While scholars such as Bott and Marcus have insisted upon the importance of conceptually distinguishing power from authority (Bott 1981, 1982; Marcus 1980), this distinction is effectively blurred in Phyllis Herda’s “Gender, rank and power in Tonga”.

Much has been made in the literature of the lack of pule (authority) of the sister and the father’s sister, it being claimed that, at most, she influenced her brothers and their children in their rule, but did not effectively control or influence affairs [here she cites Kaeppler 1971, Rogers 1977, and Bott 1981]. This appears, in many respects, to be an anachronistic understanding of the extent of ‘influence’ in traditional Tonga. Before the advent of fixed hereditary title succession which was initiated by the Constitution of 1875, title succession was fiercely competitive; in such a situation, the power to influence the choice of successor would be great (Herda 1987:197).

She goes on to elide the distinction between specifically feminine powers, the powers a woman exercises over her brother and his children, on the one hand, and the jurisdiction a chiefly titleholder exercises over people and land, arguing that “some Tongan women held politically active titles distinct from those which emphasized their sacred sister role” (ibid.:198-199) and concluding therefore that “chiefly women were not, categorically, denied titles or legitimate access to political power” (ibid.:198). Herda’s test case is Tupoumoheofo.

Tupoumoheofo’s actions have been explained away by foreign observers as ‘subversive’, ‘tyrannical’ and ‘odious in the extreme’ and Tupoumoheofo herself described as an ‘extremely ambitious, scheming woman’, a ‘meddling wife’ and ‘nothing at all to the credit of her sex’. However, it will be argued that Tupoumoheofo’s actions were in accordance with her position and, more importantly, her rank in Tongan society and that her negative portrayal represents the European misunderstanding of female roles in pre-Christian Tonga, particularly as they related to the political sphere, rather than denoting unfeminine Tongan behaviour (ibid.:195; emphasis added).

I would agree with Herda that the father’s sister’s power to curse and a woman’s ability to choose her brother’s children’s marriage partners are substantial powers (ibid.:197), and I would also concur that some women did indeed aggrandize significant power in the political arena itself. But I would have to insist upon the transgressive nature of Tupoumoheofo’s deeds. According to the model if not the reality, in the pre-Constitutional era titleholders were not women. (This does not detract, of course, from Herda’s point that Western observers judged events ethnocentrically.) Tupoumoheofo’s actions were entirely transgressive — not only in terms of what she did but in terms of how she did it. Installation to the Tu‘i Kanokupolu title occurs within the context of a kava ceremony attended by subordinate chiefs, who thereby display their fealty and essentially elect the titleholder on the basis of attributes such as seniority, blood rank, suitability (Biersack 1991a; Bott 1972; Valeri 1989). Though Tupoumoheofo placed her back against the koka tree that all Tu‘i Kanokupolus placed their back against when they were installed, she could not thereby appoint herself. Herda knows this, for she writes: “No individual could simply appropriate the title, he had to be called to it” (Herda 1987:100). If indeed Tupoumoheofo did install her son as Tu‘i Tonga while Tu‘i Tonga Paulaho was still alive (see n.17), this act also was transgressive (“Discussions” 1:44; Bott 1982:41-42). Similarly, if her intention was to reunify the Kauhala‘uta and the Kauhalalalo, the “sacred” and “working” titles, this too constituted a major violation (see Ellem 1981:61; see n.18). In fact, it could be argued that the transgressions as such were appropriate to Tupoumoheofo’s position, for as moheofo she could operate across and not just within categorical boundaries (Biersack 1982 [1974]).

When King George Tāufa‘āhau Tupou IV, the present king (Figure 1), ascended the throne, he was hailed as the long-awaited embodiment of the blood of all three ancient lines. Through his mother, daughter of Lavinia Veiongo, he traced descent from Laufilitonga, the last Tu‘i Tonga, the Tamahā Lātūfuipeka, and the half-sister of Tāufa‘āhau; through his father he traced descent from the last Tu‘i Ha‘atakalaua and the Tu‘i Pelehake, in whom the Tu‘i Tonga’s kava privileges were vested and who symbolized the Tu‘i Tonga and the Kauhala‘uta side (Ko e Kalonika Tonga, June 30, 1967; see Figure 5). The title he held prior to his accession and still holds is his father’s title, Tungī. His younger brother was made the Tu‘i Pelehake. Now at the “summit” (tumutumu) of a transformed Tongan society, the present King, descendant of a nineteenth-century usurper, is the living emblem of a unity that was historically achieved through a weaving together of bloodlines.

The King’s body is emblematic of the fact that in Tonga centralization is achieved historically and horizontally, through marriage, rather than structurally and vertically, through descent. This contravenes the divisions rehearsed herein: blood and garland, marriage and descent, a domain of feminine empowerment and a domain of masculine empowerment. If the present King governs as the embodiment of the blood of the three branches (Figure 1), then the distinction between these domains is effectively collapsed, and transgression assumes the status of an “organizing principle”. My account is far more replete with examples of male transgressions than it is with examples of female transgressions. For example, whereas the father’s sister theoretically chooses her brother’s children’s spouses, it is clear enough from this history that men exploit marriage in pursuit of their own political ambitions. As usurper, Tāufa‘āhau’s career was largely transgressive.

Normative sociology places structure and event in causal relationship. As a result, it can explain conformity and reproduction but not transgression and transformation. Is there a sociology consistent with the Tongan historical record?

The Tongan title system is part of an encompassing totality that turns on an axis of blood and garland. Formal analysis is thus centred upon this axis. A rule of matrilateral cross-cousin marriage would integrate and align blood and garland statuses, allowing the system as such to be operationalized. However, as Kaeppler has emphatically stated, élite marriage is a matter of choice. Matrilateral cross-cousin marriage “was not prescribed, proscribed, or preferred, but was occasionally practiced among the highest chiefs to prevent social repercussions that might result if a Tu‘i were outranked by a collateral line” (Kaeppler 1971:192-193, n.25). It is therefore impossible to construe élite marriage as the mere “execution” of a rule (Bourdieu 1977:24) or as an “instantiation” of structure. Rather, Tongan élite marriage must be construed as a political tool for positioning actors across status fields. Élite marriage is always strategic — a matter of consolidating or upgrading rank and of mobilizing resources and labour in the pursuit of chosen ends. As Bott observed more than twenty years ago: “In the traditional system the object of the social and political game was to use one’s standing in one system [kāinga] to increase one’s standing in the other’s [ha‘a], marriage being one of the main devices for doing so. The process took several generations” (Bott 1972:219).

Whether a line rises or falls, élite marriage is therefore always a practice. But it is a practice in which the blood/garland system is fully implicated, as the ground of the practice’s pragmatics and politics. Rather than suppressing time, blood/garland creates a number of possible futures. These futures are conservative or radical, depending upon the goals of the players and the strategies they employ. The moheofo institution, for example, resolves the uncertainties of the blood-garland relationship conservatively, by way of reinforcing existing status asymmetries in the title system. Thus, until a certain point, the Tu‘i Ha‘atakalaua, directly subordinate to the Tu‘i Tonga, supplied the Tu‘i Tonga with his principal wife or moheofo (Figure 1). Yet juniors could also withhold a sister from a senior or appropriate the senior’s wife, thus altering the status quo. As the Tu‘i Tonga title began to decline in effective power, the Tu‘i Kanokupolu began sending wives to the Tu‘i Tonga, just as other chiefs began sending women to the Tu‘i Kanokupolu (Biersack 1982 [1974]; Bott 1981, 1982; Fusitu‘a 1976:10).

Sahlins has defined the event as “a relation between a happening and a structure (or structures)” (1985:xiv). For Tonga the word happening is problematic, since it implies happenstance rather than the motivated political entrepreneurship that is so much in evidence in the Tongan historical record. I would reformulate Sahlins’ definition of an event as follows: an event is a relation between a practice — motivated and intentional — and a system — in this case, the blood/garland system. Regardless of whether conjugal practices result in reproduction or transformation, since the system is always fully implicated in the practice as the ground of its pragmatics and its politics, practice remains “internal” but without being determined by the system.

If the relationship between practice and system is thus redefined, formal analysis acquires new powers and also new limitations. Since there is no direct, causal relationship between system and practice, formal analysis alone cannot explain behaviour. However, since the system is the ground of the pragmatics and the politics of practice, formal analysis becomes a preliminary step in understanding the field of historical action and historical action itself. Actors act out of their interests as their position stipulates these interests and by way of exploiting the possibilities for empowerment the system opens up. Historical analysis thus centres on the relation between the system and (rather than the happening) the practice — that is, on the event, construed as a relation between the two (cf. Sahlins 1985:vii). The resulting theory of Tongan history is at once event-centred (Biersack 1991b) and highly political. It understands actual events in terms of an actor’s possible moves and their political consequences, and it is as equipped to explain moments of transformation as it is to explain moments of reproduction. This is precisely the strength, it seems to me, of Goldman’s theory of status rivalry. As he argued in Ancient Polynesian Society, “structure … allows for shifts in power relationships, offering several leverage points for such shifts” (Goldman 1970:304-305). Structure energizes (to use Goldman’s figure of speech) or motivates a field of political actors and informs their initiatives. What becomes characteristic, therefore, is a range of practices, a historical culture (cf. Sahlins 1985).

Attention thus shifts from structure to structuring (Giddens 1979) in the very strongest sense of the word: to activities of making and remaking. Tongans themselves speak of such activities in terms of fa‘u. As a verb the word means “to bring into existence; to make, construct, put together; to build …; to found, institute; to formulate, draw up, making, bring in (a law, etc.) …” (Churchward 1959:147); and as a noun the word means “bringing into existence, constructing, etc.; plan, measure, or institution; thing formed or constructed” (ibid.). These words accent the activity of construction rather than its product; they call attention to a certain kind of agency. Integrating blood and garland in such a way as to uphold the status quo, the moheofo institution was one such fa‘u: not a structure of alliance so much as a structuring through alliance. Whether the action maintains the status quo or encourages change, in either case the hierarchy of titles is generated through marriage as a fa‘u or artefact of historical practices. The Constitution of 1875 is itself the much celebrated fa‘u of King George Tupou I. It retained some titles, dropped others, and strategically added new ones — the Tungī title, for example — erecting a new order of nōpele titles upon the ancient foundations (Biersack 1990a).

The rank of the body is a personal rank, the rank of the historical individual. The present King’s body represents the totality in terms of a historical genealogy: the many acts of structuring and restructuring that the last two hundred years have witnessed. It is a fa‘u. In and of itself it signifies the historicality of political life (cf. Valeri 1985, 1990b) and the historicity of the Tongan ramage.