Rereading "Sex and Temperament": Margaret Mead's Sepik Triptych and Its Ethnographic Critics

by David Lipset
Rereading "Sex and Temperament": Margaret Mead's Sepik Triptych and Its Ethnographic Critics
David Lipset
Anthropological Quarterly
Start Page: 
End Page: 
Select license: 
Select License



RereadingSex and Temperament: Margaret Mead's Sepi k Triptych and its Ethnographic Critics

David Lipset

University of Minnesota

n her autobiography, Blackberry Winter, Margaret Mead (1972) recalled the goal of her we1 I-known project in Sex and Temperament in Three Primitive Societies(193511963). She wrote there that she had wanted to show "the dif- ferent ways in which cultures patterned the expected behavior of males and fe- males" (1972:102). A few pages later, however, she admitted that "the subject on which ...[she] had particularly wanted to work" had been rather more defi- nite (1972:205). She was not then just interested in assessing the contribution of culture to defining gender stereotypes, as temperaments are now known, but she really wanted to encounter and study a culture in which the emotions or- dinarily associated with men and women contrasted with one another as well as with those in the West.

Sexand Temperament has been read and reread countless times since it was published in 1935. It has been viewed as a theoretical relic, a passing moment in the history of the American culture concept. Indeed, the imprint of Benedict's configurationalism is strong. This was the ideal that culture as a whole might be conceived as personality writ large, personalities to which actors also might find themselves in more or less conformity. The coherence and integration of culture was seen in terms of prototypical psychological patterns that were acquired or learned during childhood (Benedict 1934). Configurationalism has been dis- placed by an emphasis upon agency, actor and person. Perhaps the more last- ing theoretical contribution of Sex and Temperament might be to the history of modern feminism (Lutkehaus 1995:8). The book has both been praised for its patently constructivist view of gender roles and stereotypes(Rosaldo 1974) while it has also been lambasted for its rigid cultural determinism (Wrangham and Peterson 1966). The publication of Sex and Temperament remains an impor- tant event in the history of both the culture concept and feminist theory. But it also stands out as a unique event in the history of the Sepik ethnology.

It remains an unmatched ethnographic comparison of three Sepik groups: Mt. Arapesh, Mundugumor and Tchambuli. This latter project, the fieldwork for which was done with her husband, Reo Fortune, certainly makes up the bulk of the book. While specific claims about each of the three groups have been the sub- ject of critical attention, no one has bothered to reread these debates by way of putting together an overall evaluation of their claimsas a whole. I want to return, as a Sepik ethnographer, to this dimension of Sexand Temperament. I shall there- fore resume the major claims Mead made about each group and the specific crit- icisms to which they have given rise. I shall then argue that Nancy McDowell's conclusion about Mead's work on the Mundugumor is not quite right. That is, McDowell argues that although Mead was "hampered by the theoretical paradigm she espoused," (1991 :303) her ethnography remained far more complicated than her interpretations. I would say that because of Mead's theoretical interest in the cultural construction of both genders, and her interest in women in particular, the ethnography in Sex and Temperament turned out to be far more instructive, in- clusive and interesting than Mead's configurationalist model of culture permitted her to theorize. Her commitment to understanding divergent cultural patterns of personality of both women and men, and of deviant men and women, sharpened her ear for plural voices in Sepik cultures.

The Mt. Arapesh

When Mead and Fortune spent eight months in the Arapesh-speakingvillage of Alitoa in the Prince Alexander mountains in 1931-32, the people lived in pa- trilineal clans which were localized in small hamlets. In both material and symbolic senses, they were poor. They lived with chronic food shortages. Their gardens and foraging did not yield abundance. Elsewhere, Mead called them an "importing culture" (1938). They conducted trade relations with their Plains neighbors and with the maritime peoples of the Sepik Coast (Lipset 1985).

Mead's major point about the Mt. Arapesh was that both men and women uniformly saw "all life as ...g rowing things, growing children, growing pigs, growing yams and taros and coconuts and sago, faithfully, carefully, observing all of the rules that make things grow" (1963:14). They were cooperative, and passive in temperament and value-orientation rather than competitive and vi- olent. "Arapesh life is organized about ...the way men and women ...unite in a common adventure that is primarily maternal, cherishing, and oriented away from the self towards the needs of the next generation" (Mead 1963: 15). People belonged to the spirits that owned land, rather than vice versa. Ritual forced in- dividual men to act. No one wanted to lead, boast or swagger. Nevertheless, Mead listed four Arapesh categories of person, categories that clearly indicate a plurality of voices in the culture. There were people who would 1) listen and talk, 2) listen but not talk, 3) not listen but would talk, and 4) would neither lis- ten nor talk. Only members of the first category were assigned a competitive feasting partnership called buanyin that Mead designated as culturally unique.

This was not, in Mead's view, a warriors' country. Warfare was "practically" unknown (1963:23). The people lived in a society that was not "accustomed to the raids of headhunters" (1963:6). Women were left unattended. Children tended to stray. Visitors were not held in suspicion. Deaths were left unavenged. But, at the same time, Mead did allow that fighting by men over women did break out from time to time between villages, because marriage created ten- sions when it was done by elopement or abduction. When fighting did break out, it was monitored carefully and resolved via an exchange of shell rings, or through compensation paid to the mother's brother, the man from whom ego inherited his blood. A man who became too aggressive might have his firepit dumped out-a symbol of ostracism-and therefore have to leave the village for a period of time. But such a violent man might not necessarily be brought under control completely.

In Mead's survey of Arapesh socialization for this passive temperament, the mother-child relationship appeared playful and abundant, secure and trust- worthy. Children were not allowed to fight. Emotion was inhibited and not vented either against the self or the other. But even here, differences emerged between male and female which contradict the uniformity argument. Girls learned these expectations faster than boys and acquired their role in the di- vision of labor before boys. Boys, Mead granted, actually had a model of vio- lence. A boy could watch temper tantrums of big men. By the age of seven, the personality of trust and happiness was nevertheless set. He had learned not to be aggressive, to associate eating with warmth and to fear strangers from the plains as sorcerers. At puberty, boys learned to become the custodians of their own growth by avoiding masturbation and by cleansing their genitals through a bleeding procedure that involved inserting a barbed nettle into the urethra.

During male cult initiation, boys were "swallowed" by a cassowary spirit, fat- tened up, beautifully decorated and then "spat back out" into society. While in seclusion, they learned secrets; that the cassowary spirit did not really exist, that it was really a man and that the sacred bamboo flutes were blown by men. They received incisions. But the incisions and the rest of the rite were meant to pro- tect boys and help them grow, not to punish them. The values of the cult were therefore consistent, in Mead'sview, with the maternal pattern of nurture and growth that ran throughout the whole culture. The cassowary spirit was not frightening or physically threatening, but protective, nurturant and growth- oriented. After their seclusion ended, the boys became adults. Their fathers in- troduced them to their hereditary trading partners and then retired in favor of their firstborn sons. The authority of the father, which had no less been based upon nurture and contribution to the growth of his son, now concluded. "I grew the yams," the father might then declare, "I grew you!" (1963: 77). He then gave his property to his son.

Father-son rivalry was muted. Fathers sought to find little girls for their sons to marry, rather than for themselves. Girls who had been so selected were be- trothed when young; the husband was then said "to grow" the girl into his wife and used this activity as a basis for a later claim upon her labor. "Placing a carrying bag on her," was an Arapesh idiom for such child betrothal. Little girls finished growing up in the houses of their in-laws so as to become comfortable with their husbands-to-be as well as his brothers. Upon reaching puberty, a girl underwent a first menses rite. She fasted for several days, was subjected to a vaginal procedure that was held to make her breasts grow. She was bespelled, decorated, given gifts and finally "fed" by her husband. There were no secrets revealed in the girls' rites. A brideprice of shell rings and valuables was trans- ferred and sexual relations then began between husband and wife.

Arapesh wanted love to be associated with "food given and received" (Mead 1963:lOl) and with many years of sleeping together. After long years of bond- ing, Mead avered, husband and wife were made "safe" for each other. The girl was no longer a stranger who might give up her husband to a sorcerer from the Plains. Arapesh men avoided passion outside of marriage. Quick sex was feared. Rape was unknown. They avoided stranger women while on trading expedi- tions because they feared being compromised to sorcerers. Women, they dis- tinguished as being either "big fruit-bats," who were out of control, and "little, gentle bats" who lived quietly and safely in the holes of trees. They especially feared Plains Arapesh women because they viewed them as hostile and sexual- ly aggressive. But then, having acclaimed Arapesh fidelity and fear of extra- marital sexuality, Mead added a striking anomaly to it: a man might seduce another man's wife in order to abduct her.

In order to promote growth, feminine blood from sexuality and birthing had to be kept separate from masculine blood and activities. Mystical pollution resulting from sexual intercourse had to be cleansed by men. Yams would oth- erwise yams fail. Hunting would not be successful. A wife's health would suffer and the children would get sick. A woman, by contrast, only had to cleanse her- self after her first intercourse and then just after the death of her husband.

Still, both male and female possessed identical temperaments. Both were gentle. Both might initiate intercourse. Both might punish children. Their society lacked conflict between age and youth, jealousy or envy. There was little sense of struggle in their world. "To the Arapesh, the world is a garden that must be tilled not for one's self not in pride and boasting not for hoarding and usury but that the yams and the dogs and the pigs and most of all the children grow" (Mead 1963:135). Life was a path with rules to obey. The Arapesh feared what was distant and unknown. They exiled hatred beyond their borders. But then again exceptions crept into her ethnography. Certain men and women were ag- gressive and violent. The society gave them space, but the culture imparted lit- tle meaning to their behavior.

The first attack mounted against Mead's rendering of Mt. Arapesh gender in Sexand Temperament was made by Reo Fortune (1939), by then her ex-collab- orator and ex-husband, just four years after the initial publication of the book. "[Tlhe theory of Arapesh ...culture [as] having the one, uniform tendency, so called maternal, remains," Fortune concluded, "a hypothetical creation" (1939:37). He agreed that the Mt. Arapesh were not headhunters and he agreed that they did not "promote warfare to any vicious extreme." Nevertheless, war- fare did exist. It "was made dependent upon [obtaining] ...women's sexual con- sent in extramarital liaisons outside locality borders, and it was regarded with considerable distaste" (Fortune 1939:37). Land was not fought over. It was less important than people to Arapesh thinking. "The Arapesh express more concern for replenishing the land with children than they do for finding land for their children" (Fortune 1939:26).

Abducting women was the cause and goal of Arapesh warfare. A man might covet another man's wife living in another community. A middle-man, called a bera libere, was then engaged to arrange a tryst in order to seduce such a woman away from her husband. The bera libere was a relative of the woman and so might freely go to her. A disgruntled "clan-brother," he felt slighted by the brideprice distribution and might act as the go-between to gain retribution by fomenting warfare.

She had first to be seduced and to be found willing to run away from her husband. Her seducer had to possess the support of his clan, moi- ety and locality in arranging for her elopement to himself. If the seducer failed to arrange the divorce and the war, the jilted woman was believed to be likely to cause his death by sending his semen to the sorcerers (Fortune 1939: 26-7).

Only one man received sexual access to the woman, the rest of the men just got satisfaction from the fighting. The battle grounds were located on borders and were named. But brawling was restrained; the death of a man or two prompted the vanguished to flee. "The chances for domestic peace, and for con- sequent peace abroad, were very high, and compare more than favorably with the chances for peace in other societies" (Fortune 1939: 37). Arapesh warfare, in any case, was memory-culture, since when Fortune studied it in 1931, it had not been conducted for 17 years.

Fortune described the practice of child-betrothal as follows: "the girl bride- to-be is usually adopted into her future husband's parents' household before she is pubescent" (1939:37). He agreed with Mead that the men believed that women acquired this way "make smooth and good marriages possible ...[The girl] is indebted to her parents-in-law for her maintenance and for her physical growth on the food they gave her" (Fortune 1939: 38). Such women, the men called "nipple-breasted women." Abducted women or widows, by contrast, were called "cut-lipped women." The former were said to be deferent to men while the latter were contrary and defiant.

I think it is evident, despite Fortune's brusque dismissal of Mead's view of Arapesh culture as merely hypothetical, that his reconstruction of Arapesh war- fare and his account of Arapesh marriage do not devastate Mead's major claim that the outstanding pattern of Arapesh gender and culture was uniformly gentle and nurturing. Among one of the neighboring plains Arapesh groups liv- ing in llahita village, moreover, a yam cult reported by Tuzin (1972) also lent some credence to her view (cf. Harrison 1982). lllahita men had very strong emotional attachments to their yams, yams which they personified and ani- mized in many ways. As children with human-like souls, both yams and people were held to have genealogies. Prior to harvest, the yams were said to be socially active in their communities and were actively engaged in gardening. Yet there were two kinds: the short ones, which were impish, and the long ones, which were referred to as "penises" by the gardeners, who viewed them as possessing a rather menacing agency.

There was a strong identification between long yams and male bodily and phallic identities. Both became erect. Female sexuality was dangerous to men and their yams as was menstrual contact. Both suffered from adultery. Long yams were decorated to look like cult initiates. A woman might try to knife her husband's yam, which amounted to an attempt to castrate or kill him. Attacks on yams constituted attacks against the status of its owner.

Yam eating among the llahita Arapesh was understood as both a physical and a mystical act. Child betrothal was meant to give the groom's family the op- portunity to feed and grow the future bride with a yam diet so she could slow- ly become an agnate. Both sets of parents gave newlyweds yams to eat until theirs were ready to harvest. Yams were kept separate until the birth of the first child, after which divorce became illegal.

The valuation of growing is certainly evident in Tuzin's account of another Arapesh yam cult. Mead, of course, recognized that the plains Arapesh were more aggressive than the mountain people and in symbolic contrasts appear between male and female in their respective yam cults, so the former was more competi- tive than the latter. Nevertheless, Tuzin acknowledged that the llahita Arapesh "did not glorify war" and suggested that their "culture of war" (1976:83) did in fact re- semble that of the noncombative Mt. Arapesh described in Sexand Temperament, rather than the more bellicose Arapesh who appear in Fortune's 1939 article.

The Mundugumor

Mead and Fortune left Alitoa village in August 1932 to start a second phase of their joint research.l They made a "perfectly arbitary decision" (Mead 1972: 203) to study a group living on the first tributary off of the Sepik River that had not pre- viously been studied by anybody else. This criterion landed them among the Mundugumor people who were living about 20 or so miles up the Yuat (now Biwat) River from its confluence with the middle Sepik. Unlike the Mt. Arapesh, the Mundugumor were affluent. They occupied a high and fertile ground, divid- ed by small channels that were full of fish. Mead viewed their adaptation and cul- ture historically rather than synchronically. Within the lifetime of the most senior generation, the Yuat River, which had once been nothing more than a little stream, "suddenly" had flooded and split up their territory. When Mead and Fortune reached the Mundugumorvillage of Kenakatem where they would spend about two and one-half months, the people remained horticulturalists and traders who feared the swift current, did not swim, built crude canoes and were only "in slight measure a river people" (Mead 1963: 169). They had only three years ear- lier been "pacified" by the Australian administration in 1929. They were little mis- sionized or subjected to labor recruitment. Mead viewed the Mundugumor as a violent people who dominated the lower Yuat all the way down to its mouth. Dubbing them "a cannibal tribe," she viewed the temperament of the Mundugumor as quite opposite of the maternal, gentle Mt. Arapesh.

The pattern of their culture was aggressive. On the one hand, the people maintained an extensive trade network. They exported garden produce, tobacco in particular, in return for potteryand mosquito baskets manufactured by swamp- dwellers living to the east. They exported the same foodstuffs for axes, bows and hunting magic from the people of the headwaters of the Yuat and they import- ed dances and, in a sense, mortuary rites, from the peoples living at the mouth of the river where it emptied into the Sepik. However embedded in or depend- ent upon regional trade relations, they still dismissed their trading partners as con- temptible, as little more than enemies to be attacked and cannibalized, at least to the extent that they did not eliminate their sources of pottery and so forth.

The external combination of reciprocity and antagonism was no less prevalent within their society. "On their own, which they hold by virtue of a greater ferocity and recklessness than any of their neighbors, the Mundugumor live a state of mutual distrust" (Mead 1963: 173). There was no central plaza. The mas- culine ideal, rarely achieved, was to live alone in an isolated compound, with ten or so wives and sons. "There is no genuine communi ty...Instead Mundugumor so- cial organization is based upon a theory of a natural hostility that exists between all members of the same sex, and the assumption that the only possible ties be- tween members of the same sex are through members of the opposite sex" (Mead 1963: 176). Avoidances divided men from both younger and elder broth- ers. Loyalty to mothers separated men from their fathers. When boys were near- ing adolescence, fathers began to look for new, younger wives. Women were no less hostile to and suspicious of each other. They would curse each other's fish- ing, gossip maliciously, or mock each others' sense of fashion.

The Mundugumor practiced a "form of organization" they called "ropes" (Mead 1963: 176). Note that Mead did not call them groups, clans, or lineages and, at least by implication, actually contrasted them to all of these categories. "A rope," she said, "is composed of a man, his daughters, his daughters' sons, his daughters' sons' daughters; or if the count is begun from a woman, of a woman, her sons, her sons' daughters, her sons' daughters' sons, and so on" (1963:176).The rope was not a way to reckon the inheritance of land, which Mead claimed was so plentiful that it was not highly valued. It was rather a way to reckon the inheritance of cosmic powers, that took the form of water, bush and other spirits, and their manifestations in flutes, masks, etc.

Unlike Mt. Arapesh marriage, which apparently institutionalized pre-Oedipal dependencies, Mundugumor marriage institutionalized Oedipal rivalry. For men, plural marriage was a source of labor, given husbands by wives who grew and cured tobacco, the most important trade good, and by wives' brothers, who of- fered brideservice. While brothers and sisters were separated by contrary rope af- filiations, they were allied through a system of sister-exchange marriage. Brothers became rivals with each other over rights to their sisters, elder brothers having pri- ority over younger brothers. But fathers also became rivals of their sons. just as a son would want to trade his sister for a wife, so his would father, e.g., trade his daughter, in this case, the same girl, to marry again. Fathers came to hate their sons "just in proportion as the son is sturdy and masculine ...The father's jealous regard for his daughter is outraged by his son's claim upon her, and he has a deep- seated hostility to permitting her exchange at all unless that exchange is made at his behest and results in direct sexual satisfaction to himself" (Mead 1963:180). Unlike sons, fathers and daughters were already allied through rope ties. She worked for him, gardened for him, used kinterms reckoned through him and may have actually slept with him in his mosquito basket at night.

While mothers were separated from daughters, they were united with their sons, with whom they slept. Wives therefore became rivals with husbands over the loyalty of daughters to be exchanged so their sons might marry. "Every growing boy has dinned into his ears by an anxious mother the possibility that his father will rob him of his sister, and so of his future wife" (Mead 1963:179). Each compound was divided, as sons matured, into hostile camps composed of disgruntled cowives and jealous, aggressive sons who were set to demand and assert against their father rights to their sisters in order to marry. A pervasive sense of guilt and angry defiance, of fathers defrauding sons over daughter, of obligation betrayed and ignored, beset Mundugumor society. "Although women choose men as often as men choose women, the society is constructed so that men fight about women, and women elude, defy and complicate this fighting to the best of their abilities" (Mead 1963:210).

Girls and boys flirted intensely. By contrast to Arapesh courtship, passion was key, not gradualness. Secret, one night stands in the forest or mosquito basket might turn violent. The kinship system called for aggressive joking relations mocking sexual morality with matrilateral and patrilateral kin, depending up- on the gender of ego (cf. Lipset 1997). Lovers had to step carefully through the taunting of ubiquitous joking partners. In the event that a couple wanted to marry in the absence of sisters, which is to say, should a couple decide to elope, conflict would result between the male kin of the couple. Mead ob- served that "about one-third" (1963:220) of Mundugumor marriages began amid this kind of trouble. Fortune wrote practically nothing about Mundugumor. However the one piece he did publish substantiated this point about conflict over marriage rights (1948).

Men did not want children and became annoyed when their wives got preg- nant. There were mystical health hazards associated with having intercourse with a pregnant woman. Moreover, men tended to refuse to believe in their pa- ternity and would charge their wives with infidelity upon discovering their pregnancy. Wives therefore anticipated that their husbands would desert them when they got pregnant. This male attitude about children was consistent with the rope system: a man had no heirs. His sons were not his heirs and his daugh- ters, whom he favored, were taken from him via the marriage system.

In terms of socialization practices, Mead referred to the background of such a contrary society as one of "maternal rejection" (1963:199) Newborn babies were not loved. They were carried about in stiff bags. They were nursed quick- ly and in anger. The child had to learn to be independent but not to walk off and fall into the river. Weaning was abrupt. The child grew up anxious and fret- ful and faced a world which he divided into a series of binary categories: e.g., play vs. respect relations, elder brother-younger brother, brothers and sisters, ego and affines, water and land, female and male, among others. Children reached a harsh maturity.

The male and female cults reflected this fissile sociology, spartan upbringing and agonistic ethos. There was no central building, no place for men to gath- er. The flute or crocodile sacra were stored in the dwellings. There were sever- al cults, rights to which were inherited through ropes. Initiation into them was a game of prestige for the individual man who sponsored the rite of passage. It was not puberty linked and did not create an age-graded solidarity, or even a common experience. Instead, initiation promoted status differences. "It is all organized about the idea of exclusion and the right of those who have been initiated to taunt and exclude those who have not been" (Mead 1963: 181). Initiation for girls might take place to the degree that they demanded it, while for boys it was constructed as a penalty they might not escape.

At one point in their history, the Mundugumor apparently entertained the idea that male and female descendants of brother-sister pairs inherited ritual prerogatives, for which they might be relatively well compensated, to perform services in alternate generations, eventually to marry in the fourth generation. The idea of hereditary prerogatives led to

the conviction of every Mundugumor that he is doing wrong and that he is being wronged by others ...[Tlhis meticulous observance of obligations through ...g enerations is too difficult for the aggressive individuality of Mundugumor...[A] great number of people are always angry because someone else has been asked to perform the ceremony that they have in- herited the right to perform ...Thus these fantastic provisions for social co- operation between kindred over several generation~ not only do not operate to integrate the society, but actually contribute to its disintegra- tion (Mead 1963:183-4).

The only salient moments of unity, according to Mead, arose at the behest of the most ambitious, arrogant, powerful, treacherous and aggressive individu- als. Such leaders were known as "really bad men," (Mead 1963: 186) of whom there might be two or three in each "community." They were the most suc- cessful polygynists, and used the labor of their many wives and the loyalties of weaker men to cater the planning of war parties and to organize large feasts. In 1932, now that warfare had been banned, they oversaw the creation of the art that would be ritually displayed. The men would be assembled "under the arrogant direction of the master artist" (Mead 1963:188) and then would spend the day together in good spirits preparing the sacred paraphernalia, a 20 foot long crocodile painted with elaborate designs, or a bark triangle, about 30-40 foot tall, that was also decorated. "For several weeks, men who ordinarily dis- trust each other's every move, and hesitate to turn their backs to one another for an instant, work together, while the more level heads scheme to bring ad- vantage to themselves out of the temporary lull in hostilities" (Mead 1963:188).

At the end of her truly compelling discussion of Mundugumor society, Mead briefly introduced male and female deviants who did not conform to its ag- gressive standards. These were men, such as her key informant, Omblean, and women, such as Kwenda, who were disrespected because they were nurtu- rant, gentle and cooperative, good-natured and intellectually gifted, rather than constantly enraged. They were "maladjusted persons, whose gifts were spent in a hopeless effort to stem the stream of an uncongenial tradition, where both men and women were expected to be proud, harsh and violent, and where the tenderer sentiments were felt to be as inappropriate in one sex as in the other" (Mead 1963:233).

The major response to Mead's account of Mundugumor appeared in 1991. Based upon her own research in a village just upriver from Kenakatem in the 1970ssand upon having made a brief visit to this village in 1981, Nancy McDowell then produced a monograph drawing together the bulk of Mead and Fortune's 1932 unpublished fieldnotes (McDowell 1976, 1980, 1991). McDowell confirmed Mead's views about a variety of subjects, conflict, village composition, social structure, e.g., joking relations with matrikin, father-daugh- ter bonds, mother-son bonds, socialization, trade, the sexual division of labor, male initiation ritual, warfare and cannibalism, the direction of change in the society and pollution beliefs. Mead, McDowell concluded, had stood above and beyond 1932 models of change and history. She had not viewed Mundugumor society as in any sense static. But she had seen it as degenerate or broken. The moral order was no longer meaningful; nobody obeyed rules and everybody felt guilty about it.

McDowell's major contention with Mead's account of the Mundugumor con- sisted of two related problems. First, Mead did not give sufficient credit to patri- lineal relationships and the corporate groups they constituted. And second, Mead misunderstood rope relations. The patriclans in Mundugumor society were the small, localized, exogamous groups that owned land, bush spirits, slit-drum beats and personal names. These groups no longer assembled because they were di- vided by rivalries over women. Households remained located on agnatically in- herited land that was safeguarded by similarly inherited ancestor-spirits. Only their masks, flutes, and other insignia, together with ritual services associated with them, were transferred down the ropes, just as Mead had claimed.

If Mead underestimated the import of patriclans in the society, she misun- derstood rope relations. Writing in the margins of Mead's fieldnotes, Fortune is quoted as saying that marriage among the Mundugumor was concerned with a "metaphor taken from the manufacture of fish line, netting, cord or rope as the interweaving of pairs of descent lines of alternating sexes with successive generations as if such lines were the plies of a two-ply cord or rope" (McDowell 1991: 245). However, although she had evidently gotten the image right, Mead's view of rope relationships was, according to McDowell, vague and imprecise. Both Mead and Fortune minimized the relationship of sister-exchange marriage to them. Ropes were not descent related. They were rather a "metaphor for this complex of intergenerational exchanges in which strands of descendants of

intermarrying pairs of brothers and sisters were wove together and reunited in the fourth generation" (McDowell 1991: 266). The firstborn offspring of the founders of two ropes were supposed to marry five generations later.

Mead and Fortune's understanding of rope relations was full of discrepan- cies and ambiguities. Evidently, the ropes were not strictly based in the alter- nation of the sexes. Some of the things passed from father to daughter or mother to son were motivated by sentiment rather than structural rules. While granting that the Fortesian revolution in the study of kinship and descent had yet to occur in the early 1930s, McDowell still argued that she had conflated "de- scent and affect, structure and sentiment" (McDowell 1991: 285). Mead and Fortune gave to the concept of rope "a structural concreteness" it did not de- serve (McDowell 1991: 285). The ropes, in short, were only relationships, not groups, relationships that were constituted by practices, sentiments and ex- changes, not jural rules.

Mead was a splendid ethnographic observer, and if she says that there was hostility between members of the same sex, close ties been father and daughter or between mother and son ..., and that only people of the same sex who were connected through opposite sex relatives, such as mother's brother and sister's son, were friendly ..., then it is probably that these sentiments were characteristic of the relationships. However, positive sentiment is analytically distinct from formal structure (McDowell 1991 :285).

What of Mead's main ethnographic claim about the Mundugumor? Were they a fierce and proud group, consisting of violent, aggressively sexed, ruthless individuals among whom both sexes shared the same temperament? Drawing upon her own impressions of the people she observed in 1981, and drawing up- on impressions of longterm missionary residents among the people, McDowell agreed with this general picture. Moreover, she conceded, "there is no evi- dence to doubt that Mundugumor men and women were ...similar to each other in emotional configuration" (1991:300).

The Tchambuli

Mead and Fortune left the Mundugumor in late December, 1932. One thing led to another (Lipset 1980), and they set up a camp among the Tchambuli people a couple of weeks later. They proceeded to do four months of research in the Tchambuli three villages. The people were not gardeners, but fishermen and traders who lived on the banks of a lake that was connected to the middle Sepik by two channels. A few years before, the Tchambuli had been rousted by neigh- boring latmul and had spent more than a decade hiding out under the pro- tection of their trading partners, a hill people who had been their sago-suppliers. They returned to take back the lake in 1926, following Australian pacification, and were in the midst of rebuilding their communities, particularly the great many male cult houses and holding the rituals through which they were empowered, when Mead and Fortune turned up. Every man was an artist, or so it seemed to Mead, and every man was primarily concerned with and acutely sensitive about his ritual appearance.

Tchambuli society was based upon localized patricians, an exogamous moi- ety system, as well as other feasting relations. Quarrels were frequent among men over slights of honor. The level of conflict reflected incompatible claims by the multiple groups over male allegiances. Men, high strung, suspicious and nervous, were left bickering, to defend themselves as individuals, as fops, fuss- ing over their ornamental appearances, playing a series of charming roles in the ritual system. Mead likened rivalries within the male cult were to those in a bal- let company where the narcissistic ambition of individual dancers conflicted with the needs of the larger choreography.

There were more neurotic men in Tchambuli society, Mead claimed, than she ever observed anywhere else in the nonwestern world (Mead 1963: 307). Women's lives, by contrast, possessed a happy unity. They might not control marriage selection, the exchange of bridewealth and so forth, but they did im- pose their wills upon the domestic side of their husbands' lives. This was the central contradiction in the culture and this was Mead's central claim about the temperaments of the genders within it: beneath the patriarchal forms, "women dominate the scene" (Mead 1963: 270). The "real position of power in the society" (Mead 1963: 253), in spite of the patrilineal ideology, was held by the women who produced and controlled daily resources, who traded the mosquito bags they manufactured and controlled the shells they brought home in return for them. jurally, men could might expect to rule women, but experience showed them that women ruled the day. Tchambuli women nev- ertheless took a tolerant and beneficent attitude toward their ethereal men. They enjoyed their dances and engaged in good-natured horseplay with their masked figures.

The double entendre of the situation, the spectacle of the women court- ing males disguised as females, expresses.. .the complexities of the sex-sit- uation in Tchambuli, where men are nominally the owner of their names, the heads of their families, even the owners of their wives, but in which the actual initiative and power is in the hands of the women (Mead 1963:256).

Boys grew up hearing that their fathers had exchanged bridewealth for rights over their mothers. They grew up seeing that men were stronger physi- cally than women. At the same time, they saw that women got pregnant by their wishes and they saw their fathers continually asking women about their wish- es. They saw that the ritual life was held for the sake of the women and they overheard what a large voice women had in economic matters.

Tchambuli is the only culture in which I have worked in which the small boys were not the most upcoming members of the community, with the most curiosity and the freest expression of intelligence. In Tchambuli it was the girls who were bright and free, while the small boys were al- ready caught up in the rivalrous, catty and individually competitive life of the men (Mead 1972:214).

Of deviants to the contrasting ethos of Tchambuli gender, Mead allowed that it was the men who were notable rather than the women. They appeared in- competent in any number of ways or became unpredictably subject to violent rages. One had forgotten how to speak the vernacular while away doing planta- tion labor. Another had become hysterically deaf. A third man, covered with boils, became increasingly inactive. And then there was Kaviwon, who lost his tem- per and speared his wife in the face, having been "seized by an ungovernable de- sire to thrust a spear into [a] ...g roup of chattering women" (Mead 1963: 263). Tchambuli women, meanwhile, should they deviate from power, as quiet and in- conspicuous "slip along within the comfortable confines of the large women's group, [where they may be]. ..overshadowed...[and]. ..directedn (Mead 1963:272).

Deborah Gewertz, who began to do longitudinal fieldwork among the Tchambuli in 1974 at first by herself and later with Frederick Errington, at first rejected Mead's claim about the power of Tchambuli women. She accepted that Tchambuli women did control daily work and marketing, as Mead described the situation in Sexand Temperament(Gewertz 1981, 1984, Errington and Gewertz 1987, Gewertz and Errington 1991). However, with the possible exception of the temporary and specific historical moment when Mead and Fortune were doing their research in 1933, she disagreed with her claim that the women dominated Tchambuli men. Why? For two reasons. First, Tchambuli gender asymmetries had to be viewed historically or contextually rather than as fixed and static. Secondly, women and men lived according to separate value-systems; which was to say that the women were motivated by different ambitions that did not re- ally overlap with those of the men.

In 1933, Mead encountered Tchambuli men at work rebuilding their com- munities, having returned seven years earlier from twenty years exile with their sago suppliers in the Sepik Hills. She found the men "decorating their ceremo- nial houses with beautiful carvings, manufacturing the graceful double hooks up- on which to hang the highly patterned net bags that they impo rt..., and plaiting the various masks that belong to the different clans and ceremonial groups" (Mead 1963: 244). The women, meanwhile, were reestablishing their barter markets with the sago-suppliers. During the period they lived together, they had learned each others' skills and began to intermarry (Gewertz 1981:105). However prior to this time, at least from the standpoint of the Tchambuli women, their relationship had been hierarchical. The fishwives were superior to the hill women. The political status of men, nine of them to be exact, had been undermined by intermarriages with their sago-suppliers that had taken place during the past years. "After the Chambri returned from exile in the Sepik Hills, the barter system between fish suppliers and sago ~[roducers had to be reestab- lished before ...competition between Chambri men could recur" (Gewertz 1981: 104). When Mead was doing fieldwork in Tchambuli, moreover, one-half of the men between the ages of 15-45 were evidently out of the villages working, Gewertz speculated, on plantations elsewhere in the country. "With so many men away, Tchambuli women had begun to assume control of the trading partner- ships previously dominated by their husbands and fathers" (Gewertz 1984: 621). The nervous, watchful demeanor of the men in 1933, according to Gewertz, had to do with the mixing of affinity with barter relations that was then taking place.

By the 1980s, Gewertz did not see women in control of the goods they pro- duced or for which they bartered. Men did. Patriclan rivalry, and the political system in Tchambuli society, were mediated or conducted in terms of men's ex- change of bridewealth and recipriocal prestations between avunculate kin dur- ing rites of passage. Gewertz gave more methodological weight to male perspectives on exchange and marriage than had Mead.2 Building upon her an- drocentric orientation, Gewertz then developed a second challenge to Mead's view of Tchambuli women, namely, one that was based in Tchambuli concepts of personhood rather than being distorted by the unacknowledged imposition of Western ones. In order to become and remain a maximal person in Tchambuli culture an individual had to belong to a patriclan, possess totemic names, bequeathed both from it and from the mother's clan, and be privi- leged to participate in the ceremonial transaction of bridewealth. The posses- sion of names and the engagement in affinal exchanges were particularly crucial. One provided the cosmic power necessary to transact the other. That is to say, the jural identities of persons, both male and female, were defined by and through formal relations, not emotions or other capacities or subjectivities. Of course, such a view of personhood placed women as incomplete citizens in and of jural society vis a vis men. A woman might choose a man to marry, but her attraction to him was "understood as a measure of his totemic power" (Errington and Gewertz 1987: 55). Women were, in her view, subordinates ma- nipulated by members of rival patriclans.

A few years later, however, she and Errington reckoned that such a view of Tchambuli gender "no longer seem[ed] ...defensible" (Errington and Gewertz 1987: 48). Instead of subordinate, they now deemed Tchambuli women as be- longing to separate "spheres of activity" from men (Errington and Gewertz 1987: 104-10). Since women had neither access, nor the wish, to seek men's power, and since they had their own abilities to achieve goals, "it would neither be accurate to describe [Tchambuli] as dominating [Tchambuli] ...women, nor, as Mead would have it, the reverse" (1963: 254).

While women could bear children in order to reproduce themselves, what- ever men symbolically did to repay their wives' kin was insufficient when meas- ured in terms of the latter's gift of a woman. As a consequence, Tchambuli women were not subject to the same stresses with affines that troubled the men. They did not suffer the humiliation of inequality that the men did, as wife- takers. At the same time as men lived rather isolated lives competing against other men in affinal exchanges, women enjoyed themselves together. Nor were they enthralled by the men's engagement with what Mead had called in 1935 the "minor war-and-peace that goes on all the time among" them. According to Gewertz, "Mead was wrong to describe the war and peace that goes on among the men as minor and unimportant. She is right to suggest that women do not wish to engage in these battles (Gewertz 1984:627).

Day and night, literally, Tchambuli men lived in a constant fear that their wives might betray their totemic powers by revealing the secret names in which those powers were invested, names which the women might learn from men talking in their sleep. Although acknowledging their vulnerability to Tchambuli women when dreaming, the men were not dominated by them. By the same token, men had to "accommodate themselves to [the women's] ...wishes, for as a discontented wife, she is likely to steal her husband's secrets and give them to a lover of her choice" (Errington and Gewertz 1987 :82). Nevertheless, Tchambuli men and women should rather be understood in separate Tchambuli categories, idioms and concepts, rather than in terms of asymmet- rical power relationships as both Mead and then Gewertz and Errington them- selves had once done.

Here, I think it is fair to say that Mead's major ethnographic claim about the cultural patterning of gender in one of the three groups-basically that Tchambuli male-female temperaments were contrary rather than uniform, that women, despite the men's jural privileges, were powerful, unified and cheerful while the men were dependent, isolated, prickly and rather desperately narcis- sistic-received its most comprehensive re-examination. To what degree does it stand or fall, as a result? Iwant to defer my discussion of the methodological crit- icisms raised by Gewertz and Errington for the conclusion of the article. Let me just observe, for the moment, that, despite all their protestations that Mead had ignored the effects of the historical moment upon Tchambuli gender, and that Mead had forced unsuitable constructions of personhood upon her data, the picture of Tchambuli men and women by which they adorned the cover of their 1987 monograph, Cultural Alternatives and a FeministAnthropology, is one of male absence; it rather presents a woman and her daughter, their gaze directly cast at the photographer, without hint of reticence.

Mead's Prescience I have presented Mead's views of the pattern of gender stereotypy found in the social structures, ritual systems, socialization processes, etc. in each of the three Sepik cultures appearing in Sex and Temperament. I have rehearsed the criticisms to which her views have been subjected by Fortune, McDowell, Gewertz and Errington. It remains for me to account for my own views of both the ethnography and the commentary upon it. I think it is possible to identify two streams of criticism, minor and major, weak and strong, that have been lev- eled against the ethnography in Sex and Temperament.

Both McDowell and Gewertz attacked Mead for failing to distinguish between emotion and formal structure. McDowell argued that the Mundugumor "ropes" did not constitute descent groups, as Mead had understood them to do. The ropes were not groups. They were merely categories of relationships constitut- ed through exchange, sentiment and actor-initiated events. Similarly, Gewertz ob- jected to the privileged position of emotion, intention and experience in Mead's view of Tchambuli gender, a view that she criticized as an ethnocentric distortion of the indigenous constitution of male and female personhood.

All three of her critics essentially argued that she gave insufficient weight to men's voices and male values and, implicitly, that she overstated those of women. Fortune denied the view that Mt. Arapesh men and women were uni- formly gentle and oriented toward growth. The men were men. They were war- riors, not just nurturing mothers. While accepting Mead's portrayal of a uniformly fierce Mundugumor temperament, McDowell argued that Mead underestimat- ed the import of patrilineal ties in their society. Lastly, Gewertz (and Errington) rejected Mead's characterization of men's subordination to their wives, again be- cause of having underestimated the import of patriliny and male cosmogony.

The minor issue, the conflation of emotion and structure, results from Mead's adherence to the personality writ large, configurationalist view of culture. Because it essentially accuses Mead of failing to make use of theoretical dis- tinctions that had yet to develop, I find it less interesting than the major prob- lem, e.g., that Mead diminished men's contributions to, or definitions of, their cultures. Whether or not we accept the validity of this second line of criticism, it might be apt to ask why, when taken together, this problem seems to emerge so consistently in this critical discourse. There are probably several kinds of re- sponse to this question. But Ithink an important point to make is that Mead was trying to study both Sepik women and men during the Sex and Temperament project in a way unprecedented for its evenhandedness. Writing in her Tchambuli fieldnotes in 1933, she herself said as much: the "key to ...emotional tone [at least among the Tchambuli] to be found in the organization of the women's groups and the relationship of children of both sexes to women rather than to men" (Mead quoted by Gewertz 1984: 616). There is no shortage of male-cen- tered ethnography in the three sections of Sexand Temperament, But what is re- markable is the rendering in this book of female-centered ethnography.

If Mead's ethnography was influenced by her own concerns and those of both Sepik women and men, then Iwould think that a methodologically more interesting reconceptualization of it might be through a concept of culture in which both male and female voices play on and off, with and against, each oth- er (see Lipset and Stritecky 1994; Barlow and Lipset 1997). The positivist frame- work employed by her critics according to whom they discriminate correct (male) from incorrect (female) versions of one institution after another, I find much less convincing or compelling. What is striking to me upon rereading Sex and Temperament in 2003, in other words, is not the failure of configura- tionalism, or the excesses of Mead's cultural constructionism, or even her mis- understanding of particular institutions or relationships. I am rather struck by how Mead's ethnography of the Mt. Arapesh, Mundugumor and the Tchambuli seems to have succeeded to privilege or foreground actions, voices and dispo- sitions of both women and men, almost a half-century prior to the emergence of a feminist anthropology (see also Reiter 1975), not to mention long before the recognition of polyphony and dialogism in ~ulture.~


This article was originally presented at the Annual Meetings of the American Anthropological Association in 2001. It has benefited from critical readings from James Boon, Donald Tuzin, and Eric Silverman.


'What happened to the people of Alitoa following Mead and Fortune's departure? Paul Roscoe reported that they were occupied by the Japanese during W.W. II. The soldiers wiped out game and plundered their pigs and gardens. American and Australian airforces then bombed communities in the region: Alitoa was hit directly. In 1991, Roscoe himself saw scorch marks still visible on coconut palms in the old village site. After the war, thepeople slowly abandoned the peaklands, eventurqally to resettle near Dagua on the coast. In the 1990s, after some period of land disputes with their hosts, plans were made to return to Alitoa. But by 1997, only a few families had actually built new houses on ancestral lands, however. Paul Roscoe,'ln the field: return to Alitoa.'

*Gewertz explicitly admitted that she privileged men's voices during her research. "I was ini- tially prohibited entrance into any of the's houses, which line the shore of Tchambuli Lake...l protested the decision. My argument was a simple and effective lie. I told my hosts that Ihad to visit them in their men's houses if I was to adequately complete my work. If they didn't wish me in them, I'd move to their enemy village, where I'd be quite welcome every- where" (Gewertz 1984: 618).

30ne could argue that she shared this dual orientation with Bateson, who also wrote about the dual ethos of men and women among the neighboring latmul of the middle Sepik River (see Bateson 1958).


Barlow, Kathleen and David Lipset. 1997. "Dialogics of Material Culture: Male and Female in Murik Outrigger Canoes." American Ethnologist 24 (1): 4-36. Benedict, Ruth. 1934. Patterns of Culture. Boston: Houghton Mifflin

Chodorow, Nancry. 1974. "Family structure and feminine personality." In Women, Cultureand Society. Rosaldo, Michelle and Louise Lamphere, eds. Pp. 43-66. Stanford: Stanford University Press.

Errington, Frederick and Deborah Gewertz. 1987. Cultural Alternatives and a Feminist Anthropology: an Analysis of Culturally Constructed Gender Interests in Papua New Guinea.

Cambridge: Cambridge University Press. Fortune, Reo. 1939. "Arapesh Warfare." American Anthropologist 41 :22-41. . 1948. "Law and Force in Papuan Societies."American Anthropologist 49:44-59.

Gewertz, Deborah. 1981. "A historical reconsideration of female dominance among the Chambri of Papua New Guinea." American Ethnologist 8:94-106. . 1984. "The Charnbri view of persons: a critique of individualisim in the works of Mead and Chodorow." American Anthropologist 86:615-629. Gewertz, Deborah and Frederick Errington. 1991. Twisted Histories, Altered Contexts: Representing the Chambri in a World System. Cambridge: Cambridge University Press. Lipset, David. 1985. "Seafaring Sepiks: Ecology, Warfare and Prestige in MurikTrade."Annual Review of Research in Economic Anthropology 7: 67-94. a n d Jolene M. Stritecky. "The metaphor of mutedness: gender and kinship in Seaboard Melanesia." Ethnology 33 (1994): 1-20. McDowell, Nancy. 1976. "Kinship and exchange: the Kamain relationship in a Yuat River vil-

lage." Oceania 47:36-48.
. 1980. "It's not who you are but how you give that counts: the role of exchange in a
  Melanesian society." American Ethnologist 7:58-70.
. 1991. The Mundugumor: From the Field Notes of Margaret Mead and Reo Fortune.
  Washington D.C.:Smithsonian Institution Press.

McLean, Ann. 1991. "In the footprints of Reo Fortune." In Ethnographic Presents: Pioneering Anthropologists in the Papua New Guinea Highlands. Terence E. Hays, ed. Pp. 37-67. Berkeley: University of California Press.

Mead, Margaret. 1935/1963. Sex and Temperament in Three Primitive Societies. New York: Morrow. . 1938. "The Mountain Arapesh I. An ImportingCulture."Anthropological Papers of The American Museum of Natural History 36 (Part 3).

. 1972. Blackberry Winter: My Earlier Years. New York: Morrow. Reiter, Reyna R. 1975. "Introduction." In Toward an Anthropology of Women. Pp. 11-19. NewYork: Monthly Review Press.

Rosaldo, Michelle. 1974. "Woman, culture and society: a theoretical overview." In Woman, Culture andsociety. Michelle Z. Rosaldo and Louise Lamphere, eds. Pp.17-42. Stanford: Stanford University Press.

Roscoe, Paul J. 2001. "In the field: return to Alitoa."

Tuzin, Donald. 1972. "Yam symbolism in the Sepik; an interpretive account." Southwestern

journal of Anthropology 28 (3): 230-54. . 1976. The llahita Arapesh: Dimensions of Unity. Berkeley: U. of California Press Wrangham, Richard and Dale Peterson. 1996.Demonic Ma1es:Apes and the Origins of Human

Violence. New York: Houghton Mifflin.

  • Recommend Us