin between (rather than within) Local neighborhood groups to be, in many respects, quite similar to the situation detailed earlier by Oliver for the Siwai, a conclusion implicit in what has been said here as well. Even at the local Level where a difference between ascribed vs. achieved status Looms Large, he observes that the actual functioning of neighborhood groups in Buin much resembles the workings of Siwai men's societies.

3 Thurnwald gives further details, also (1937 : 4-5). Frau Thurnwald has reconstructed the process of conquest in detail : Von der Küste aus erblickt man die Shortland-Inseln Alu und Fauro, und das weiter westlich gelegene Mono. Die Bewohner dieser Insein haben entscheidend Ln das Schicksal des Buin-Volkes eingegriffen. Es waren Banden der kriegerischen seefahrenden Mönner aus Mono und Alu, wahrschein-Lich auch noch von Roviana auf der Insel New-Georgia (zentrale Solomonen), die vor ungefähr zwei Jahrhunderten anfingen in Buin einzudringen. Sie holten aus der dort ansässigen Bevölkerung zunächst Frauen und junge Männer als Sklaven herüber nach Alu, kamen wieder und wieder und ein Teil von ihnen siedelte sich schliesslich in Buin an. Mitteis ihrer überlegenen kriegerischen und zivilisatorischen Ausrüstung gelang es ihnen in Buin ein Herrschaftssystem aufzurichten, das die ansässige Bevölkerung zur abhängigen Unterschicht, ja zu Hörigen mochte und ihre alte Lebensordnung weitge-hend abwandelte (Hilde Thurnwald 1938 : 214-215).

4 As this paper was being completed, and after Kothleen Fine had typed the final draft of all but the concluding section, I received a copy of Professor Jared Keil's dissertation, "Local Group Composition and Leadership in Buin" (Harvard, August 1975), kindly sent me by Dr. Keil after a delay created by a Canadian postal strike. His study is based on field work in Buin from November 1971 to September 1973. He indicates that Buin society is, indeed, stratified into two classes, as Thurnwald reported. He believes the invasion hypothesis, however, is unwarranted and unnecessary. If it is rejected, Keil feels that the evidence recorded by the Thurnwalds, in fact, illustrates many of the processes he examines in his own study. Keil's discussion is elegant and his conclusions dealing with so-called "big-man politics" are of porticular interest. He finds the conduct of political life in Buin between (rather than within) Local neighborhood groups to be, in many respects, quite similar to the situation detailed earlier by Oliver for the Siwai, a conclusion implicit in what has been said here as well. Even at the local Level where a difference between ascribed vs. achieved status Looms Large, he observes that the actual functioning of neighborhood groups in Buin much resembles the workings of Siwai men's societies.

Leave a Reply

Your email address will not be published. Required fields are marked *