1. See Vicinus (1982) for a consideration of hydraulic theories of desire.

2. This is a veiled reference to Foucault and his arguments in The History of Sexuality (1978). See Stoler (1995) for a critical consideration not just of how Foucauldian theory might be applied beyond Europe but of how his theory ignored the racial and colonial character of the transformations of sexuality in Europe itself (cf. Stoler 1991, 1992). See Herdt (1993b) for a reinsistence on the need for a theory of repression.

3. The previous few sentences repeat some thoughts from the introduction to Jolly and Manderson (1997) but further develop them.

4. But see the recent work by Marilyn Strathern (1992a) and others on the new reproductive technologies, which observes how this maternal rock has been shaken by distinctions between the woman who gives the egg and the woman who supplies the womb. This has entailed a slippage in notions of the real as against the surrogate mother.

5. This is part of a more general caution about unduly assimilating past and foreign same-sex practices with the erotics and identities of contemporary gays in places like Australia and the United States. There has been much debate by Halperin, Herdt, Weeks, and others about the need to recognize the historical and cultural specificity of ancient Greek, earlier European, and contemporary non-European examples and experiences (but see Elliston 1995 for a brillant critique of Herdt). Thus Herdt (1993b) acknowledges not just the emphasis on growth rather than pleasure, but indeed the pervasive violence and terror that the seniors inflict on the juniors in the Sambia rites. Moreover, in contradistinction to conservative Western models, same-sex practices do not here compromise gender identity of either the passive or the active partner. Inseminating boys in no way compromises but

6. I do not deny men's role or engagement as fathers, nor do I deny that contemporary reproduction in the West can still be seen in terms of male dominance, as is most obviously the case in the politics of abortion in the United States (Ginsburg and Rapp 1995). But individual men are less confidently endowed with patriarchal controls over reproduction than in the past, and women are more often in the United States and Australia the privileged subjects of reproductive choice or control. Of course, the situation is very different between the developing states with predominantly antinatalist policies and those developed states where pronatalism prevails, at least for those of the right ethnicity, class, and ability, where the new reproductive technologies aim to promote, not constrain, fertility for the right sort of folks who can pay. For development of these arguments, see both my introduction and later chapter in Borders of Being (Jolly and Ram 2001). For reflections on the stratified nature of global reproductive politics, see Ginsburg and Rapp (1995), and for a scintillating examination of the new reproductive technologies, see Strathern (1992a).

7. There are important differences and exceptions to this statement. Thus the corpus of Annette Weiner (1976, 1988) on the Trobriands suggests not only that women are central to reproduction there but that men are marginal. The same argument could be developed not just for the matrilineal societies of Melanesia but for many cognatic societies as well. But, as Bonnemère (Chapter 2) shows, the degree to which women are made central or marginal to reproduction is not just a matter of descent or residence principles.

8. Central references in this large literature include Read (1952–1953); Langness (1967); Hogbin (1970); Faithorn (1975); Tuzin (1980); Herdt (1981, 1984); Poole and Herdt (1982); Meigs (1984); and Jorgensen (1983).

9. This is an important concept in Bislama, referring both to indigenes in general as against foreigners and to local custodians of the land as against migrants. I elsewhere discuss its masculinist referent, and the way in which men and women are differently located in the language of place (Jolly 1994a, 1999).

10. Compare Roscoe's analysis of male sexual avoidance (Chapter 12), which he sees as productive in the sense of being constitutive of self. But he focuses on the male self, refraining from considering female sexual avoidance in similar detail. He notes male sexual avoidance before warfare, hunting, fishing, ritual, gardening, artistic production, and football. No doubt all of these require a potent self, but whether they all require minacity is arguable. His argument relies on a neurophysiological argument about sex and aggression, common to male and females it is claimed, and an argument about sexual relapse, which is singularly male. In Sa interpretations, I stress, avoidance produces a more vital and empowered self for both men and women.

11. Compare the insights of the late Jeff Clark apropos Huli sexuality. As well as aptly criticizing Heider (1976), he notes that for the present, the reduction in abstinence and pollution avoidance and greater sexual promiscuity are seen to endanger both health and beauty—of men in particular.

12. I allude to the extraordinary presumptions of some demographers in their analyses of "demographic transition," namely, that prior to modern mechanical and chemical contraceptives

13. This contrasts with those places where intercourse continues since the semen is seen to feed the fetus, or where the blood is seen to be wrapped around or enveloped by the semen (Bonnemère, Chapter 2; Biersack, Chapter 4).

14. In those foods proscribed for the pregnant mother there is a sympathetic connection imagined between the food and the fetus. For example, if she ate clams, the child would cling to the womb and not come out; if she ate octopus, the child would be covered with warts. Coconut is proscribed since the fetus would split in half like the coconut shell and become twins. Such food taboos deny to pregnant women many delicious and nutritious foods. They also underline the clear substantive connections made between foods and human corporeality. As examples of behavioral proscriptions, a pregnant woman should never use a digging stick, lest the water break prematurely, and she should not walk over coral or white rocks, lest the baby be born an albino. Both parents should avoid sacred sites where ancestral spirits may be lurking. If an expectant father wants to impose a taboo on his fruit trees, he must delegate another man to cut and place the mwil (the cycas leaf, a sign of such an injunction).

15. The saltiness of the sea is descibed as nkonkon. Thinking that this might be linked to the state of sacredness, nkon, I probed for possible connections with several native speakers of Sa. They shrugged or laughed, so this was probably bad folk etymology on my part.

16. This is a very common and perduring notion both in the Austronesian and the Papuan-speaking regions of Melanesia, as is attested in McDowell (1988) and Marshall (1985).

17. Patterson (1981) claims for cognate north Ambrym, the island immediately south, that there were elements of both terror and trickery; Allen notes for the Big Nambas that boys were threatened with sodomy by older men (1998). Moreover, in South Pentecost the rites of temat, a so-called "secret society," did involve elements of violence and terror. These have been abandoned since the late 1920s, roughly coincident with the abandonment of warfare and cannibalism.

18. I thank Stephen Hugh-Jones for inviting me to detail further the diving and the way in which it might have ejaculatory overtones. It was never explicitly described to me in those terms by my local interlocutors, who instead stress male beauty and bodily therapy.

19. The higher women's titles did convey ideas of danger and destructive potential—e.g., Wohmat, loosely translated as the power of death—but successive titles were named numerically, as one, two, three, and so on. I should note that although women were not cast as hawks, they were in other contexts strongly associated with birds and with notions of flight and movement.

20. A dubious designation for the various sites of the South Coast, the Papuan Plateau, the Highlands fringe, and the Sepik River region of PNG.

21. This point is made in J. Weiner's critique (1988) and in Hill (Chapter 3) and Biersack (Chapter 4).

22. The commissioners adjudicated that, "So long as the village bure ni sa[the communal men's house] existed and the husband and wife lived in different houses, each under the surveillance of persons of their own sex, secret cohabitation was impracticable. It was made still less possible by the custom of young mothers of leaving their husband's house and going to live with their relations for a year after the birth of a child; but since the bure system has been abandoned, and an imitation of European life substituted for it, husband and wife no longer separate during the period of lactation, but rather give their parole to public opinion to preserve the abstinence prescribed by ancient custom. The health of the child is jealously watched by the other villagers for signs that the parents have failed in their duty" (1896, 146; italics added). The commissioners endorsed the practice with their own instrumental logic: given the lack of stamina of the Fijian mother, her poor nutrition, and the lack of breast milk substitutes, they perceived its benefits and satirized missionaries who saw the isolation of the nursing mother as "absurd and superstitious." But their instrumental recuperation of the practice was rendered difficult by the additional discovery that adulterous affairs by the new father could also cause dabe. Thus "in Namosi, where lactation was continued for three years, a man who had an intrigue with another woman" caused his child to sicken with dabe, or in local idiom, "alien thigh-locking" (ibid.). The commissioners ultimately concluded that "retrogression is now impossible" and that the only feasible remedy is the "use of milk from the lower animals" (p. 148)—a remedy that portended even more infant malnutrition and death.

23. This echoes the way in which women from Vanuatu and the Solomons who went as migrant laborers to Queensland in the late nineteenth century were alike sexualized and cast as prostitutes in indigenous and foreign accounts (see Jolly 1987).