Two
Substantivization and Anthropological Discourse:
The Transformation of Practices into Institutions in Neotraditional Pacific Societies

Nicholas Thomas

Anthropological thought has dealt mostly with totalities, with “society” or “culture.”^[1] These have always been partitioned; there have always been bits of one sort or another to be contextualized. Anthropologists, naturally, have not called the components of the totalities “bits” but have drawn on a confusing array of shifting terms. At different moments we have spoken of customs, institutions, habits, rites, norms, structural principles, ethics, icons, images, statements, acts, and practices. These objects may often overlap or roughly correspond, but the inflections of many of the terms were diacritical: they expressed differentiation of approach, the privileging of features marginal to an earlier vision, or rigor as opposed to imagery and interpretation.

The concern of this chapter is with an issue that cuts across and is frequently elided by this array of words: this is the difference between practices and ideas which are simply done or thought, or simply take place, and those set up as definite entities to be reflected upon and manipulated by the people in the situation under consideration. The former—like any action or statement—always express, even if idiosyncratically, broader concerns, meanings, and social relations but are not themselves regarded as symbolic entities or socially fundamental acts. Practices that are formalized and regarded as substantial entities do not necessarily have greater social effect, but potentially have quite different uses and implications. A ceremony that is named, and thus can be conceived of as an entity separable from particular enactments, can become a vital element in the self-expression or regeneration of a group, and especially may state and mark its differentiation from other groups. Attitudes toward it, or competing constructions of it, may also register crucial lines of disagreement or conflict within a particular society. I take it to be virtually tautological to assert that elaborated and substantivized notions or practices have a political and cultural potential which implicit

ideas and unreflective action do not have: to be manipulated, things must be “there” in some sense. It is of great importance whether a “custom,” “code of behavior,” or whatever is reified as such by those who do it and think it, or only by the anthropological analyst.

The problem is not, of course, reification or substantivization as such: the whole point of research and writing is presumably to order things and argue about them in a different way to the one in which they present themselves. Analytical reification may thus be justified or justifiable by the theoretical objective of disclosing implicit meaning. The form of substantivization is, however, consequential for an understanding of cultural dynamics, and for the interpretation of particular beliefs and statements. A conventional, unhistorical anthropological analysis may take a specific custom simply as an element in a culture, to be “explained” on the basis of its functional or signifying coherence with a totality; a particular institution or belief would be anthropologically substantivized as an expression of a “web” of signification or kinship (Hooper 1985, 13; Fortes 1949). The argument here does not deal with anthropological reification and essentialism (but see James Carrier's Introduction), but is concerned instead with the fact that indigenous reifications of culture have been passed over. I suggest that there is always a general cultural potential for substantivization, and that substantivization can be a topic of broader theoretical importance for anthropology, but the process of naming and reifying customs and beliefs takes place in a particularly marked and conspicuous fashion in the course of colonial history. Abstracted notions about practices can be invented on the basis of daily life and colonial interaction or borrowed from the practices of other groups, and come to be represented—by indigenous people or anthropologists or both—as “customs” in the strong sense, without it being recognized that their existence as such derives from the oppositional dynamics of the colonial encounter. Cultural objectification in noncolonial contexts—between saltwater and bush people in Melanesia, or rivalrous religious and migrant communities in Europe—may emerge in essentially the same manner as the type of substantivization discussed here, and could be based on a range of oppositional kinds of social difference; in the Pacific, however, the history of colonialism has been crucial to the recognition of culture and the elaboration of difference.

Kerekere in Fiji

In what remains one of the most detailed ethnographies of neotraditional Fijian society, Sahlins discussed transactions, reciprocity, and redistribution in the economy of Moala (in the Lau group of eastern Fiji) at considerable length. The analysis included an extended discussion of what was described as “the famous custom of kerekere ” (Sahlins 1962, 145); this was “the prevailing form of economic transaction among kinsmen as individuals” (1962,

203). The verb from which the term derived meant “request,” or, in economic contexts, “solicit”: kerekere involved soliciting goods, resources, services, or use rights in goods or resources (1962, 203). Only houses and house platforms could not be subject to kerekere; while land was inalienable, use rights in land could be solicited, as could virtually any kind of product, food, or service.

Sahlins drew the attention of readers to the fact that kerekere had frequently been mistranslated as “begging,” which it was “most emphatically not” (1962, 203, and cf. 441 nn. 2, 4). This was misleading because it obscured “the essential kinship ethic [and] the implied reciprocity” (1962, 203). Kerekere stood as a kind of formal operator as opposed to certain informal transactions; it played a significant part as a “mechanism” in a system that tended to “produce material equality in a community of kinsmen”; which “obliterate [d] the material inequality among households and regions by transforming it into social inequality”—that is, into asymmetries based on the prestige and superiority of givers as opposed to receivers. The singularity of kerekere is thus attributed to its place within a socioeconomic system based on principles quite different from those of Western capitalist economies, and we could see it being drawn into the polarities so succinctly articulated by Gregory in Gifts and Commodities —between clan-based and class-based societies, between transactions based on qualitative rank, directed at relations of social value, and those turning on quantitative equivalence, directed at profits or use-values; I have argued elsewhere (Thomas 1991) that Gregory's analysis provides a stimulating point of departure, but stress here that this is because his concepts outrun the limitations of these dichotomous categories. The theoretical error arises from a failure to notice that the properties of the gift economy are simply conceptual and ideological inversions of those of the capitalist economy, rather than attributes derived from particular studies of gift economies that have not been caught up in colonial entanglements. I shall argue that the peculiar cultural and political circumstances of Fijian colonial history produced a neotraditional system that corresponded in some of its effects and elements to the stereotype of a clan economy structured by solidarity and reciprocity.

Fijians have differing views today about customs such as kerekere, which have been taken to be emblematic of the “communal” character of Fijian society, and these constructions and adjudications have also changed over time. My point here, however, is that the reification of kerekere as a “custom” and a possible reference point for such debate is something that occurred during the colonial period; there is no evidence that kerekere existed in this sense during the earlier phases of Fijian history. That is, kerekere apparently did not exist as a reified custom in indigenous Fijian society during precolonial phases of its history. In the numerous accounts of Fijian culture, belief, behavior, and politics, in general and particular, as these were

observed before the formation of neotraditional Fijian culture in the years of expanded settlement and extensive official codification of early colonial rule—that is, before the 1860s—there appear to be no references to kerekere as a recognized practice, as a custom.

This statement may appear categorical, and needs to be specified and qualified in three ways. First, I am not making a claim about what was absent or present in precontact Fijian society, which is unknowable, but am instead contrasting the period of trading relations and contacts (roughly from 1800 to the 1860s) with the subsequent period of more extensive settlement and (from 1874) formal British administration. Second, the unsatisfactory character of evidence for a negative proposition is in this instance ameliorated by the extent and nature of ethnohistoric documentation. Third, my claim is not about the types of transactions called “kerekere,” which no doubt occurred in precontact society as well as at other times. It is rather about the process of objectification.

Let us consider the problematic character of strictly negative evidence. It is true that most of the descriptions by missionaries and traders were oriented toward the spectacular, and more specifically, toward the features of indigenous life which provided ideologically useful images, such as warfare, cannibalism, and widow-strangling. This is the obvious and familiar line of criticism of preanthropological ethnographic writing, and it has some basis, especially in relation to the missionary publications that were worked up by metropolitan editors mindful of the appetites of mission supporters and the need to secure future funding; such works often seem especially laden with the burden of stereotyping and racial imagery. I have argued elsewhere, however, that the most significant distortions in missionary texts operate at a more subtle level, and that, in general, the representational value of these accounts may move beyond the apparent constraints of premises.^[2] This is so especially if unpublished discursive manuscripts, rather than popular publications, are drawn upon. In fact, much of the missionary ethnography of Fiji is of a very high and detailed standard, and concerned itself with relatively mundane and unspectacular aspects of Fijian life. There was a bias toward reportage of the activities of chiefs, but it did not overwhelm entire descriptions. The manuscripts and publications of writers such as Williams (e.g., 1843–1852, 1858) and Lyth are, moreover, not the only sources; I have found no discussion of kerekere in the texts of naval visitors or beachcombers, who were from very different social backgrounds, and who of course acted and reported on the basis of very different practical interest (e.g., Wilkes 1845; Erskine 1853; Diapea 1928). Nor could it be suggested that these preanthropological writers were simply unable to notice a custom when they saw one, as there is a great contrast between the invisibility of kerekere and the extensive comments upon the rights of nephews (vasu ), especially chiefly nephews, to appropriate anything from within the maternal uncle's domain (e.g., Wil-

liams 1843–1852, I, 61; 1858, 28; Wilkes 1845, 63; Diapea 1928 [late 1840s], 89, 110). Given that one can hardly expect to find positive evidence for the absence of a reification, the absence of reference to kerekere from extended discussions of a variety of customary usages in a substantial literature justifies my claim that kerekere was not talked about as a recognized practice in Fijian culture before the 1860s.

This is not, of course, to say that the sorts of transactions described by Sahlins in his early monograph did not take place. It is almost certain that they did. Nor do I suggest that the word “kerekere” did not exist. There is evidence from Hazelwood's dictionaries (1850, 61; 1872, 54) that it had roughly the same meanings as were noted later. The point is, rather, that the claim made in Capell's later dictionary, that kerekere was “a recognized system in Fijian society” (1957, 112) would have been unintelligible earlier. In earlier usage the word may not have had such specific associations with a particular kind of transaction; Hazelwood's reference to a “petition” implies that services or favors, rather than property, might often have been sought. Capell noted some associated usage such as i kere ni vanua, ground rent paid in kind” (1957, 112); the translation may carry misleading connotations but does suggest a more diffuse pattern of meanings.

Why, then, should this practice have acquired a distinct status and become, as Sahlins put it, a “famous custom”?

The answer is connected with the singular form of indirect rule developed by the British administration in Fiji in the years immediately after cession (which took place at the request of a group of leading chiefs in October 1874). The influence of the first governor, Sir Arthur Hamilton Gordon, was crucial. He had interests in anthropology, and saw parallels between native Fijian society and the clan system which had existed in his native Scotland a few centuries earlier. He also had a unique opportunity to shape the direction of British policy in Fiji. Experience in other colonies, and especially New Zealand, had convinced him that trampling upon native rights, especially wide-scale alienation of indigenous land, produced highly undesirable conflict and instability. He was determined to avoid such developments in Fiji and, contrary to the vigorously expressed wishes of planters and settlers, he set about generating policies and establishing institutions that perpetuated what he saw as the indigenous Fijian chiefly hierarchy; more importantly, in the hands of his successors, a fabricated and homogenized but “traditional” system of corporate ownership of land by Fijian descent groups was institutionalized.

As the historian France has stressed, Gordon's construction of the basic principles of the traditional Fijian system derived less from detailed observation and inquiry than from general expectations about the nature of a society at a certain level of development; “communal ownership and inalienability”

were the cornerstones of a codification which solved the political problem of the moment by precluding the free sale of land to settlers and the dispossession of the Fijians (France 1969, 102ff.). A few years later, the missionary-anthropologist Fison, in correspondence with Morgan, determined that Fijians had reached the “Middle Period of Barbarism.” His 1880 discussion of Fijian land tenure, which stresses the paramount importance of communal ownership, was twice reprinted by the Administration, which used it and apparently regarded it as authoritative until at least the 1950s (France 1969, 118–119). Gordon and his successors were thus engaged in the creation of a substantivized Fijian culture, toward which they were sympathetic—Gordon in particular seeing himself as a Fijian paramount chief—but which was there to be acted upon in whichever ways were deemed necessary (see Thomas 1989b ).

Fijian “society,” or more particularly the “Fijian communal system,” was thus not something that always existed, either in Fijians' own minds or in practice. This is not to say that there was no substantivizing of Fijian customs before colonialism: certain practices were indigenously identified as vakaviti, Fiji style, as opposed to the vaka Tonga; for instance, in Fiji women were tattooed, whereas in Tonga and Samoa this was done only to men. In a similar manner, particular polities (vanua ) within Fiji had their distinctive customs of respect and ways of preparing kava. In some cases, what was presumably important was the specificity of local identity, rather than any oppositional concept of a type or social totality. However, where there was a broader oppositional relationship, of the kind that clearly is often characteristic of relationships between foraging and agricultural neighbors, it is clear that colonial circumstances would be conducive to the substantivization of a totality manifested in particular features, which would have lacked such salience earlier. The general idea of a Fijian customary order—which always glossed over considerable internal diversity in kinship, ritual, social structures, and language—drew on a set of basic components, which were therefore meaningfully revalued upon being incorporated as key constituents of the system in general. Some of the customs and institutions seen by European administrators and, to some extent, by other Europeans as the fundamental elements of the system were lala, the system of chiefly requisitions; kerekere, “begging”; the land tenure system; and solevu, the practice of largescale feasting.

The question of whether chiefly requisitions for goods or services should be tolerated, regulated or proscribed by the administration was a contentious political question which fueled tempers and consumed a great deal of paper from the earliest days of the colonial administration. Lala in fact often over-shadowed kerekere in debates about the desirability of retaining or modifying elements of Fijian society. Here I do not attempt to identify the origins of

kerekere as a vital representation in European discourse but turn instead to a document of considerable importance in Fijian colonial history: the 1896 report on the Decrease of the Native Population.

Reflecting an ambitious, ideologically charged, pan-Pacific concern with indigenous depopulation, the inquiry attempted to map out in great detail the cause of the nineteenth-century decline in the Fijian population, as well as evaluating potential means of reducing or reversing the trend. An initial section of the report set out an array of prospective causes for consideration; the opinions of respondents—mostly Europeans in Fiji—to a circular questionnaire were summarized and evaluated, and the judgment of the commissioners was expressed. The “Communal system” was among the prospective causes specified under the general class of “predisposing” causes “tending to the Degeneracy of the People as a Race” (Fiji 1896, 6–7). More immediate causes affecting “Welfare and Stamina,” such as water supplies, modes of treating the sick, “abuse” of yaqona (kava), and so on, were also discussed extensively.

A prefatory observation in the section on the communal system acknowledged that this was a site of contest: “This system, which is to some extent synonymous with the native policy of the Government, has been frequently referred to in criticisms of the Colonial Authorities, and has been the principal source of contention between the Government and the European Colonists” (1896, 45). The centrality of lala and kerekere in the communal system was stressed: “After lala —or perhaps before it—kerekere (the mutual appropriation of property) is the principal feature of the communal system” (1896, 47).

The statements of evidence were generally vehemently opposed to both the “system” in general and to the specific practices:

The perpetuation of the communal system keeps the people in their primitive state…. A good deal is said [in the respondents' statements] about the bearing of the communal system upon native industry. If a man acquires anything, he cannot retain it. It is lavaka' d by his chief, or kerekere' d by any relative. The former he cannot deny for fear of punishment, and the latter he must appease for custom's sake. (1896, 47–48)

The last sentence shows the dependence of the specific comment upon a much larger conception of the “custom-bound” savage condition. This was expanded in the note of the commissioners' own views. It was noted that the Fijian system was not wholly communal, that private property existed in such things as pigs and canoes, and therefore that Fiji probably represented “the first stage of evolution from the state in which the proprietary unit was the tribe” (1896, 57).

It is difficult to imagine how primitive society could exist without some such custom as communal lala and kerekere within the limits of the clan. So long as

usage prescribed an universal standard of industry the system worked well enough; it is only with the decay of that stimulus of fear that it has become mutilated. (1896, 57)

At this point it needs to be mentioned that two of the three authors of this document took a great interest in anthropology and were subsequently associated with the Royal Anthropological Institute. Both Bolton Glanville Corney and Basil Thomson wrote on aspects of Fijian customs and history, but Thomson's contribution is especially significant, as he later elaborated the theme of the “decay” of the “custom-bound” society at considerable length (Thomson 1908). In this context, the suggestion that kerekere was formerly essential and unobjectionable because a “standard of industry” was customarily enforced, but had become pernicious as “usage” had lost its force, was a neat way of expressing criticism of the particular custom without calling into question the government's overall policy of maintaining the native system: the problem was that change had brought about the “mutilation” of this specific element.

The report's recommendation on kerekere was singularly uncompromising, given that a positive view was taken of lala and that generally the authors were for administrative policy and against the settler view: “We condemn kerekere ” (Fiji 1896, 209).

Kerekere has in fact always been alluded to, or more extensively discussed, specifically in relation to the question of the inhibition of enterprise and development (e.g., Brewster 1922, 92; Roth 1953, 36, 49; Belshaw 1964, 126; Sahlins 1970 [1960], 84). I do not suggest that this overall process of definition and inscription had particular or necessary social effects apart from upon the consciousness and behavior of the administrative actors themselves. That is, the ideological construct did not automatically lead to the refashioning of actual Fijian societies. Fijian social and cultural change was obviously a complicated process that affected different parts of Fiji unevenly (partly because administrative control was initially far more limited in some areas, such as upland Viti Levu, than elsewhere); the adoption or imposition of official models was clearly only one aspect of the process. However, the indirect rule system in general made chiefs especially, and to some extent all Fijians, complicit in the transformation of the administrative construction of Fijian society into a model of considerable practical salience. This is apparent especially in relation to land tenure, which has become firmly enshrined in Fijian custom and has developed a renewed and deeper relevance in the political conflict and rhetoric associated with the end of parliamentary democracy in 1987. The point is thus not that a certain invented or imposed tradition is “inauthentic,” because through its enactment what begins as a mystification can become a genuine element of the system, a notion to be resorted to, or a model to be adduced.

The vital point is that what become elements of the system do not all have the same kind of significance; they should not be seen simply as parts of an expressive totality or functional or dysfunctional system of social relations. In the case of kerekere, such a view is inadequate because it obscures the fact that kerekere was not an isolated term, but one which operated in relation, and specifically in opposition, to others. Numerous efforts were made from about the time of the report onward to proscribe kerekere, and its true meaning and positive and negative features were extensively debated by Fijian native councils at all levels, as well as in correspondence to the government-published Fijian language gazette, Na Mata (Macnaught 1982, 20–21). Some Fijians, especially a few who had begun cash-cropping, supported the administrative moves to abolish or restrict the procedure; others defended it on various grounds. In one case, a provincial council in the eastern interior of Viti Levu voted overwhelmingly at one meeting to abolish kerekere for a trial period, yet at a subsequent vote were almost evenly divided. They were also confused about the application of the proscription in a particular case, and sought clarification from the white Resident Commissioner: “the people want to know whether when people kill pigs and make feasts and then ask for property if that is kerekere. ”^[3]

Here, then, the Fijians needed to obtain information from Europeans about the nature of what was supposedly an important institution of their own. While the observations of many Europeans cannot be taken to reflect adequately developments in Fijian culture and consciousness, it is clear from this example that as they responded to and debated colonial regulations, Fijians had to deal with European codifications of customs such as kerekere. While indigenous objectification was not ideologically homogeneous, and did not directly mirror European invention, the disputes and debates did turn upon newly substantivized terms. Whether individuals felt that the system should be maintained or suppressed, they were talking about a recognized entity that had not previously been defined as a topic of discussion, an ideological token. The meaning of kerekere as a substantivized practice derived largely from the fact that it was the target of policies that sought to foster individualism and dismantle the communal social order. These challenges were very ineffectual, and laws against kerekere were mostly ignored (Macnaught 1982, 21). In a private comment upon one of the legislative attempts to ban kerekere, Hocart effectively sums up my argument:

I think the government is mad or more ignorant than it imagines. There is a new law forbidding “begging”; I suppose the last governor was told that it killed initiative, and he imagined that it was a well defined custom like tauvu, circumcision, etc. & the result is that according to the law a Fijian who begs yams of his neighbour, to plant, or sends for a fowl to feed his guest, or brings a whales tooth and mats and asks for taro in exchange, is liable to the same penalty as a thief. The best part of the law is that it cannot be enforced. (Hocart 1912, n.p., emphasis added)

Sahlins noted that “breach of [the law] is not only common, it is a way of life” (1962, 203). This way of life, though, was not simply a pattern of behavior or an array of “well defined customs” of the type mentioned by Hocart, but a neotraditional existence that was profoundly affected by the rigidifying efforts of a paternalistic administration, and, more particularly, partly structured in opposition to the incursions of planters and a cash economy. The communal forms were recognized and magnified through contrast to what they were not, to what occasionally threatened to encroach upon them.

My point here is thus not that anthropologists have fabricated an order that did not really exist, or that they have regarded recently introduced practices or notions as manifestations of timeless tradition. An interpretive or analytic endeavor can deal with a system, irrespective of the time depth of some or all of its elements. The point should not be their origins but coherence and positional value. This sort of critique was, of course, the basis for the rejection of “historical” perspectives such as diffusionism by Radcliffe-Brown and others, and led to debates about “the relevance of past events” (Lewis 1968, XV-XX ). These claims have been contested on a variety of grounds; the objection I develop here is specifically that neglect of the colonial and historical processes leads to a misconstruction of the system; it fails to recognize the factors that make some features of culture or social life more prominent or important than others. The notion that some particular theme, process, or metaphor may be seen as the big thing for a group of people is never elaborated in detail but looms large in the way anthropologists informally and implicitly try to come to grips with particular bodies of evidence, whether from their own fieldwork or that of others (see Damon 1980, 289). Certainly, if one is attempting to learn about a group of people it is presumably necessary to identify their preoccupations or central concerns. However, given that the most elementary aspect of anthropological method involves the contextualization of specific items in terms of a social or cultural totality, what is problematic is the contextualization of the “big thing” in systemic terms which fail to recognize the reactive interpenetration of local and colonial systems. This point will be developed in relation to two other cases, which will help me to draw together a set of concepts which might provide a better basis for the analysis of colonized societies.

Reciprocity and Wamira.

In discussions of economic processes in egalitarian societies, anthropologists frequently stress that food is constantly shared and redistributed beyond the boundaries of immediate domestic groups. The fact that this point is also often emphasized in television documentaries about “disappearing” peoples alerts us to the ideological import sustained by the observation: it is a way of saying that in this type of simple society we find a kind of sharing amongst

kin, or amongst members of a community, which is quite unlike the privatized, impersonal, and even unpleasant interaction that pervades extrafamilial interaction in modern societies. For us, the family is a haven in a heartless world; for them, we suppose, the family is coextensive with society: pervasive kinship engenders cosy solidarity and a fuller, unfractured, and ultimately more natural existence.

The unsubtle versions of this vision, while continuing to flourish in mass-circulation anthropology, are becoming rarer in professional discourse. Everyone is attuned to the dangers entailed in projecting Western categories, either through eternalizing our forms or by romanticizing what are misrepresented as their inversion. Or rather, there is a widespread understanding at a general level that these projections are unsatisfactory; that Western hermeneutics can often lead simply to another level of tautology (e.g., Strathern and McCormack 1980). However, scepticism of this kind has been selectively directed: while some objects of anthropological discussion such as ritual have long been assumed to be complex and refractory for a Western vision, others have been taken to be transparent and relatively unproblematic. Economic transactions of the barter type have, for example, been taken to belong to a general class of straightforward exchanges of use-values (e.g., Sahlins 1972) that was not seen to demand especially complex cultural interpretation, but recent explorations of the field have led to rather drastic rethinking (see Humphrey and Hugh-Jones 1992).

I will argue that some images—such as that of the witch doctor invoked in the rationality debate, and that of tribal sharing—cannot be effectively contextualized in a way that subverts their ideological inflection without grounding in colonial history. Kahn's discussion (1986) of the attitudes of the Wamiran people (of Milne Bay province, Papua New Guinea) toward food and reciprocity permits me to elaborate this point.^[4] I choose her study because it is a fine and subtle description of people's attitudes and behavior; it does not attempt to ignore or minimize the contemporary importance of Christianity, the involvement of men in wage work, or the significance of purchased commodities in the Wamiran economy. It is also a down-to-earth, particularistic work that conveys a lively sense of individuals and events, of the flow of village life. It is thus not a book that is preoccupied with symbolic abstractions, that one would obviously expect to reify and dehistoricize its cultural subject matter.

The monograph's main concern is with food symbolism. This is explored initially through elucidation of an initially paradoxical state of “famine” lacking any objective basis in actual food shortage. This is attributed to a deeper “obsession” with control. Desires for food and sex are regarded ambivalently and therefore repressed, controlled, and channeled outward in a variety of ways. The myth of Tamodukorokoro, for instance, is shown to compound the images of hunger and sexual desire: the external movements

entailed in marriage involve “getting meat,” which is unavailable at home; home is a place of famine and alliance is therefore “socially necessary” but difficult and stress-ridden. In another line of argument it is suggested that women's sexuality and reproductive capacities create a problem for men: they are dependent upon processes they cannot control or even (allegedly) “represent” (Kahn 1986, 74). Kahn suggests that men “solve” this problem at a metaphoric level by exchanging and symbolically manipulating pigs: in this domain of “female surrogates” men see themselves as controllers and “become independent of women” (1986, 75).

The overall analytic procedure of the text is thus to make connections between food and other aspects of social reality; once the field of such connections acquires the status of an explanatory or interpretively satisfying context, the domain of food can be seen to “express” features of kinship, affinal competition, sexual relations, and so on, which are problematic in other ways. One register of thought and practice thus resolves complications or contradictions that are inescapable within another.

We might wonder whether these symbolic transpositions actually provide solutions (a sort of functional closure) or whether, in displacing the issue to another domain, they simply express tensions or contradictions (which might be taken to meet a psychological requirement of the actors). My main concern here, however, is the way Kahn's contextualizing procedure resorts to systemic links and coherence.

It is clear from Kahn's description of Wamiran food-sharing and generosity that this is something ethically crucial: “Wamiran social etiquette centers on, and can be translated into, rules about sharing food. Generosity is accorded the highest social value…. The emphasis on sharing was brought to my attention within the first week of my arrival in the village” (1986, 41). Although it is noted that the moral injunctions are routinely violated in practice, and that in fact ways of hiding food, avoiding sharing, or demanding that others share, are highly elaborated, the ideals and rules are “mechanisms to control the fear of greed” and are pursued further through the deeper meanings ascribed to hunger and famine. The line of argument, however, neglects the context of many of the statements quoted: “We are not like white people, we share our food” is thus taken simply as an expression of the Wamiran view of the world, rather than as an indication of the opposition in which the statement is most salient. Kahn's account indicates that an insistence upon food transactions and sharing was constantly enunciated against an image of Western practice:

Wamirans constantly reminded me that they define their world, themselves, and their relationships in terms of food. To bring their beliefs into a framework that I could more easily grasp, they often explained, “We are taro people, but where you come from, people are money people.” Time and again I was told,

“In your place people are different. They work for money. That is their life. That is who they are.” (1986, 154)

The last sentence is especially revealing in that it implies that, just as in America, work for cash is identity, food and transacting food are identity in Wamira—which is clearly the ethnographer's point. I suggest that the repetition and stress upon reciprocity and sharing is contrastive in the first instance: this ideological salience can only arise in opposition to what are perceived as Western values; it is not a self-subsistent expression that could have existed in the same form prior to the recognition of the terms to which it is now opposed. In 1984, in the Marquesas Islands, I was frequently told that people there shared food, that if one could not grow or buy any, one would be looked after by others, but that they knew that where I came from, no money-no food-you die. The persistence of this type of statement across a variety of colonized situations suggests that it was more than the Wamirans' helpful effort to put their ideas “into a framework that [Kahn] could more easily grasp” that in fact motivated those resonant polarities.

Egalitarianism in Hawii

The suggestion is essentially that, in fundamental respects, modern Pacific cultures and practices are organized oppositionally. This condition can be further documented with material from contemporary traditionalist Hawaiian society, through Jocelyn Linnekin's (1985) study of a settlement on Maui. Keanae is “a taro place,” a marginal settlement of about fifty households, seen both by outsiders and residents as having a “real Hawaiian” feel; it was still, in the mid-'seventies, a “reservoir of tradition” (Linnekin 1985, 18–22). This traditionalism was of course starkly different from that of anti-Christian Melanesians such as the Malaitan Kwaio and Sa speakers of Pentecost (Keesing 1982; Jolly 1982), and even from that inscribed in Christian “kastom,” because contact history in Hawaii has greater temporal depth and was marked by some savage moments of dispossession or displacement (for an overview, see Kent 1983). Continuity with early religious beliefs might be expected to be virtually nonexistent, given the celebrated chiefly abrogation of the kapu system in 1819 and the extent of subsequent development of various missions and indigenous churches.

Linnekin effectively demonstrates that what is now taken to be authentically Hawaiian as opposed to what is haole (foreign or white) is largely invented tradition:

A key concept in the organization of the cultural revival is that of 'ohana, a term for extended family … [which] appears rarely in texts and archival materials from the nineteenth century, but today it refers to an idealized version of the Hawaiian family unit, characterized by cooperation (kokua ), internal harmony,

and aloha (love, affection). The word evokes Hawaiian kinship and solidarity…. 'Ohana also describes an ethic of egalitarianism and thus represents the rejection of certain historical aspects of Hawaiian society [i.e., stratification and chieftainship]. (1985, 12)

This point is made in relation to contemporary and partly urban Hawaiian nationalism, but the overall pattern applies in such places as Keanae: “the egalitarian ethic does have a correlate in rural society, although aloha and kokua represent the ideals rather than the reality of village social relations” (1985, 12). The assertion of rural Hawaiian identity was and presumably still is thus based more on generalized notions about the importance of wide bilateral networks of relatives (in which one has aunties and uncles) and the value of gift-giving as opposed to purchase and sale, rather than upon a Melanesian-style retention of specific practices such as revalorized gender taboos and menstrual segregation.

Exchange is one of the central concerns of Linnekin's study. She notes that while Keanae Hawaiians are deeply and indisputably embedded in various capitalist relations, there is a strong ethic that within the village things, and especially food, should be given rather than sold. In practice food is offered constantly, both in the contexts of casual visiting, and in luau —feasts celebrating marriages, anniversaries, and the like. “The luau celebrates relatedness and the ideal of aloha” (1985, 114).

The larger interpretation of local gift exchange foregrounded in Linnekin's book is entangled with a knotty set of issues in the history of anthropological ideas: the Melanesia-Polynesia division, and the conflation of that pair of ethnological terms with the social conditions of “equality” and “inequality.” It might seem that Linnekin's work makes a further healthy step toward undermining these tired and very misleading labels, since her suggestion is essentially that the “important people” of Keanae, those who mobilize labor and offer major luau—are like Melanesian big men (1985, chap. 8) (although certain points of contrast, and especially the fact that the Hawaiians understate prestige, are consistently noted). Yet whereas the concept of Polynesian ascribed rank has retained currency (despite its limitations), the stereotype of the egalitarian, competitive Melanesian system is now discredited (see Douglas [1979], Jolly [1987a ] for extended critique). Linnekin's argument introduces the most dubious and problematic aspects of the Melanesian construct into the contemporary Polynesian situation.

As was the case with Kahn's book, it could not be complained that Children of the Land is an ahistorical work that suppresses the extent to which the people studied have been affected and indeed absolutely transformed by historical changes. It is not only an unlikely work to be flawed by colonial history; taken together with some of Linnekin's articles (e.g., 1983), it directly addresses the mutability and social basis of notions of tradition in Hawaii.

My interest here, however, is not in criticizing Linnekin's work, which more recently (1990) has included a valuable synthesis and development of the debate about tradition and identity in the Pacific; I am concerned rather with the way in which anthropologists' analytical frameworks can often obscure the oppositional character of substantivized traditions.^[5] Hence the elucidation of the status of “important people” through reference to Melanesian big men, which is central to Linnekin's book, seems to deflect a more effective contextualization of Hawaiian ideology in history.

Although it is true that Melanesian big men, like chiefs (among other Pacific political figures), might be restrained or cut down to size if perceived to be despotic or overly domineering, the antihierarchical nature of Keanae egalitarianism seems to have a rather more specific character. Linnekin notes at several points that the ideology is at variance with actual inequalities of property, but that much is made in practice of diminishing these differences: “the Hawaiian social ethic demands that individuals deny their achievements, repudiate special status, and avoid differences in social position” (Linnekin 1985, 135, 142–145, 209–210, etc.). This case thus almost amounts to an inversion of what Sahlins reported from Moala, where mechanisms such as kerekere removed material inequalities by transforming them into “social” asymmetries, that is, into prestige relations associated with giving and receiving (1962, 146). In Hawaii, actual differences are sustained while equality in interaction and behavior is upheld: “the Hawaiian big-man's role is notable in its understatement and in the public denial of any special status” (1985, 212). While Linnekin discusses the contrast with Melanesian systems in this regard, it is not explained beyond further reference to “the Hawaiian ethic of egalitarianism” (1985, 235–247).

This “ethic” need not, however, be the end point of analysis; it has some singular features that are consistent with a comparative model of egalitarianism in poor and marginal societies. I refer here to Jayawardena's argument about the interrelationship between “the existence in a complex society of a local community whose members occupy a uniformly low status in the wider system” and “a dominant egalitarian ideology that provides the norms [and] the basis of social solidarity” (1968, 425, 426). The people

are, or believe themselves to be, economically and politically under-privileged. They are not reconciled to their position and therefore are antagonistic to the social order and the upper class that represents it, whom they think are responsible for their deprivation and subjection…. The norms derived from the [egalitarian] ideology conflict with a degree of differentiation that actually exists [within the marginal community]. (1968, 425–426)

This argument is based primarily on Guyanese plantation workers of Indian origin who were obviously extracted from their own societies and

severely oppressed in a way that was rare in the colonial Pacific (although the conditions of Chinese laborers on Hawaiian plantations were probably as severe [Kent 1983, 41]). There are few parallels to the “violent collective protests” reported by Jayawardena (1968, 418–423; see also Jayawardena 1963).

However, Jayawardena's argument becomes more appropriate to the Pacific case if the significance of egalitarian ideology is seen less in terms of its function for a lower-class community, than in terms of an oppositional cultural process whereby the ideals of a particular group are produced through an inversion of the values of the dominating external world. The members of the marginal community invert the values they perceive as the ones that energize a larger set of discriminating relations. Hence formulations such as “you are money people, we are taro people.” This is part of a reactive process of positive collective self-identification within which a local group distinguishes itself, and expresses its own worth, by articulating and elaborating features of local practice which contrast with attributes of wider social relations. This process frequently operates in conjunction with a sense that the Europeans who perpetrate discrimination in workplaces or elsewhere, are also (generically) the producers of valuables, the possessors of wealth, and the means of obtaining and appropriating such valuables. The foreigners must often therefore be dealt with, perhaps through indentured labor, and are therefore represented in specific, ambivalent, and emotionally salient ways.

Partly through the influence of the “invention of tradition” paradigm, Pacific anthropologists have become aware of the extent to which traditional culture now operates as a “political symbol” (Keesing 1989), but it is important to examine the sense in which foreigners and colonizers were not a self-evident intrusion and presence but were also worked into an image by Melanesians and Polynesians, which then provided a foil for their own reactive self-identifications. The strength of fraternal (male) solidarity and egalitarianism among those who engage or have engaged in migrant labor must clearly be seen in this context (see Jolly 1987b ).

Hawaiian egalitarianism has a longer history in a whole sequence of displacements and cannot be seen as an immediate reaction to specific work conditions; it is rather a generalized reversal of the principles that order the external world. Gift-giving, as one of the main elements of the asserted local culture, is not a self-subsistent category but something that is defined in opposition to the sorts of transactions conducted by outsiders:

Hawaiians say that monetary transactions have no place within the village. As one informant explained, “Keanae is a small place. The minute you sell, you going to get in trouble. You give, don't sell. When you give, something tastes good; but when you sell, it not going to taste so good.” (Linnekin 1985, 137)

The “ethic” of egalitarianism of course upholds precisely what is absent, both in the general relationships that constitute the minority or oppressed group as such and in the differentiation within that group. Jayawardena argued that the combination of an egalitarian ideology with actual discrepancies of wealth led to “frequent disputes between individuals over real or alleged breaches of egalitarian norms” (1968, 426). It is notable that there are direct parallels between the “eyepass” disputes he reports from Guyana and a process noted by Linnekin of “talk stink”: critical gossip about objectionable behavior, in reaction to unseemly status seeking, reflecting the acute disapproval of those who “act high,” leading in some instances to the almost complete ostracism of offenders (1985, 146). As a Portuguese garage foreman told Linnekin, “[T]hey cannot stand the minute you little bit higher than them” (1985, 145).

Since writing the body of this paper I have spent some time in the western interior of Viti Levu, the largest island of the Fiji group. It would be inappropriate to extend the arguments of this essay with substantial ethnographic detail, but it may briefly be said that contemporary rural Fijian culture has a strong “oppositional” character of the kind postulated here. The constitution of “the Fijian way” (which has been discussed by Toren [1984] among other recent writers) places great emphasis on sharing, kinship, reciprocity, and respect. Fijian customs (itovo ) or manners (varau ) also entail hospitality and the welcoming of others. Here, the cruder and more rhetorically positive constructions seem to have been influenced by, or at least share imagery with, tourist constructs of “Fiji—the way the world should be.” However, from the rural perspective the emergence of these values from a distinctive Fijian-Christian social order constituted both hierarchically (through chieftainship) and on the basis of loloma (Christian affection and kin solidarity) is more significant. The polarities are between the way of the land (na itovo vakavanua ) and the path of money (na calevu ni lavo ); Indians and foreigners are dedicated to the latter, and are rhetorically said to be inhospitable, indifferent to kinship, and to be concerned much more with work and the money it generates than with children and wider family ties.^[6]

It should perhaps be made clear that the argument about the oppositional character of these Fijian cultures—which I think is true with permutations and to a greater or less degree of all contemporary Pacific societies—does not in some sense apply to the “whole” of Fijian culture. The reactive selfconstruction does influence practice in many spheres of daily life, but is particularly salient in rhetorical discussions that are specifically concerned with collective self-presentation. It follows from this that visiting foreigners, and particularly ethnographers, are extensively exposed to this form of ideological difference, but it would be wrong to suppose that these statements are made only to foreigners. Their internal relevance has, of course, been

magnified as the specter of interracial political conflict has been raised over recent years—small wonder that a Fijian anthropologist should bring out a work entitled The Fijian Ethos in the year of the coup (Ravuvu 1987). But the contrast between the way of the land and the path of money is perhaps even more significant as a way of encoding social difference within the Fijian community. People of the interior and others in villages and outlying islands make much of their commitment to the way of custom, setting themselves up as the “real” Fijians in opposition to those in town, who may have done well in money terms but have abandoned their Fijian-ness. Speaking of this, a man who had fought the Japanese most of the way through the Solomons before settling for the rest of his life in a tranquil interior village, said of Fijians in town, Eri sa qulu na itovo ni vavalagi —they have picked up the customs of foreigners—because, he said, they watch videos. This is, of course, a stereotype: urban Fijians are no more Westernized than the lives of rural Fijians are really mirrored in the positive but highly selective constitution of “the way of the land.”

A Wider View

Neotraditional customary sharing and egalitarianism is thus quite different from the condition of equality and the practice of reciprocity which might be thought to exist in “simple” societies in general. That state tends to be seen as the absence of varieties of social differentiation and inequality which are associated with more complex stages in social evolution. Although the evolutionary rhetoric is now overtly discredited, much of the intellectual baggage persists in implicit associations. Both Kahn's and Linnekin's studies are open to history, and disclose the impact of historical processes, but set up sharing and egalitarianism as elements consistent with cultural form rather than as effects of historical and cultural dynamics. Kahn neglects the extent to which the ideal of sharing is substantivized as a marker of the singularity of the Wamiran people and their difference from whites. Linnekin's study appears more conscious of the reified and abstracted character of the “ethic of egalitarianism,” but takes this simply as a central theme or—although she does not use the term—“key symbol” which provides a cohesive point of reference for analysis. Practices are not quite transformed into institutions, as kerekere in Fiji was transformed into a “custom,” but are substantivized as “ethics” or cultural propositions in an equivalent way. The open attitude of both texts toward history thus masks the fact that the core of the analysis depends on systematization in synchronic registers of meaning. The alternative perspective I have advocated here is not a reactive historical particularism that denies the systemic form of social and cultural phenomena, but rather a different systematization that takes process rather than coherence as its primary trope.

From this perspective certain cultural terms—such as kerekere and the practice of insisting upon sharing—are to be read at least partly as meanings predicated upon their opposites: they thus derive directly from indigenous confrontations with, and representations of, the other (i.e., intrusive foreigners). A system of positive and negative ideals may consist of items +A, -B, +C, -D, and so on, which might stand for “you must share food,” “you must not ‘act high’,” and so on. The argument of this paper began with a query as to why some ideas should be reified and enunciated anyway, and partly answers this by stressing the dependence of the terms mentioned upon another set, -A, +B, -C, +D, and so on, which are represented by the indigenous people as the values of the system they are—to a highly variable degree—oppressed by or confronted with. This is not peculiar to colonial contexts, and might fruitfully be analyzed elsewhere, but culture contact of all kinds does produce, in a particularly powerful manner, essentialized constructs of selves and others within which particular customs and practices are emblematic.

I do not wish to abstract from Jayawardena's argument any quasipositivistic hypothesis or an overall model for colonized Pacific societies. Substantivization is a general process, but there is, of course, a great deal of local specificity: the way of kastom for non-Christian traditionalists in Vanuatu and the Solomons is very different to the elite chiefs' version of the “way of the land” in Fiji or Tonga, and these differ again from reconstructions of Maori and Hawaiian culture within white-settler states. The general intellectual configuration at present is such that those who attempt to think about modern Pacific societies—and indeed tribal or marginalized societies in many parts of the world—must choose talk about power, work, exploitation, colonialism, development, and history, or opt for apparently more subtle and expressive words which bear upon culture, discourse, ethos, and metaphor. From a variety of directions, steps have been taken which enable us to break down this division, a theoretical task that seems urgent in the south Pacific, since recent upheavals in Fiji, Papua New Guinea, and elsewhere underline the political resonance of “custom” and tradition in both local “cultures” and nationalist rhetoric. The elaboration of Jayawardena's argument helps us to think differently and fashion an anthropology for these circumstances.

References

Belshaw, Cyril S. 1964. Under the Ivi Tree. London: Routledge.

Brewster, A. B. 1922. The Hill Tribes of Fiji. London: Seeley, Service and Co.

Capell, Arthur. 1957. A New Fijian Dictionary. Sydney: Australian Medical Publishing Company.

Damon, Frederick. 1980. “The Kula and Generalized Exchange: Considering some Unconsidered Aspects of The Elementary Structures of Kinship.” Man 15:259–292.

Diapea, W. [a.k.a. W. Diaper, John Jackson]. 1928. Cannibal Jack: The True Autobiography of a White Man in the South Seas. London: Faber & Gwyer.

Douglas, Bronwen. 1979. “Rank, Power, Authority: A Reassessment of Traditional Leadership in South Pacific Societies.” Journal of Pacific History 14:2–27.

Erskine, J. E. 1853. Journal of a Cruise among the Islands of the Western Pacific. London: Murray.

Fiji. 1896. Report of the Commission Appointed to Inquire into the Decrease of the Native Population. Suva: Government Printer.

Fortes, Meyer. 1949. The Web of Kinship among the Tallensi. Oxford: Clarendon Press.

France, Peter. 1969. The Charter of the Land: Custom and Colonization in Fiji. Melbourne: Oxford University Press.

Gregory, C. A. 1982. Gifts and Commodities. London: Academic Press.

Handler, Richard. 1984. “On Sociocultural Discontinuity: Nationalism and Cultural Objectification in Quebec.” Current Anthropology 25:55–71.

Hazelwood, David. 1850. A Feejeean and an English and an English and Feejeean Dictionary. Vewa, Feejee: Wesleyan Mission Press.

——. 1872. A Fijian and English and an English and Fijian Dictionary. London: A. Low, Marston & Co.

Hocart, A. M. 1912. Letter to W. H. R. Rivers. MS in Rivers Papers, Haddon Collection, Cambridge University Library.

Hooper, Antony. 1985. “Introduction.” Pp. 1–16 in Antony Hooper and Judith Huntsman (eds.), Transformations of Polynesian Culture. Auckland: The Polynesian Society.

Humphrey, Caroline, and Stephen Hugh-Jones, eds. 1992. Barter, Exchange and Value. Cambridge: Cambridge University Press.

Jayawardena, Chandra. 1963. Conflict and Solidarity on a Guianese Plantation. London: Athlone Press.

——. 1968. “Ideology and Conflict in Lower Class Communities.” Comparative Studies in Society and History 10:413–446.

Jolly, Margaret. 1982. “Birds and Banyons of South Pentecost: Kastom in AntiColonial Struggle.” Mankind 13 (special issue) :357–373.

——. 1987a. “The Chimera of Equality in Melanesia.” Mankind 17:168–183.

——. 1987b. “The Forgotten Women: A History of Migrant Labour and Gender Relations in Vanuatu.” Oceania 58:119–139.

Kahn, Miriam. 1986. Always Hungry, Never Greedy: Food and the Expression of Gender in a Melanesian Society. Cambridge: Cambridge University Press.

Kaplan, Martha. 1988. “The Coups in Fiji: Colonial Contradictions and the PostColonial Crisis.” Critique of Anthropology 8(3):93–116.

Keesing, Roger. 1982. “Kastom and Anticolonialism on Malaita: ‘Culture’ as Political Symbol.” Mankind 13 (special issue) :357–373.

——. 1989. “Creating the Past: Custom and Identity in the Contemporary Pacific.” The Contemporary Pacific 1:19–42.

Kent, Noel J. 1983. Islands under the Influence. New York: Monthly Review Press.

Lewis, Ioan M. 1968. “Introduction.” Pp. ix–xxvii in I. M. Lewis (ed.), History and Social Anthropology. London: Tavistock.

Linnekin, Jocelyn. 1983. “Defining Tradition: Variations on the Hawaiian Identity.” American Ethnologist 10:241–252.

——. 1985. Children of the Land: Exchange and Status in a Hawaiian Community. New Brunswick: Rutgers University Press.

——. 1990. “The Politics of Culture in the Pacific.” Pp. 149–173 in Jocelyn Linnekin and Lin Poyer (eds.), Cultural Identity and Ethnicity in the Pacific. Honolulu: University of Hawaii Press.

Macnaught, Timothy J. 1982. The Fijian Colonial Experience: A Study of the Neotraditional Order under British Colonial Rule prior to World War II. (Pacific Research Monograph 7.) Canberra: Australian National University.

Ravuvu, Asesela. 1987. The Fijian Ethos. Suva: Institute of Pacific Studies.

Roth, G. K. 1953. Fijian Way of Life. Melbourne: Oxford University Press.

Sahlins, Marshall. 1962. Moala: Culture and Nature on a Fijian Island. Ann Arbor: University of Michigan Press.

——. 1970 (1960). “Production, Distribution and Power in a Primitive Society.”

Pp. 78–84 in T. G. Harding and B. J. Wallace (eds.), Cultures of the Pacific. New York: Free Press.

——. 1972. Stone Age Economics. London: Tavistock.

Strathern, Marilyn, and Carol P. MacCormack, eds. 1980. Nature, Culture and Gender. Cambridge: Cambridge University Press.

Thomas, Nicholas. 1986. Planets Around the Sun: Contradictions and Dynamics of the Fijian Matanitu. (Oceania Monograph 31.) Sydney: Oceania Monographs.

——. 1989a. Out of Time: History and Evolution in Anthropological Discourse. Cambridge: Cambridge University Press.

——. 1989b. “Material Culture and Colonial Power: Ethnological Collecting and the Establishment of Colonial Rule in Fiji.” Man 24:41–56.

——. 1989c. “Tin and Thatch: Identity and Tradition in Rural Fiji.” Age Monthly Review (Melbourne) 8(11):15–18.

——. 1990. “Regional Politics, Custom and Ethnicity in Fiji.” The Contemporary Pacific 2:131–146.

——. 1991. Entangled Objects: Exchange, Material Culture and Colonialism in the Pacific. Cambridge, Mass: Harvard University Press.

Thomson, Basil. 1908. The Fijians: A Study of the Decay of Custom. London: Macmillan.

Toren, Christina. 1984. “Implications of the Concept of Development for the Symbolic Construction of ‘the Fijian way.’” Pp. 39–52 in Christian Clerk (ed.), The Effects of Development on Traditional Pacific Islands Cultures. London: Institute for Commonwealth Studies.

Wilkes, Charles. 1845. Narrative of the United States Exploring Expedition. (vol. III). Philadelphia: Lea and Blanchard.

Williams, Thomas. 1843–1853. “Miscellaneous Notes Chiefly Concerning Fiji and the Fijians.” MSS, Mitchell Library, Sydney.

——. 1858. Fiji and the Fijians. London: Heylin.