Tongan ha‘a are agnatically grouped clusters of titles that are ranked according to a rule of historical — and, in the case of the apical ancestor, mythical — emergence or precedence (cf. Fox 1994, 1995). Since elder brother outranks younger brother and father outranks son, “elder brother” or “father” titles outrank “younger brother” or “son” titles, the “branches” (va‘ava‘a) of the Tongan ramage or “origin structure” (Fox 1988) being graded according to relative proximity to an original “root” (tefito). Mythically, if not historically, this root is the Tu‘i Tonga or high chief of Tonga (Figure 1) (Sahlins 1958:ch.1). While these botanic idioms suggest a centred, self-totalizing, and self-generating “origin structure” (Fox 1995), they are as important for what they conceal as for what they reveal. Precedence may reflect affinity and the dynamics of the system rather than descent and the statics of the system (cf. ibid.:12-13).[1]
In any effort to explore the mobilities and instabilities of Polynesian systems, Irving Goldman’s Ancient Polynesian Society offers some crucial insights. There Goldman places “status rivalry” — a jockeying for position at the top motivated by a systemic “ambiguity of rank” (Goldman 1970:24) — at the heart of these societies. Ambiguity of rank is created by the complexity of the ranking system. “Hereditary rank … is no single factor, but a compound of multiple genealogical criteria …” (ibid.:5).
In Tonga (but also elsewhere in Polynesia), senior and junior brothers, senior and junior dynasties, typically contended for top honours. “It was … a frequent theme that younger brothers and junior lines of descent, although lower in rank, came to be greater in de facto political power and eventually ‘upstaged’ the senior line” (Bott 1982:62). The competition often (though not always) aligns matrilateral half-brothers, who do not compete for the same titles, against patrilateral half-brothers, who do (Bott 1981:17; Rogers 1977:171-173). The great secular trend of Tongan history is driven by this rivalry. According to oral tradition, when the 23rd Tu‘i Tonga was assassinated in the fifteenth century, his son retained the divine privileges of the Tu‘i Tonga but surrendered the effective governance of his kingdom to his younger brother, Takalaua. From that time the government was divided between a “working chief” (tu‘i or ‘ eiki ngāue) associated with the Kauhalalalo cluster of titles and a “most sacred chief” (tu‘i or ‘ eiki toputapu) associated with the Kauhala‘uta cluster of titles (Figure 1). The term kauhala‘uta means the “bush” side of the road and the term kauhalalalo means the “lower” side of the road, most scholars interpreting this as a symboliccum-geographical division between land and sea. While the “sacred chief” enjoyed a great many privileges, the “working chief” was the effective leader, a governor of the people and the land (Ve‘ehala, pers.comm.).
The initial “working chief” was the Tu‘i Ha‘atakalaua (Figure 1). However, by the seventeenth century, the Tu‘i Kanokupolu had displaced the Tu‘i Ha‘atakalaua as the “working chief”, the Tu‘i Tonga continuing as the “most sacred chief”. These diarchic arrangements (Biersack 1990a; Valeri 1989, 1990a) were superseded in the nineteenth century, when Tāufa‘āhau, a usurper,[2] unified the archipelago, took the name King George Tupou, and substituted for the diarchy of the past a constitutional monarchy. Since Tāufa‘āhau was heir to the Tu‘i Kanokupolu title, the Kauhalalalo had in effect displaced the Kauhala‘uta.
As I shall argue here, these hierarchical transformations were underwritten by the complexity of Tongan ranking. A person acquires political status through his or her father but personal status through his or her mother and is thus complexly and often contradictorily positioned within a social and political field. Over against a ha‘a system of titles, egocentric kindreds or kāinga (Kaeppler 1971) have their own social and political force. Considering such complexity, Goldman argues that though the “structure is stable” in the sense that its “two coordinates” endure over time (ibid.:304), the “points on the coordinates” — that is, the rank order of particular historical actors — “are unstable” (ibid.). “… persons who are ambitious cannot change the absolute laws of genealogical rank …. The structure does, however, allow for shifts in power relationships, offering several leverage points for such shifts” (ibid.:304-305).
Together with Bott’s Tonga at the Time of Captain Cook’s Visits, Goldman’s analysis offers a basis for rethinking Tongan society in historical terms, albeit without precluding structural analysis. Here I develop a post-Durkheimian sociology of Tonga based on a combination of structural analysis and historical investigation. The first section provides the structural analysis. The remainder of the paper exploits this analysis as an instrument of diachronic investigation. In the concluding section I draw the lessons that Tongan history has to teach about the relationship between structure and event, the synchronic and the diachronic. My main claim is that Tongan structure becomes operational only in and through Tongan élite marriage practices and that the Tongan ramage is a product of these practices and their dynamics. In the course of developing the argument, I take up the issue of women’s status and position in Tonga — a topic that is unavoidable given the array of synchronic and diachronic material this paper surveys and the importance assigned to marriage in it.[3]
![Figure 1. A simplified representation of the Tongan title system (adapted from Biersack 1982 [1974]:197 [Figure 1]).](../images/c12f001.jpg) |
Figure 1. A simplified representation of the Tongan title system (adapted from Biersack 1982 [1974]:197 [Figure 1]).
Traditionally and still today, Tonga sustains two ranking systems (Biersack 1982 [1974], 1990b, 1991a; Bott 1981, 1982; Goldman 1970:449-459; Kaeppler 1971; Rogers 1977), each operating in its own context and cultural domain. The kāinga (kinship, kindred) system rank orders persons as persons and as they participate in kindreds, while the ha‘a system rank orders titles and their holders according to an agnatic calculus: elder brother:younger brother: ‘ eiki [or chiefly rank]:tu‘a [or commoner rank]::high:low.[4] Within the title system, at least in theory, the consanguineal relationship among actual titleholders is immaterial. What matters is the genealogical proximity of eponymous ancestors to an apical ancestor: ‘Aho‘eitu, the father of the first Tu‘i Tonga (Sahlins 1958:ch.1). Also, whereas eligibility for office is unilineally conferred, personal rank is not only unique but is bilaterally rather than unilaterally and unilineally determined.
Within the context of the kāinga, sister and sister’s children outrank brother and brother’s children. Thus, in life crisis ceremonies women and their children are the ranking figures (Biersack 1982 [1974]; Bott 1972, 1981:18-20; Kaeppler 1971; Rogers 1977). The highest position in such ceremonies is the fahu position, which is usually occupied by a sister’s child or a father’s sister’s child. A position of honour, the fahu receives the best of the valuables distributed in such ceremonies.
In the past sororal hegemony had its most dramatic expression in the extraordinary status accorded the sister of the Tu‘i Tonga as the “female” Tu‘i Tonga or Tu‘i Tonga Fefine. The status of the Tu‘i Tonga Fefine and that of her eldest daughter, the Tamahā, were conferred within the kāinga of the Tu‘i Tonga rather than within the ha‘a system. Of the Tu‘i Tonga Fefine, John Thomas, the first continuing Methodist missionary in Tonga, wrote that the position gave “her no power or authority, but she is raised by it, and greatly honoured” (Thomas 1879:15). The Tu‘i Tonga Fefine was married to a chief of the Ha‘a Fale Fisi, the “ha‘a of the house of Fiji”, a line of foreign provenience (Figure 1). The foreign paternity of the Tamahā or eldest daughter of the Tu‘i Tonga Fefine meant that, although the Tu‘i Tonga Fefine’s children had the highest possible personal rank through their mother and outranked their mother’s brother, the Tu‘i Tonga, their rank in the ha‘a or title system was inferior to his. Chiefs of a foreign line, they constituted no serious political threat to the paramount (cf. Sahlins 1976:42).
In an idiom of “garland” (kakala) (or “name” [hingoa]) and “blood” (toto), Tongans themselves distinguish between the person and his or her rank vis-à-vis kin, on the one hand, the title and its rank, on the other. The title is metaphorically a “garland” the titleholder wears. In contrast, personal rank is “blood” rank, the rank of the “body”. A key aspect of the blood/garland distinction concerns the kind of power associated with each kind of rank. The sister and her children, vis-à-vis the brother and his children, exercised power over individual and biological life (ibid.), while the power of the titleholder was a public power exercised over macrosocial geopolitical units, the village and the estate (tofi‘a) on which a village or multiple villages formed. Sororal power was manifested most dramatically in a father’s sister’s cursing power, which placed organic life at risk, while the latter was manifested in the juridical authority of a chief.
The distinction between ha‘a and kāinga rank, “garland” and “blood”, is gendered. The rank acquired through women is personal while the rank acquired through men is the impersonal public rank of title and office (Biersack 1982 [1974], 1991a; Rogers 1977). Thus, while paternal rank is a factor in determining a child’s personal rank, the blood of the mother is more important than the blood of the father (Bott 1981:19; Gifford 1929:113; Marcus 1977:227, n.8, 1980:89; Spillius n.d., ch.5, BSP 4/3/31[5]). “If an aristocratic woman and an aristocratic man of roughly equivalent rank marry commoners of low rank, the children of the aristocratic woman will have much higher rank than the children of the aristocratic man” (Spillius, “Kinship and Ha‘a in Tonga”, BSP 5/6/5-6). It follows that the fount of aristocratic blood is feminine. “To be ‘aristocratic’ … means to be descended from the Tu‘i Tonga Fefine or the Tu‘i Tonga” (Bott 1982:61) or the Tamahā (Bott 1981:36; see Rogers 1977:170-173). Of a child whose high rank depends primarily on the mother it is said that the child is tama tu‘u he fa‘ē, “child standing on the mother” (BSP 17/6/n.p.) or “child standing in the mother” (Collocott and Havea 1922:118).
Whereas women and their children effectively dominate the contexts governed by personal rank (life crisis ceremonies), “their ascendancy was supposed to be confined to the realm of rank” (Bott 1982:76). Titleholders, however, assumed the command of particular villages and districts and all the people living therein; and it is men rather than women who are associated with the title system. Except in rare cases, titleholders were and are male.
Men have authority (pule) over their younger brothers and over their children — also over their wives. They control access to land and titles; they control and organise economic affairs. Reciprocally, men owe allegiance, subservience, respect, and tribute to their fathers and older brothers, real and classificatory. Sisters, however, have higher rank than brothers, even though they have no secular authority over them. They are more ‘ eiki than brothers, but have no pule over them (Bott 1982:58).
Similarly, Garth Rogers refers to “the ideology that the ‘father’s side’ of the family has the pule over the children and not kāinga ‘i fa‘ē ‘the mother and her immediate relatives’ … ‘father’ is normally considered the head of the household with pule over its members …” (1977:159). He goes on to observe — drawing on remarks made by Tupou Posesi Fanua, a well-known authority on Tongan culture — that the “rights of a fahu at a wedding, first-birthday, or funeral are symbolic and decorative but not political; in fact, Tupou Posesi said that “the fahu have no say in anything; they are not the pule” (ibid.:168).[6] Moreover, the idioms in which the relationship between and among titles is couched — as father and son or older and younger brother (Biersack 1982:196) — is agnatic. It is said, I am told (‘Okusitino Māhina, pers.comm.), that “from women blood flows” (“ko e fefine ko e fakahokohoko toto”) but “from men titles flow” (“ko e tangata ko e fakahokohoko hingoa”). This is not to deny that men no less than women had personal rank and that men no less than women functioned as fahu on particular occasions. Since sister’s children outrank brother’s children, men outrank their mother’s brother and his children and have the same kāinga or kinship superiority toward them that the sister enjoys vis-à-vis the brother or the father’s sister enjoys vis-à-vis the brother’s children.
Public authority is acquired through titleholding; and no matter how high the rank of a particular person, rank could not be translated into command without the bestowal of a title. Elaborating upon Bott’s (1981, 1982), Rogers’ (1977) and Kaeppler’s (1971) work, as well as my own early work (Biersack 1982 [1974]), I have argued that title and the authority that attaches to it are inherently impersonal — authority emanates from the office rather than the person and all titleholders are the same — while rank is inherently personal, an attribute of bodies (Biersack 1990b, 1991a). Thus, the Tu‘i Tonga Fefine, second in rank only to her own children, had no heir. Instead of being succeeded, the Tu‘i Tonga Fefine was transcended — by a daughter who outranked her: the Tamahā or “most sacred child”. A daughter of a chief of the Ha‘a Fale Fisi (Figure 1) and the Tu‘i Tonga Fefine, the Tamahā was fahu to the Tu‘i Tonga.
In distinguishing the two ranking systems, Bott dwells in particular on one Lātūnipulu, who as Tu‘i Lakepa of the Ha‘a Fale Fisi (Figure 1) had a relatively low status within the title system but who as sister’s son or fahu of the Tu‘i Tonga had exceptionally high personal status. In meeting Lātūnipulu in 1773, Captain Cook was immediately impressed with his apparent power and prestige. Cook “saw that he was a man of some consequence by the extraordinary respect paid him, some when they approached him fell on their faces and put their heads between his feet and what was still more no one durst pass him till he gave them leave” (from Cook’s journals; quoted in Bott 1982:16-17). Lātūnipulu’s prestige, a matter of blood and not title, stemmed from his mother, the Tu‘i Tonga Fefine;[7] and in this was feminized. It was (and still is) possible for a man to have high personal rank but no title; a man could be, in the local idiom, “chiefly in body” (sino ‘i ‘eiki [Bott 1981:10-11] or an aristocrat without being appointed to title (‘ eiki fakanofo, “chief who has been installed”) (ibid.).
There is a sense in which title rank is inferior to blood rank, the rank being an attribute of the title but not the person. A titleholder having no royal blood is said merely to wear a garland (“Discussions with Queen Sālote” 1:50; hereafter referred to as “Discussions”). “The title [or] ‘garland’ (kakala) can be taken away [but the] ‘blood’ (toto) is one’s own for ever” (Bott 1981:38; see also “Discussions” 1:50; Marcus 1980:18-19). “It was typical for titles to be disdained”, Marcus observes (ibid.:61). High ranking chiefs “frequently passed them on to younger siblings, adopted children, or other persons of lower kin rank” (ibid.); and James Spillius writes that “very high ranking [male] aristocrats considered it lowering to have a title, and were known by their personal names rather than the name of their titles. Some had no titles. Titles meant political obligations; high rank meant deference from everyone” (Spillius, “Kinship and Ha‘a in Tonga”, BSP 5/6/6). Even though Lātūnipulu was Tu‘i Lakepa, “he rarely used the title; … the very great ‘ eiki (like Lātūnipulu) were scornful of titles, which they considered to be appropriate to lesser … chiefs” (Bott 1982:60).
[1] This paper was initially drafted while I held a grant-in-aid from the American Philosophical Society and was a Visiting Fellow with the Comparative Austronesian Project, Research School of Pacific and Asian Studies, The Australian National University. It was revised for publication in light of further research while I held a NEH Travel to Collections grant. Dr James Fox was the convener of the project. It is based in part on research I completed under grants-in-aid from the Wenner-Gren Foundation for Anthropological Research and the American Council of Learned Societies and a summer research award from the University of Oregon, all in 1986. I thank also Dr Niel Gunson and the Department of Pacific and Southeast Asian History, Research School of Pacific Studies, for a visiting fellowship in that department in January 1987. Others have also been encouraging and helpful: Tupou Posesi Fanua, Futa Helu, Adrienne Kaeppler, Sione Lātūkefu, “Okusitino Māhina, George Marcus, the late Garth Rogers, Tavi, ‘Ofa Tulimaiau, Takapu, Pohiva Vaimo‘unga, Valerio Valeri, and the late honourable Ve‘ehala. I am indebted to Drs Fox and Sather and especially to Drs Ian Campbell and Niel Gunson for their thoughtful critiques of a prior version. I also wish to thank Richard Brown, Suzanna Layton and Clive Moore, all of The University of Queensland, for assistance in locating a key text in the eleventh hour of preparing this article for publication. Any errors of fact or interpretation remain my own.
[2] I use the word usurper in a rather loose sense. I do not mean to imply that Tāufa‘āhau appropriated a title that another person by rights should have succeeded to. Tāufa‘āhau rightly claimed the Tu‘i Kanokupolu title; and since the royal title, King, was of his invention, it is impossible to argue that he illegitimately held that title. However, the King title in effect displaced the Tu‘i Tonga title and it placed Tāufa‘āhau, as King George Tupou I, at the top of the title system. Since Tāufa‘āhau had no birth right to the “summit” (tumutumu) of Tongan society, in the broad sense of the word he was a usurper.
[3] The literature on ancient Hawai’i similarly emphasizes the political uses of marriage. Valeri’s “Le fonctionnement du systéme des rangs á Hawaii” was the first to discuss this matter. The topic has been pursued by Sahlins (1981 and 1985), Valeri (1985, 1991) and Linnekin (1990).
[4] I hope I do not misrepresent her argument in saying that Kerry James envisions the source of status — “rank”, in her terminology — as ultimately feminine (James 1992). While the Tongan polity (traditionally and yet today) is clearly religious, as James claims, to emphasize blood- and female-derived status over title, with its masculine associations, is to miss the crucial distinction between blood and garland that Tongans themselves seem to draw and which informs my argument. After all, the first Tu‘i Tonga was ‘Aho‘eitu, a male; he was killed by his heavenly older brothers; to reward him ‘Aho‘eitu’s father, ‘Eitumatupu‘a, sent him back home on earth and installed him as the first Tu‘i Tonga; and to punish the brothers, the same father sent them with ‘Aho‘eitu to serve as his assistants and subordinates (Biersack 1991a; Bott 1972; Māhina 1986; Valeri 1989). The charter of the political system as a religious system could be neither more patriarchal nor more focused on fraternal rivalry (see also the exchange between James (1991) and Valeri (1994)).
[5] BSP refers to the Bott Spillius Papers in the New Zealand and Pacific Collection of The Library, University of Auckland. The tripartite number scheme refers to box number, file number, and page number, respectively. These papers are a compilation of the work of three authors: primarily Elizabeth Bott (then Elizabeth Spillius), James Spillius and Queen Sālote Tupou III. During the period 1958-1960, when Bott was in the employ of the Tonga Traditions Committee, Bott interviewed Queen Sālote as their principal informant on Tongan history and culture; and Spillius did so as well, although to a lesser extent. In the two-volume compilation of interviews with Queen Sālote called “Discussions with Queen Sālote”, Bott records the answers Queen Sālote principally, but also occasionally Ve‘ehala, gave to her questions, as well as the questions themselves. In citing documents from this collection, where no author is designated, the author of the cited document is Bott.
[6] While Bott here associates pule with the juridical power of the chiefs, the power they exercised over land and people, pule or “authority” may refer to the authority of the father’s sister as well (Lātūkefu, pers.comm.). Similarly, whereas I have distinguished the ha‘a from the kāinga system, the idiom of the chief-villager relationship was an idiom of kinship (Lātūkefu, pers.comm.; 1975:9, 1980:65-66), an idiom that reflects the moral character of governance in Tonga (Biersack 1990b, 1991a; Valeri 1989, 1990a).
[7] Bott notes that Lātūnipulu is sometimes described as a male Tamahā, but dismisses the possibility. “What seems most probable is that the actual title of Tamahā was bestowed only on … three women …, but that there was also a less precise usage by which all the people who stood in the right relationship were called Tamahā even if the actual title had not been specifically granted to them. The evidence of Cook strongly suggests that this was so. In any case, it is clear from Cook’s account that Lātūnipulu and his sisters, as well as their mother, were of higher rank than the Tu‘i Tonga” (Bott 1982:35).