1. Mangrove and the names of all individuals used in this work are pseudonyms. I have told the people of Mangrove that I would not use their names in anything that I publish. I use a pseudonym for the community to provide some privacy from the outside world.

2. Though some construe the words "action" and "behavior" to signify a different order of things (e.g., Harré and Secord 1973 and Reynolds 1976 in Klama 1988; Taylor 1964 in Gergen 1984), I use them synonymously. In using either word, I intend to indicate that I am talking about something that people do rather than something they only feel or think, but I never mean to imply that thoughts and feelings are not a part of specific human behaviors.

3. There is a growing literature that attempts this kind of self-reflection. For a recent example, see Kondo (1990).

4. Speaking of "human conflicts," Martin Daly and Margo Wilson write,

Self-report methods, so often adopted out of expediency, are of dubious validity at the best of times, and perhaps especially so for studying conflict. People may or may not be prepared to discuss their hostilities and affections, and they may or may not be willing to predict their own behavior in various hypothetical situations, but even if we could be sure that our subjects were being forthcoming, we would not have much reason to trust their introspections. (1988:11-12)

Not only did I find Aboriginal people willing to discuss their "hostilities" and "predict their own behavior in various hypothetical situations" but I also found many reasons, which should be evident in these pages, to "trust their introspections'' (see also Burbank 1988). Indeed, Aboriginal "introspections" have provided the impulse and much of the substance of this book's presentation.

Daly and Wilson go on to say that "homicide, by contrast, is drastic action, with a resultant validity that all self-report lacks" (1988:12). It should be noted, however, that insofar as they did not observe at firsthand the homicides they discuss, they too rely to some extent on the reports, predictions, and introspections of others. Undoubtedly, the majority of homicides are reported to the police rather than witnessed by them. The obvious role of report, predictions, and introspection in the collection of any kind of "case" is underlined when Daly and Wilson mention that they include in their analyses cases of homicide in which the "police have identified the killer to their own satisfaction, regardless of prosecution or conviction'' (1990:85; 1988). While the police may be regarded as "experts" with regard to Western homicides, so may Aboriginal people be recognized as knowledgeable authorities on the kinds of aggressive behavior manifest in their community.

5. Of interest here is Susan Abbott's (1983:173-174) discussion of conflict with a Kikuyu woman that she experienced during her fieldwork in rural Kenya. According to Abbott, it was precisely because the women perceived Abbott to be powerful that she attempted to thwart her research project, competing with her "to gain influence among local women." While I think I share the motives of writers who characterize ethnographic subjects as "oppressed," I must agree with Diane Bell who points out that this stance can be "both misleading and demeaning for all parties" (1990:158-159). It strikes me that in characterizing other women in this manner, we deny both the ways in which they can be powerful and the ways in which we can be powerless, two possibilities that must be considered in discussions of ethics and responsibility.

6. Some might argue that my access to such items as Land Rovers, fans, and refrigerators bespeaks the "very existence of privilege that allows the research to be undertaken" (Patai 1991:137). In my experience, however, possession of these items did not give me power over Aboriginal people. Such power depends on very specific constructions of material objects and concepts of "possession," constructions that are not necessarily shared by Aboriginal people (Gerrard 1989). I should also point out that while my material circumstances in 1977-1978 and 1981 were unlike those of many Aboriginal people, by 1988 they were quite similar insofar as people at Mangrove had gained much greater access to Western material goods, including refrigerators and vehicles.

7. I believe that this participation is also based on some knowledge of what I do. I have from the beginning of my work at Mangrove told people that I am an anthropologist and have talked with them about the general nature and substance of my work. As is the case with any Aboriginal community today, to work at Mangrove, it is necessary to obtain permission from Mangrove's Aboriginal Council and from the Northern Land Council. In addition to writing to the council at Mangrove about my proposed research, I tried to discuss it in general terms with as many people at possible. Following the guidelines of the Australian Institute of Aboriginal and Torres Strait Islander Studies (formerly the Australian Institute of Aboriginal Studies), the Australian research institute that has sponsored my work at Mangrove, I have sent copies of reports (Burbank 1980 b , 1987 c ), a paper (Burbank et al. 1989), and my previous book Aboriginal Adolescence (1988) to the council, clinic, school, and specific people

in the community, providing them with an opportunity to see what I do with what I learn. Many people at Mangrove read, often for pleasure, and several have told me that they have received, read, and found what I write of interest.

8. It is within this largely integrated community that I formed my primary relationships, and it was from men and women in this integrated community that I came to learn about Aboriginal life. Discussions of past social forms are confined to those of the predominant language group at Mangrove.

9. Deborah Rose reminds us that the American Anthropological Association statement of Professional Ethics (1973:1) includes the following: "[Anthropologists] bear a professional responsibility to contribute to an 'adequate definition of reality' upon which public opinion and public policy may be based" (quoted in Rose 1986:28). While Rose would focus on the "physical, economic, and political deprivation which people suffer" (ibid.), she also acknowledges "that not all Aboriginal communities are experiencing such distress and there may be much to be learned from comparative studies" (ibid.:24). I would add that an "adequate definition of reality'' must also include the identification of factors that ameliorate suffering and engender well-being.

10. Bronislaw Malinowski's book on the Aboriginal family in Australia includes the following statement:

Ill-treatment [of wives] is—in the primitive state of the aboriginal society—in most cases probably a form of regulated intra-family justice; and that although the methods of treatment in general are very harsh, still they are applied to much more resistant natures and should not be measured by the standards of our ideas and our nerves. (1913:82)

I wish to make it absolutely clear that my argument bears no resemblance to Malinowski's assertions about Aboriginal senses and sensibility. When I propose differences in the experiences of American and Aboriginal women with male aggression, I focus on differences in meaning and social context, creations of social experience, not of putative differences in physiology.

1. I occasionally recorded Aboriginal speakers on tape. My transcription of this account reads as follows.

[Man speaker]: All the Aborigines been here. [Woman speaker]: Here we been only on top, ceremony, we've been dancing in olden days. [Man]: Aborigine been all over, Billy Macoy story, before, shooting olden day. Used to come in a helicopter and shoot people. They used to hate them, white fellow used to shoot black people. Him [refers to woman speaker] husband working on those boats. [Woman]: My husband, I been little girl yet, I no been married, spear, this way [name], him been get killed. One girl him been taking. [Man]: No more one girl, big mob prisoners. [Woman]: All the prisoners they been taking. [Man]: Billy Macoy, they speared him and let the people go.... [VKB]: What for those whitefellas been coming shooting everybody? [Man]: Long time ago, before like me been coming, white like you skin but black like me...colored man...came here and made a lot of trouble.... Billy Macoy and Mr.McCall.

Keith Cole mentions that in 1932, Aboriginal people killed five Japanese pearlers at Caledon Bay, an area to the north of Mangrove. He also reports that Aboriginal people then killed the police investigator of these murders, Constable McColl. Soon after these deaths, two "white beachcombers were also killed." According to Cole, members of the Church Missionary Society's "Peace Expedition" persuaded the "self-confessed killers" to return with them to Darwin to stand trial, thus averting "a demand for a police punitive party to go and 'teach the Caledon Bay Aborigines a lesson''' (1977:194; 1982). According to Nancy Williams (1987:1), news of the planned punitive expedition reached the south, and a "public outcry'' resulted. A consequence of the outcry was the establishment of a mission by the Methodist Missionary Society at Yirkalla in 1935. The mission was conceived as a "buffer" between Aboriginal and non-Aboriginal people.

2. This group of people, described by Warren Shapiro (1977), had only a minimal presence at Mangrove in 1977 and 1978. According to Cole (1982), almost all of them had left the settlement by 1981. In March 1988, however, according to my residence survey, there were approximately three dozen people from this group living at Mangrove. I had very little contact with them.

3. H.C. Coombs, B. G. Dexter, and L. R. Hiatt (1980) have described the events and policies leading to the establishment, in 1974, of a federal program that provides financial support for outstations. In 1977, groups of Aboriginal people were leaving Mangrove to set up small communities on their clan lands. By 1981, eleven outstations had been established. In 1988, the outstations seemed to hold little attraction for most of the people I knew. One young woman periodically took supplies to her father who along with his wife were the sole inhabitants of one station. Others were visited occasionally at best. I only visited two in 1988. One of these was deserted by all but the occasional visitor, its garden abandoned, its tin houses empty and falling down. I was told that a third outstation, one that had previously held the largest year-round population, had recently been devastated by a cyclone.

Natural disasters aside, there are many possible reasons that the outstations have not provided more of a long-term alternative to settlement residence. Among these are both the attractions and compulsions of Mangrove. In 1978, for example, when parents took school-aged children to the outstations during the school year, there was talk of fines for absenteeism and withholding of "child checks," a government allowance paid to the parents of schoolchildren. I do not know if such financial penalties were ever actually imposed. The perceived threat, however, may have contributed to the smaller outstation population in subsequent years. Though outstation life can be seen as a means of regaining autonomy and preserving Aboriginal practices (Altman 1987; Burbank 1988), their residents, nevertheless, continued to depend on many of the goods and services provided at Mangrove. Trips to the settlement were made, for example, to obtain foodstuffs from the shop and medicine from the health clinic. The outstations' relatively small populations and lack of popular entertainment may also contribute to their small population. In 1988, the people I knew expressed much more interest in visiting "town" or other settlements to attend such events as Bible meetings, dance festivals, or rock concerts. Alcohol

may also be a factor affecting outstation residence. Like Mangrove, its surrounding outstations are legally "dry." Thus Aboriginal people who want to drink, as a number do, are more likely to visit towns or settlements where alcohol is available than an outstation where it is not. The use of outstations as "punishment" for adolescents who misbehave may also reduce the desire of young people to visit them voluntarily (Burbank 1988).

Abandonment of Mangrove and a permanent exodus into town, while an option taken by some, does not appear to be seen as an attractive alternative by the majority of the settlement's residents. Attachment to land and kin, in and of themselves, are probably sufficient reasons for people to remain on or near the settlement.

4. According to Nancy Williams, Aborigines have always been "fully subject to Australian law," although their status as Australian citizens has varied over the years since 1788. Her work on law at Yirkalla suggests how the "two laws" have worked at Mangrove.

Just as Yolngu consistently associated big troubles with Australian intervention, so they linked little troubles with the legitimacy of Yolngu modes of dispute settlement. Their creation of these categories represented a prediction that white intervention would occur in big troubles but would be unlikely in little troubles.(1987:129)

In 1986, the Australian Law Reform Commission recommended limited recognition of Aboriginal customary law. Marcia Langton (1988) notes that in most areas of remote Aboriginal Australia two systems of law have been in operation. She also observes that in some parts of Australia police tolerate the Aboriginal "law maintenance devices" of swearing and fighting. Other recent anthropological discussions of the relations of the "two laws" are to be found in Williams (1988) and Bell and Ditton (1980).

5. My determination of a "case" is guided by my own definition of an "event" as well as the perceptions of Aboriginal people. Generally, when I observed or was told of an aggressive event such as a ''fight'' or an "argument," I counted this as one case of aggression. Sometimes, however, a series of aggressive events that would, say, be described as a "fight" by Aboriginal speakers include an act that can be regarded, either according to Aboriginal categories or mine, as another form of aggression. For example, in the course of fighting with his wife, a man might strike her, then throw a spear at the shop. In order not to "lose" the act of spear throwing (an act we might conceptualize as displaced aggression) by subsuming it in the fight, I have scored two cases, noting, however, that the act of throwing a spear was related to the fight. In 145 instances of the 793, I have coded aggressive cases as "related." The number and content of aggressive cases, then, are as much a reflection of how Aboriginal people and I have chosen to slice up a stream of action as they are of that stream itself. This scheme should be compared to those employed by others, such as Pilling's "trouble cases" (1957) and Hiatt's "dispute cases" (1965), before any intercommunity comparisons are made.

Beyond defining "aggression" and "cases," I have also made a series of methodological decisions, necessary in any research, about what to include and what to exclude in the total number of cases. I have, for example, not included cases of sorcery, sexual violence, vandalism and juvenile delinquency, or the

punishment of children—all acts that could be regarded as aggressive. During the years that I have visited Mangrove, I have been told of many cases of vandalism and juvenile delinquency and perhaps four or five cases, both past and present, that might be defined as rape, attempted rape, or sexual molestation. I have excluded these cases from my analysis because at the time of my fieldwork they seemed to be regarded as a different order of things. I remind readers that the specific acts detailed in these pages must be kept in mind when discussing "aggression" at Mangrove (Macaulay 1985).

The reader should also note that these cases of aggression are not a random sample. It must be asked if the women who told me of the vast majority of these cases tended to talk more about their own aggressive acts, thus introducing a form of sex or gender bias into the data. I examine this possibility elsewhere (Burbank 1992 a ). To summarize briefly, a series of statistical tests have suggested that the possibility of gender bias is negligible.

6. A major division made by Aboriginal women in their categorization of aggression is between physical and verbal acts. I also use this categorization but go beyond it for theoretical reasons that are evident in the various discussions of aggressive behavior in this book. For example, the Aboriginal women I know tend to call any kind of physical aggression a "fight." I make further distinctions; for example, I have coded acts of physical aggression into six different categories: "fights" (where physical aggression is threatened or used between at least two adults), ''displays" (where an object rather than a person is attacked; see chap. 6), "displays within fights" (where the attack of an object takes place during a fight or appears to be a substitute for attacking a person in a fight), "discipline'' (adult aggression directed against an adolescent who occupies a "child" relationship to the aggressor; see chaps. 2 and 3), "murder" (at least one person is killed), and "mirriri" (the aggression of a "brother" against a "sister" following a breach of etiquette; see chap. 6). In addition, I use the category "fight?" for cases where people use the word "fight" in an account but it is not clear if physical aggression actually was threatened or took place. In organizing types of aggression, I have also created categories for "miscellaneous acts" (including abduction and destroying property but not in a display), "cursing" (as described in chap. 6), "threats" (both verbal and physical), "suicide attempts," and acts that are "not clear."

7. All z scores are derived from one-sample proportion tests (Hintze 1990).

8. These cases probably include less verbal aggression than occurred because of a coding artifice. For reasons of economy, I coded only one action—the one judged to be the most severe or destructive—for each participant in a case. For example, if an aggressor both picked up a rock and struck her target with her hand, the latter action would be scored. Thus, in cases in which verbal aggression precedes physical aggression, the verbal aggression is "lost."

9. I use the word "girl" to indicate that I am talking about an immature female, one who is not regarded as a "woman" by her community (see Burbank 1987 a , 1988). I also use the word "girl" in my conversations with Aboriginal women because they use this word to talk about themselves.

10. These numbers should be regarded as an approximation because ages, necessary for calculating the number of adults in residence, were not in all cases

known. Following the Department of Aboriginal Affairs' census practice, people age 15 and over are counted as adults.

1. Though I tend to follow the spelling of Heath (1982), I do not use his orthographic system for the sake of readability.

2. As I understand this speaker, the husband would "get jealous" because the women's laughter indicates that they are talking about lovers. I was not aware of lesbian relationships at Mangrove. Women's homosexual practices in Aboriginal Australia are discussed, if only briefly, by Berndt (1965), Reay (1970), Kaberry (1939), and Roheim (1933).

3. For accounts of trysting in Arnhem Land, see Shapiro (1981), Warner (1937), Berndt and Berndt (1951), and Burbank (1988).

4. The word "tease" in Aboriginal English lacks all connotations of playfulness. I think it is best understood to mean "attack," either verbally or physically. I find this understanding of a word borrowed from the English vocabulary particularly interesting with respect to Miller's (1991) supposition about the construction of masculine anger in Western society. Briefly, she proposes that particular constructions of anger and aggression develop in interactions that fathers have with their sons. The boys are teased but not allowed to express anger and hurt, because it is, after all, ''just teasing." If boys express their anger, the father withdraws from the "game" and may punish the child. This leads, says Miller, to a more general masculine stance in which, ''the game is played with the pretense that no one really is hurt" (191). If we say that Aboriginal constructions of anger allow greater recognition of potential harm, are we not also saying that some constructions better reflect human experience than others?

1. Others have made similar observations about aggressive events in Aboriginal Australia (see Langton 1988; Macdonald 1988; Sansom 1980; Stanner 1968; Warner 1937) and elsewhere (e.g., Bohannan 1960 b; Cummings in press; Fox 1968, 1975; LeVine 1961).

2. Other observers of Aboriginal social life have postulated a relationship between patterns of aggressive events and aspects of the childhood environment. Hamilton (1981 b ) has drawn a parallel between the "violent and unreasonable demands" that men make of women and temper tantrums displayed by

children over food. Myers (1988) sees children's anger as "paradigmatic" of that expressed in adult life. Grayson Gerrard (n.d.:17) sees wife beating by adults and mother beating by children as "part of the same continuum." David McKnight has made the following remarks about links between childish and adult aggression.

Although I do not hold that the violent behaviour of the Mornington Islanders can be attributed simply to their childhood experience, nevertheless there is some connection and it is plain that the aggressive behaviour of children and adults is all of a piece. Children are continuously exposed to violence and they soon learn to regard violence as a way of life. Initially when I observed children screaming and crying when their parents were involved in fights I thought the children were terrified. I eventually learnt that I had misinterpreted their reactions, for they in fact often enjoyed the spectacle. This was borne out by people's comments and by the children's animated and carefree accounts. Most children seemed to have strong egos and they would stubbornly assert themselves. The were apt to strike out when thwarted. They frequently threw tantrums, smashed things and took sulk ( bunme ). The children fought each other and took sides in the traditional windward and leeward manner—even my five-year-old son was recruited. The childhood game was throwing tin lids at one another which the children blocked with sticks, or else they rolled the lids on the ground and tried to hit them, using their sticks as spears.... Needless to say these games are excellent training for hunting and fighting. (1986:156)

3. Some of the people known by any individual at Mangrove are described as "close" or "full" ''relations.'' We might call this segment a "kindred" (Shapiro 1979, 1981). Individuals are said to be "full relations" on the basis of perceived genealogical links (e.g., a woman might say, "We are all one mother's mother"), shared clan membership, or a shared "Dreaming." The totality of an individual's "full relations" are described as "family."

4. According to Warner's (1937; see also Williams 1987) accounts of makarata , a very similar form of ritual revenge, a spear thrust through the offending party's thigh indicates forgiveness, but a shallow wound or no wound at all indicates a lack of forgiveness and/or intentions to take further vengeance. What may be a contemporary deviation from the earlier form might represent a modification designed to placate the legal concerns of the larger Australian society.

5. Past and present forms of these body operations include circumcision, scarification of chest and upper arms, and nose and ear piercing.

1. In 736 of the 793 cases, I have been able to ascertain the sex of the aggressors and targets. I have devised seven categories for a variable I call "Sex of Aggressors/Targets." These include cases where (1) men attack women (MAW); (2) women attack men (WAM); (3) men and women are aggressors, but it is not clear which sex began the attack (MWAG); (4) men attack other men (MAG); (5) women attack other women (WAG); (6) a male aggressor

does not have a specific human target (SMAG) (see chaps. 3 and 6); (7) a female aggressor does not have a specific human target (SWAG).

The cross-tabulation of "Sex of Aggressor/Target" with the variable "Reason" is significant (chi-square with 90 degrees of freedom = 240.2872, p < .0001). The chi-square value in the cell in which MWAG and "Subsistence" intersect is 4.1, that is, greater than 3.0, a number conventionally accepted to indicate significance (Fienberg 1977). I interpret significance to indicate characteristic reasons for aggression for the various gender combinations of aggressors and targets.

It should be noted that this use of cross-tabulation is intended as a close heuristic rather than as a direct test of the significance of the differences. The chi-square is approximately z ² and therefore approximates a difference of proportions test for the hypothesis that the particular cell does not differ from its expected proportion under an equal distribution. The chi-square value of 3 is approximately a z of 1.73 with a one-tailed probability of .05. This value is sufficiently close to acceptable levels of significance to warrant further evaluation of the relationship (see Fienberg 1977).

2. The actual phrase used in my question was "young boy," the Kriol and Aboriginal English words for adolescent males.

3. This is not to say that men's contributions are not important to the well-being of women and children. See Burbank and Chisholm (1990) for a discussion of this issue.

4. This statement reflects the old adage, "It is a wise child who knows its own father." Men cannot always be sure that the child their wife bears is their own. Women, however, unless their children are removed at birth and placed in the hospital nursery, know that they are the mother of their child. Similarly, relatives of a woman can be more certain than relatives of a man that their kin's children are indeed their genetic relatives. The potential difference in "paternity certainty" is an important theoretical issue in sociobiological discussions of gender relations (e.g., Daly, Wilson, and Weghorst 1982; Burgess and Draper 1989).

5. Aboriginal women at Mangrove use the words "find" and "found" to speak of a child's mother and presumed genitor. According to Shapiro (1979), this terminology reflects the practice of dreaming the child. In a dream, a man encounters the spirit of the child and directs it to its mother. At Mangrove, people dream children, but fathers do not usually dream their own children (Burbank 1988).

1. Maccoby and Jacklin include in their summary of an early review of the research on "sex differences" the following passage:

We have seen that the greater aggressiveness of the male is one of the best established, and most pervasive, of all psychological sex differences. We have also seen reason to believe there is a biological component underlying this difference. (1974:368)

There have been rebuttals of their conclusion that there is a biological substrate to this difference (e.g., Tieger 1980; Fausto-Sterling 1985) and discussions that cast doubt on the quality of data derived from experimental psychology (e.g., Deaux 1985; Frodi, Macaulay, and Thome 1977; Macaulay 1985). However, aggression manifest in observed "real world" patterns of male and female experience continues to be "one of the best established and most pervasive" gender differences. For example, cross-cultural explorations find that men are, with few exceptions, the warriors in any given society (Adams 1983; Ember and Ember 1971; Ember 1981; Whyte 1978). There is also consistent evidence that killers are usually men (e.g., Bohannan 1960 a; Daly and Wilson 1988; Fry 1992; Knauft 1985, 1987 b; Lee 1979).

When the focus is on nonlethal interpersonal aggression, however, the question of difference becomes more complex (see Frost and Averill 1982). There is evidence both for and against our perceptions that men are more aggressive, and we must always specify what we mean by "more" (Burbank 1987 b , 1992 a; Björkqvist and Niemelä 1992 a ). David Levinson, for example, in a cross-cultural study of family violence, finds wife beating "the most common form of family violence around the world," occurring in 84.5 percent of the societies surveyed; husband beating, in contrast, occurs in far fewer societies (26.9%) and generally occurs less often than does wife beating (1989:30-31). Similar cross-cultural distributions of spousal violence have been found by Broude and Greene (1983) and Whyte (1978). Cross-cultural comparisons have also found boys to be more aggressive than girls (Ember 1981; Maccoby and Jacklin 1974, 1980; Rohner 1976; Whiting and Edwards 1973). However, Rohner's (1976) cross-cultural survey did not find that men were significantly more aggressive than women. Some reviewers of the largely experimental psychological literature have found few consistent or robust sex differences in aggression (e.g., Epstein 1988; Frodi, Macaulay, and Thome 1977; White 1983). Differences, however, always in the direction of greater male aggression, continue to be found (e.g., Eagly and Steffen 1986; Hyde 1986, 1990).

2. Overviews of feminist studies of woman abuse are provided by Breines and Gordon (1983), Bograd (1988), and Liddle (1989). Feminist thought characterizes studies done within psychological, sociocultural, and sociobiological frameworks. Examples of psychological approaches are provided by the work of Walker (1979, 1984); examples of sociocultural approaches, by Campbell (1992) and by the papers edited by Counts (1990 b ) and Counts, Brown, and Campbell (1992). Examples of feminist approaches within sociobiology are provided by Burgess and Draper (1989) and Smuts (1992).

3. In 606 cases, I have been able to code a reason for an aggressive event. (See table 3, p. 99, for details of the coding categories; see chap. 5, n. 1, for a discussion of my use of statistical procedures.)

4. Liddle has said that feminists make a series of assumptions about "individuals, consciousness, and interests" when they attribute the motive of dominance to men's use of violence against women. He also points out that "feminist

characterizations of 'male interests' or 'the interests of men' are quite undeveloped" (1989:764). Gregor has used a similar strategy for interpreting rape in our society. He argues that there is little to support the idea that it is intended by men as a political act. He says with reference to women's experience, however, that "the implications for gender politics are far-reaching" (1990:482).

5. Yengoyan's (1990) analyses of 29 dreams of aggression from Pitjantjatjara men suggests the extent to which cultural constructions may be internalized in the individual psyche.

6. I asked this speaker why the dugong would be cursed when a man danced "really beautifully." He replied that it was because people thought it was a really good ceremony.

7. Several women at Mangrove have told me that women do use supernatural means to woo men as lovers or husbands. One woman labeled some of these acts jarrada . In the literature on women's ritual activity, the term "jarrada" is more commonly used to refer to a complex of women's rituals that may include attempts to manipulate male/female relationships (see Burbank 1989, for an overview of the literature on women's ceremonies in Aboriginal Australia). One older woman who attempted to introduce jarrada of this kind at Mangrove told me that some of the older men told her to stop after a couple of performances. The ceremonies were too successful, and young women were running off with inappropriate partners. Currently, women's attempts to entice men through supernatural means appear to be isolated, individual acts.

Although this is clearly manipulation of the supernatural and an attempt to compel someone to do something they might otherwise not do, I do not think that jarrada should be included in a discussion of aggression. Neither the means nor the ends of these acts are intended to harm another. Although magically separating a woman from her husband, or a man from his wife, could be described as an aggressive act insofar as it leaves an injured spouse, the purpose of jarrada seems to be that of obtaining a spouse or lover rather than of hurting the previous husband or wife.

8. Before phrasing this question, I had heard Aboriginal people using the word "frightened" to describe their reactions to seeing things like spears, snakes, and fights (see Burbank 1988). The seven kinds of aggressive events I asked about were "gun fired in the air," "you are hit with a stick,'' "you are chased with a spear," "shop is cursed,'' "you are cursed," "someone wrestles with you," "someone swears at you."

1. Anne Campbell (1982, 1984 a , 1984 b ) and Marsh and Paton (1986) have found overtly aggressive behavior to be accepted and employed by British working-class schoolgirls, "lower-class teenage girls," and American black and Hispanic gang members, girls and women ranging in age from 15 to 30.

2. Many of these studies can be found in the volume edited by Björkqvist

and Niemelä (1992 b ) and in a special issue of Sex Roles edited by Fry (in press). Other references include Browne (1987, 1988), Burbank (1987 b , 1992 b ), Campbell (1982, 1984 a ), Cook (1993), Marsh and Paton (1986), Schuster (1983, 1985), and Schuster and Hartz-Karp (1986).

3. At Mangrove, children are sometimes hit, but physical aggression that is injurious is proscribed and punished:

If a mother hits the child for something too hard or makes it bleed the father will hit her.

If a father hit [a child] too hard, the child's brother, uncle, or [grandfather] would make a big fight and tell the man not to hit hard.

According to these criteria, there is little child abuse apparent at Mangrove. In 1981, for example, a "European" nursing sister, who had worked on the settlement for just under two years, reported that no cases of child abuse had been brought to her attention. In 1988, an Aboriginal health worker reported that one mother had "bruised" her daughter but also said that she had ceased this behavior after the health worker and other women spoke to her.