On Yanomami Warfare: Rejoinder

by Bruce Albert
On Yanomami Warfare: Rejoinder
Bruce Albert
Current Anthropology
Start Page: 
End Page: 
Select license: 
Select License



KINNES,I. A., AND T. L. HIBBS. 1989. Le gardien du tombeau:
Further reflections on the initial Neolithic. Oxford Iournal of
Archaeology 8:I 59-65.

LE ROUX, C-T. 1979. "Stone axes of Brittany and the Marches," in Stone axe studies. Edited by T. H. McK. Clough and W. A. Cummins. CBA Research Report 23.

1981. Circonscription de Bretagne. Gallia Prehistoire 24:395-423.

-. 1983. Circonscription de Bretagne. Gallia Prehistoire

-. 1984. Apropos des fouilles de Gavrinis (Morbihan):
Nouvelles donnees sur l'art megalithique armoricain. Bulletin
de la Societe Prehistorique Frangaise 8I :240-45.

LE ROUZIC,Z. 1927. DepBts rituels de haches en pierre polie de- couvert dans la region de Carnac. Bulletin de la Societe Prehis- torique Frangaise 24:I 56-60.

-. 1930. Carnac: Restaurations faites dans la region, les
cromlechs de Er-Lannic, Commune de Arzon, de 1923 a 1926.

Vannes: SocietC Polymathique du Morbihan.

1. 1983. Les idoles qu'on abat, ou Les vicis- sicitudes des grands stkles de Locmariaquer. Bulletin de la Societe Polymathique du Morbihan 119:57-68.

MEILL AS SOUX,C. 1967. Recherche d'un niveau de determination dans la socikte cynegetique. L'Homme et la Societe 624-36. RENFREW,

C. 1976 "Megaliths, territories, and populations," in Acculturation and continuity in Atlantic Europe. Edited by S. de Laet, pp. 198-220. Bruges: de Tempel.

SHEE TWOHIG,E. 1981. The megalithic art of Western Europe.
Oxford: Clarendon Press.

Yanomami Warfare:


ORSTOM-Universidade de Brasilia, Caixa Postal 07-1121, 70359 Brasilia, D. F., Brazil. 24 VII 90

In my critique of Chagnon's 1988 article [CA 30: 637- 40) I showed that his measure of "the amount of vio- lence in Yanomamo culture" as well as his statement ;hat an alleged status of "killer" ["unokai") promotes higher male reproductive success are questionable on both ethnographic and theoretical grounds. I also traced the historical roots of the cultural assumptions that underlie his image of the Yanomami as "fierce people." Since his only comment on this point is that I find Dar- winian theory "repulsive" [CA 3I : 5 I) I will let it go at that2 and examine his other assertions [pp. 49-53).

I. Several months of work with physicians among the Yanomami of Brazil have prevented me from writing this rejoinder sooner. I am grateful to D. Buchillet, P. Menget, A. Ramos, and A. Quesnel for their helpful comments.

2. Except to note that on a recent rereading of Leviathan I have found that Chagnon is closer to Hobbes than I had suggested (CA

30:639 n. 10): "It may peradventure be thought, there was never such a time, nor condition of Warre as this; and I believe it was never generally so, over all the world; but there are many places, where they live so now. For the savage people in many places of America, except the government of small Families, the concord whereof dependeth on natural lust, have no government at all; and live at this day in that brutish manner, as I said before" (Hobbes 1988[1651]:187).

I. Warfare and interethnic contact. I showed that the Shamatari Yanomami of the Mavaca-Siapa region (Orinoco River basin Venezuela) studied by Chagnon present a much higher percentage of male mortality in warfare than any other Yanomami subgroup on which such data are available and that this percentage is smaller than those for other Amazonian groups3 whose ethnographic images are much less conspicuously war- like. Stressing Shamatari specificity [which Chagnon once emphasized and now plays down [see 1974:chap. 4 and 1988:991 n. ZI]), I stated that it may be related to early historical changes introduced in this population well before "first contacts" with whites through direct or indirect contact with surrounding indigenous groups [see Posey 1987). This is the only part of my argument to which Chagnon objectsI4 accusing me of "invoking un- known and undocumentable factors such as mysterious effects of a 'population explosion' or the alleged hostility of unknown Arawaks who lived in this region before the Yqnomamo penetrated it" [p. 51). Ironically, my "invo- cation" of these "undocumentable factors" was based on Chagnon's (1966: 167) own attribution of the Shamatari's demographic expansion to their early acquisition of steel tools and their free access to open territories [on the Shamatari's higher rate of population growth, fission, and warfare than their Namoweiteri neighbours, see Chagnon 1974:129-32). AS a matter of fact, although ethnohistorical research on the Yanomami is still inci- pient, it has been established that, during the 19th cen- tury, in several regions they were at war or trading- often to acquire steel tools-with various neighbouring Caribs, Arawaks, and other groups whose territory they now occupy in whole or in part (see Albert 1985:40-42).

As for the "unknown Arawaks," the record shows that the Yanomami who preceded the Shamatari in the Mavaca-Siapa drainage (see CA 30:638 n. 5) were en- gaged in fighting with the Arawak-speaking Anauya and Kuriobana (see Lopes de Araujo 1884:54, 56; Cerqueira 1928:78-79; Stradelli 1889:23; Chaffanjon 1889:247, 292, 295).= It-is also known that the Bare of the Cassi- quiare Canal used to take captives from among these Yanomami [see Spruce I908: 3 I 6, 3 56) and that the Man-

  1. Waorani and Achuara-to which we could add the Mekranoti Kayapo, with pre-contact male mortality in warfare calculated as 42% (Werner 198324, table 7).
  2. In a footnote (n. 7) Chagnon attempts to support his assertion of "the signal importance of violence as a determining factor" in Yanomami culture by mentioning Alks (1984) on the Parima high- land Yanomami, but Alb explicitly disagrees with his utilitarian conceptions about competition over women and the related theory of Yanomami warfare (see pp. 97-99 and p. 111 n. I). Moreover, Smole's data on the Parima highlands (1968-70) suggest a low war- fare intensity in that area (1976:74, 233 n. 105).
  3. During the 19th century another Arawak group, the Mandawaca, also occupied the Siapa (Schomburgk 1841:249) and Pacimoni (Spruce 1908:427) Rivers, while the Mawaka, also Arawakan, were on the Mavaca River (Spruce 1908:408). Humboldt (1819:572) en- countered Yanomami at the colonial settlement of La Esmeralda (upper Orinoco) in 1800, and in 1857 Michelena y Rojas (1867:354) heard from acculturated Indians of Santa Isabel de Mavaca (upper Orinoco) that they used to trade with the Yanomami (known as Guaicas or Guaharibos).

Volume 31, Number 5, December 1990 I 559

dawaka and Yabaana (also Arawakan) were expelled by

the Yanomami from the Cauaboris and Marauia Rivers

(upper Rio Negro, Brazil) at the beginning of the 20th

century (Knobloch 1975 : 143-44). Finally, two recent

studies have provided persuasive data and discussions on

the influence (ancient and/or recent) of techno-econo-

mic change and contact on warfare intensity among the

Yanomami (see Colchester n.d., Ferguson 1989b). All

this evidence seriously challenges Chagnon's rendition

of the Yanomami he studied as an isolated group retain-

ing pristine patterns of warfare (1983[1968]: I; see Head-

land and Reid 1989 on the anthropological attachment

to the representation of small-scale societies as pristine


2. Unokaimu ritual and "killers" record. I obiected to Chagnon's taking the percentage (44% ) of males 25 years or older who had performed the unokaimu ritual6 as a measure of "the amount of violence in Yqnomamo cul- ture" (1988:989) on two grounds, one empirical and the other logical. I noted that the accuracy of his record of men who "have participated in the killing of someone" ("unokais") might have been impaired by the distorted way in which he used the unokai category and then pointed out that "participation" in a killing in this con- text cannot be lum~ed with the Western notion of homicide. Chagnon's counterarguments regarding my first objection transform it into a methodological straw man. He attributes to me several fictitious assumptions about his research in order to contrast them with his "real" procedures of recording victims and "killers." My criticism was centered on the misuse of the unokai ritual category, and the only assumption underlying it was, of course, that he must have used that category in his inquiry. His 1988 paper indicates as much when it states that the "unokais" are "widely known within the village and in most neighbouring villages" (p. 987). In a subsequent publication he says that he "eliminated 'symbolic' unokai" from his record of "killers" through employment of an alleged native distinction between "false" and "true" "unokais" (1989a:24), again implying that he used that concept. Now (p. 50) he suggests that he never employed it except to confirm the obvious. i.e., that the "killers" he identified (with what concept but unokai?) had undergone the unokai condition.

To my second objection, that is, the validity of his "unokais" record for "measuring" Yanomami "violence" in a comparative perspective, Chagnon has no response. Yanomami killings in warfare include collec- tive arrow shooting of the same victims in combat7 and subsequent wounding of dying or even dead victims (see also Lizot 1989:109). A11 warriors who have injured an enemy in these different ways consider themselves to be in the state of unokai (see Albert 1985 : chap. I I). Can a

  1. Unokai is a condition of symbolic impurity resulting from the physical or supernatural killing of an enemy; unokaimu is the ritual that neutralizes it (see Albert 1985: chaps. 7-11).
  2. Forty-eight percent of the victims recorded in Chagnon's data

were killed by more than one warrior and 24% by 3 to 15 warriors (P.50, fig. 11.

record of these "unokaied" men be equated with statis- tics on homicides in the Western sense of the term (see Knauft 1987:463)? It is doubtful, and that is my point.

3. Raiding and abduction of women. In questioning Chagnon's speculation that the abduction of women by "unokaied" warriors might enhance their reproductive success, I quoted the only quantitative data available at the moment on marriages by abduction in the context of Yanomami warfare-the firmre of 0.8% calculated for a


region very close to Chagnon's research site (see Lizot 1988:540-41). I could have added that in the Parima highlands, adjacent to Lizot's area of study and taken to be the historical center of Yanomami territory, abduc-


tion in warfare is said to be practically nonexistent (Smole 1976:230 n. 22; Ales 1984:97). Chagnon claims

(p. 51) never to have said that Yanomami raids begin exclusively as fights over women or are exclusively motivated by abduction strategies. Nevertheless, many of his descriptions of Yanomami conflicts over women come very close to this:'

. . . most drastically, the woman shortage is remedied

by raiding other villages to abduct women. [1966:69]

. . . most Yqnomamo warfare and intra-village fighting

is directly attributable to competition over women.


. . . wars between villages usually begin in a contest

over the possession of some woman. [1976:17]

. . .much of the fighting. . .is explained, by the Yqno-

mamo, as attempts to get revenge. Still, the wars al-

most invariably begin in dispute over women.

[1977:87 n. 21

Most fights begin over sexual issues: . . . The most

common explanation given for raids . . .is revenge . . .

and the most common explanation for the initial

cause of the fighting is "women." [1988:986]

He goes on to raise doubts about the accuracy of Lizot's data, despite the fact that he has made favorable remarks on Lizot's data collecting and on the collaborative work they developed together (see Chagnon 1975:1og n.2; 1976:14-IS).

Finally addressing my point, Chagnon argues first that to calculate a percentage of abductions only on the basis of current marriages underestimates their rate and that Lizot's data are irrelevant for his research area. I agree that the rate of abductions in current marriages is not a satisfactory index, but, although not ideal, it is some- thing that Chagnon has never provided. He would be the one to provide quantitative data to support his assertion on the marital success of Yanomami "killers" through mate abduction, and until my 1989 CA criticism this does not seem to have crossed his mind.9

  1. Even his more nuanced formulation of the cause of Yanomami warfare stresses that "although few raids are initiated solely with the intention of capturing women . . . once raiding has begun be- tween two villages . . . the raiders all hope to acquire women" and that "the Y~nomamo themselves regard fights over women as the primary causes of their wars" (Chagnon 1983[1968]:175-76)
  2. Similarly, for over ten years he insisted on a link between Yanomami warfare and female selective infanticide without giving


Challenged by my criticism, Chagnon comes up with

a (preliminary) rate of marriage by abduction in his area

of study: 17%." Unfortunately, he does not distinguish

abduction from enemy villages and from allied villages

as Lizot did (0.8% and 0.9%)~ although he notes that

"most abductions are not the consequence of raiding" (p.

51). Two conclusions can be drawn from this: that the

rate of abduction of women in warfare in his area of

study is low, as in Lizot's research area and in the Parima

highlands, and that his 1988 hypothesis is that "un-

okaied" warrior might gain reproductive success not by

taking captives on raids from their enemies but by secur-

ing women from their allies. These points modify his

portrayal of Yanomami warfare.

Nevertheless, the discrepancy in levels of abduction of women from allied villages between Lizot's and Chag- non's data remains a puzzle. In view of the differences between the Shamatari and their Namoweiteri neigh- bours (Chagnon 1974: 127-32), whose populations were pooled in Chagnon's data base, it might be interesting to check whether this rate is characteristic of both "popu- lation blocs" or specific to the Shamatari. At any rate, Chagnon has not yet provided any statistical evidence of the gains "killers" ("unokais") may have in marital and reproductive success through the abduction of women, whether from allied or enemy villages.

4. Unokai and "cultural success." Pointing out the precariousness of his ethnographic evidence, I challenged Chagnon's postulation of a prestigious status for warfare "ki1lers"l"unokais" (see also Chagnon 19go:g~) that, as an indicator of cultural success, would confer upon them a special "attractiveness" as mates in mar- riage arrangements. Chagnon's reply to this is anecdotal except for the statement that he is "aware of other meanings and nuances" that the concepts of unokai and waitheri "have in other contexts" but his article "was not intended as a contribution to semantics or lexicog- raphy" (p. so). The issue at stake in my comments on these categories is, of course, more serious than this.

We have seen that, because of the organization and ritual context of Yanomami raiding, many men may very well undergo the ritual condition of unokai without having mainfested bravery. Only a few men achieve a supralocal reputation for valour in intervillage raiding and are said to be waitheri.1° Having undergone the con- dition of unokai repeatedly is one of the qualifications for this reputation." Chagnon recognizes this: "Most

any statistical evidence for it. He finally abandoned this assertion (see Chagnon, Flinn, and Melancon 308-10 and Melancon 1982:227).

  1. Contradicting himself about its prestige, Chagnon stresses the (secondary) negative sense ("aggressive," "wild") that waitheri has in some contexts (see Migliazza 1972:421-22; Lizot 1989:107).
  2. Bravery is a necessary but not a sufficient condition for being considered waitheri, as this concept also connotes, outside the raiding context, being daring and tough and having authority, ini- tiative, and a generous and humorous temperament (see Albert 1985:97-98). No special status is accorded waitheri men either in my area of study or near where Chagnon worked (see Al&s 1984:108; Lizot 1989:104). Waitheri, in the sense of "resistant to

killers have unokaied once. Some, however, have a de- served reputation for being waiteri (fierce) and have par- ticipated in many killings" (I988: 987). But this nuance complicates his argument. If we consider the waitheri reputation based on multiple unokai experiences rather than a hypothetical "unokai status" as an indicator of cultural success, then for his record of "killers" to estab- lish a link between prestige and reproductive success he would have to limit it to multi-"unokaied" men; only 21% of the men (29 of 137) in his unokai count have undergone this ritual condition three times or more (1988:986, fig. I). Moreover, he has recently suggested that men who engage in lethal violence "with some moderation" do better reproductively than people who "are excessively prone" to it (1989b:566). The hypoth- esis thus becomes even more difficult to handle: its uni- verse, initially all "unokaied" men, has had to be re- duced first to multi-"unokaied"lwaitheri men and now to "moderately" multi-"unokaied"lwaitheri men. The problem, of course, is how to draw the lines between these categories.

Chagnon has presented no proof of the existence of a causal relationship between a "status achieved via . . . prowess in military activities" (p. 5 3) and reproductive advantage. At best, he. offers an apparent statistical cor- relation between participation to any degree in warfare killings-mostly by occasional warriors without special prestige-and higher maritallreproductive performance (1988:989, tables 2 and 3). As he himself has put it, he has "only speculated" on the mechanism that might connect these phenomena (1988:989) that, to him, "seem to be correlated" (p. 53). This correlation may well be due to confounding variables. Chagnon has dis- cussed some possibilities (1988: 989-90)) and so has Fer- guson in his criticism of that paper (1989b). At least one has not yet been satisfactorily resolved: that the "appar- ent higher fertility of unokais" might be achieved at the expense of higher mortality rates" (Chagnon 1988:990; see also Ferguson 198ga:564; Chagnon 1989b:5 66). Chagnon raises this question only to reject it on the basis either of anecdotal information (1988:990) or of "impressions" (1989b:566). No conclusive evi- dence has yet been provided. This is a crucial question in view of the fact that many Yanomami ethnographers (including Chagnon) have stressed that men taken to be responsible for deaths in warfare (especially the mul- ti-"unokaied" warriors) are preferred targets for revenge raids (Albert 198 5 :305; Al?s 1984: 108; Chagnon 1983[1968]:180; 1988:985; Lizot 1989:104).

If there is anv correlation between warrior re~utation and reproductive advantage left after the ethnographic and methodological shortcomings pointed out so far have been resolved, Chagnon will still have to provide a satisfactorily deductive model of causality linking these two orders of pheaomena. Such a model will of course have to be built on more than rhetoric-for example,

pain" or "ready to defend oneself," is also applied to women and children (see Bortoli 1983: 18; Lizot 1989: 107).

Volume 31, Number 5, December 1990 / 561

assertions about "tribal violence," cross-cultural mili- tary prestige, or female preference for male "winners." To explain Yanomami warfare, it will also have to be freed from what Bateson called "the old teleological trick" (see Houseman and Severi 1986:8), that is, the functionalist confusion of effects with transcendental final causes. Chagnon will therefore need to establish the mechanisms through which reproductive strategies may lead to warfare among the Yanomami and not the other way round (see Robarchek 1989:906). Then the challenge for him will be to make sense of his model in genetic terms (see Morange 1986:133-34 and Weiss 1976:369-70).12

4. "Tribal violence" as an ethnocentric construct. There is, or course, no such thing as violence per se: the characterization of anv social interaction as "violence" is culture-bound (see Jamin and Lenclud 1984:9-10, 17- 18; Michaud 1988:chap. I). The Western notion of vio- lence, rather than defining a cross-culturally identifiable unitary class of phenomena, covers a wide and varied range of behaviours and representations. The semantic root (vis) of the Latin term violentia refers to the mani- festation of physical force (the Yanomami see waitheri as an energetic emanation of the vital principle of animated beings [see Albert 1985:147]). The use of this force against someone constitutes "violence" when there is transgression of a social norm (this notion goes back to the late 18th century [see Domenach 1980:32]). There are, thus, as many forms of violence as there are norms to define it (on the complexity of Western jurid- ical and statistical classifications of types of violence, see Chesnais 1981:12-14; Michaud 1988:chap. 2). The Western notion of violence, being inseparable from the idea of transgression, when applied elsewhere inevitably conveys an ethnocentric normative referent.

From this perspective, claims to measure "violence" in "tribal societies" as if it were a universally defined phenomenon amount to reifying the outcome of a cul- turally biased assessment. It is obvious, then, that, no matter how obiective the statistics of "violence" mav seem to be, they cannot be completely free of implich value judgements. Even the cross-cultural use of the classic homicide rate per ~oo,ooo is problematical be- cause of the cultural diversity of types of killing and of subsets of the population involved (see Knauft 1987:463

12. The remainder of Chagnon's argument is a challenge to me to provide a non-Darwinian explanation of his data on higher variance in male fertility and "competition over females." This challenge is, of course, rhetorical, as it adds no supporting evidence to his 1988 hypothesis. Moreover, male differential fertility results from sev- eral factors among which polygyny is not necessarily dominant (see Howell 1979267-73). It is not, therefore, a priori an indicator of "competition over females." Chagnon has yet to provide an in- depth demographic analysis of male fertility and polygamy from his complete data base (contrast Howell 1979:chap. 13). The two figures presented [p. 52) show a higher fertility variance among males, although inflated to some extent by two men with 23 off- spring each, than among females. These figures compare men and women 40 or older living in 1988. Such a comparison seems ques- tionable in that it does not deal with completed reproductive ca- reers and does not take into account the differential mortality be- tween men and women.

for a discussion of some aspects of these difficulties).

There is no neutrality, for example, in comparing

Yanomami rates of violent death, mostly due to intervil-

lage raiding (collective/legitimate violence), with West-

ern rates of criminal homicide (individual/illegitimate

violence). Nor would mortality in warfare be easily com-

parable, the subpopulations at risk being defined by dif-

ferent social and cultural criteria. Yanomami intervil-

lage raiding avoids, for example, women and children as

targets and aims particularly at multi-"unokaied"

men.13 Things are, of course, very different in the con-

text of Western high-technology warfare.

Chagnon has published indices that are supposed to

measure Yanomami "violence." His imposing percent-

ages give the impression that this is indeed a violent

society. But on what basis is this impression created if

none of the indicators used are cross-culturally compara-


5. Cultural representations and the anthropology of war. As a social institution warfare presupposes a culturally defined interdependence among politically organized collectivities (see Aron 1984[1962]; Sahlins I98oa: 32-3 7). The minimal and widely accepted defin- ition of warfare as "armed conflict between political units" is based on these premises (see Aron 1984[1962]:34, 326; Otterbein 1973:923-24; McCauley 1990:1-2). Warfare cannot be reduced to a mere resul- tant of individual aggressive behaviours and utilitarian strategies unless we are prepared to deny it as a social institution (contrast Chagnon 1974:77 and 1ggo:79-80, 85-86). Seen in this light, Yanomami intervillage raid- ing (niyayu 'to shoot arrows at each other') is to be understood in the context of a complex cultural con- struct consisting of a classification of sociopolitical dis- tances, a theory of physical and supernatural aggression, and a system of symbolic exchanges via funerary and war rituals (see Albert 1985; 1988:89-93; n.d.).

Since his Ph.D. dissertation, Chagnon has given min- imal ethnographic attention to the intellectual and ritual aspects of Yanomami warfare, treating them as if they were exotic epiphenomena (see 1967:27; 1968: I 3637; 1983 [1968]:181-82, 186). With his new biological reductionist approach this disregard for warfare symbol- ism has been elevated to a theoretical remise. To explain Yanomami raiding in terms of universal male competition over "means of reproduction" (Chagnon 1990:81) is to deny by definition any relevance to the complex framework of indigenous representations and institutions that informs the armed action. This neo- functionalist approach precludes any comparative per- spective on the cultural diversity of warfare (cf. Menget 1985-86).

Indigenous theories of aggression and revenge, concep- tions of social difference, and the associated rituals are

13. Percentages of adult female mortality in warfare vary considerably among Amazonian societies: Shamatari 4%, Namaweiteri 7% [after Chagnon 1974:160, table 4.1o), Kayapo 23% [Werner 1983241, table 71, Achuara 27% (Ross 1988:57), Waorani over 50% [after Larrick et al. 1979:167, table 7).

essential constitutive dimensions of warfare as a social institution, regardless of military action as such and the external constraints that may shape its intensity and organization. These ideational and institutional con


structs have their own ontological autonomy and cannot be reduced to mere by-products of another order of real- ity.14 This does not mean that reproductive, environ- mental, and contact-induced factors have no effects on various aspects of warfare; it means that these effects manifest themselves through the cultural framework that constitutes warfare in a given society and do not cause it.

To neglect the ideationallritual dimension of warfare amounts to excluding a priori most of the social reality to be analysed, thus inevitably leading to impoverished (unidimensional) models of explanation. How can an an- thropological analysis of warfare omit the symbolic rela- tionship between raiding and supernatural aggression when accusations of killing by sorcery are so impor- tant in triggering cycles of intervillage raiding (see Albert 1985:302-4; Alirs 1984:99; Chagnon 1968:112; 1983[1968]:170, 175-76; Lizot 1989:106; Peters 1987:

81; Smole 1976:50)? HOW can intervillage raiding be

-. n.d. Guerre et Cchange symbolique: Reflexions ipropos du cas Yanomami. MS. AL~S,

C. 1984. Violence et ordre social dans une societk amazonienne: Les Yanomami du Venezuela. Etudes Rurales 95-96:89-114.

ARON,R. 1984 (1962). 8th edition. Paix et guerre entreles na- ticns. Paris: Calmann-Levy. BORTOLI,J. 1983. Yanomami: Politica como mediacidn de las re- laciones sociales. La Iglesia en Amazonas 14-15:16-28. CERQUEIRA,

D. E. DE C. 1928. Reminisc6ncias da fronteira. Rio

de Janeiro: F. Briguiet. CHAFFANJON,

J. 1889. L'Orbnoque et le Caura: Relation de voy- ages exbcutb en 1886-1887. Paris: Librairie Hachette.

CHAGN oN, N. A. 1966. Ygnomamo warfare, social organization, and marriage alliances. Ph.D. diss., University of Michigan, Ann Arbor, Mich.

-. 1967. Ygnomamo: The fierce people. Natural History 76122-31.

-. 1968. "Ygnamamo social organization and warfare," in War: The anthropology of armed conflict and aggression. Edited by M. Fried, M. Harris, and R. Murphy, pp. 109-59. New York: Natural History Press.

-. 1972. "Tribal social organization and genetic microdif- ferentiation," in The structure of human populations. Edited by

G. A. Harrison and A. J. Boyce, pp. 252-82. Oxford: Clarendon Press.


1974. Studying the Ygnamamo. New York: Holt, Rinehart

and Winston.

understood apart from its ritual context when the -. 1975. Genealogy, solidarity, and relatedness: Limits to lo- cal group size and patterns of fissioning in an expanding popula-

simulation of the exocannibalism of the enemies respon- 16(1):1&18.

sible for that death (see Albert 1985:353-6o)? The task

for a non-reductionist anthropology of warfare is thus to

unravel the specific cultural constructs of that phenom-


1979. "Mate competition, favoring close kin, and village fissioning among the Ygnomamo Indians," in Evolutionary

enon before analysing the non-cultural factors that may

be involved in actual wars.

Chagnon's biological determinism is incapable of ac-

counting for the multidimensional complexity (and re-

biology and human social behavior: An anthropological per-

spective. Edited by N. A. Chagnon and W. Irons, pp. 86-132.

North Scituate: Duxbury Press.

-. 1983 (1968). 3d edition. YpnomnmB: The fiercepeople. New York: Holt, Rinehart and Winston.

-. 1988. Life histories, blood revenge, and warfare in a tribal

population. Science 239:985-92.

gional diversity) of Yanomami warfare in itself or in a

comparative perspective. Reducing Yanomami intervil-

-. 1989a. [Letter.] Anthropology Newsletter 30(1):24.

-. 1989b. Response to Ferguson. American Ethnologist

lage raiding (like all "primitive warfare") to a biogenetic process-the pursuit of inclusive fitness-is simply ex- plaining it away with an all-purpose hypothesis that dis- solves its cultural content into an alleged final genetic imperative. Eliminating the fundamental issue of hu- man cultural variability and reproduction, this reduc- tionism annuls the very object of anthropology (see Sahlins 198oa:chap. I; 1g80b:gg-120) and contributes little to the study of the complex specificities and in- teractions of biological and cultural processes (see Levi- Strauss 1981, ~oran~e

1986, Paul 1987, Rogers 1988).

References Cited


B. 1985 Temps du sang, temps des cendres: ReprCsenta- tion de la maladie, systeme rituel et espace politique chez les Yanomami du sud-est (Amazonie bresilienne). Thkse de Doc- torat, UniversitC de Paris X, Nanterre, France.

-. 1988. La furnee du metal: Histoire et representations du contact chez les Yanomami (Bresil). L'Homme 106-107:87


14. This conception is not incompatible with a materialist ontol- ogy of cultural representations: see Sperber (1987).


-. 1990. "Reproductive and somatic conflicts of interest in the genesis of violence and warfare among tribesmen," in The anthropology of war. Edited by J. Haas, pp. 77-104 Cambridge: Cambridge University Press.

CHAGNON, N. A,, M. V. FLINN. AND T. F. MELANCON. 1979. "Sex-ratio variation among the Ygnomamo Indians," in Evolutionary biology and human social behavior: An anthropolog- ical perspective. Edited by N. A. Chagnon and W. Irons, pp. 290-320. North Scituate: Duxbury Press.


J-C. 1981. Histoire de la violence en Occident de 1800 a nos jours. Paris: Editions Robert Laffont. COLCHESTER,

M. n.d. The Yanoama at war: Subsistence, sover-

eignty, and social change in an Amazonian society. MS. DOMENACH,M.1980. "La violence," in La violence et ses


causes, pp. 31-42. Paris: UNESCO. FERGUSON,

B. 1989a. A savage encounter: Western contact and the Yanomama war complex. Paper presented at the School of American Research seminar "Expanding States and Indigenous Warfare," April 24-28.

-. 1989b. Do Ygnomamo killers have more kids? American Ethnologist 16:564-65. HEADLAND,

T. N., AND L. A. REID. 1989. Hunter-gatherers and their neighbors from prehistory to the present. CURRENT ANTHROPOLOGY 30143-5 I.


T. 1988 (165 I) Leviathan. London: Penguin Books. HOUSEMAN,

M., AND C. SEVERI. 1986. "Lecture de Bateson an- thropologue," preface to La cerbmonie du Naven, by G.Bate

Volume 31, Number 5, December 1990 1 563

son. Paris: Le Livre de Poche (Biblio Essais). (First published in 1936.1 HOWELL, N. 1979. Demography of the Dobe !Kung. New York: Academic Press. HUMBOLDT, A. VON. 1819. Voyages aux regions Bquinoxiales du Nouveau Continent faits en 1799, 1800, 1801, 1802, 1803 et 1804. Vol. 2. Paris: N. Maze. JAMIN, J., AND G. LENCLUD. 1984. Une ethnographie de la vio- lence est-elle possible? Etudes Rurales 95-96:9-21. KNAUFT, B. M. 1987. Reconsidering violence in simple human so- cieties: Homicide among the Gebusi of New Guinea. CURRENT ANTHROPOLOGY 28:457-82. KNOB LOCH, F. 1975. Lungo il Cauaboris (Storia di una missione). Salesianum 37(1):131-58.


1979. Patterns of health and disease among the Waorani Indians of eastern Ecuador. Medical Anthropology 3:147-89. LEVI-STRAUSS, C. 1981. Culture et nature: La condition humaine a la lumikre de l'anthropologie. Commentaires I 5 :365-72.

LIZOT, J. 1988. "Los YanBmami," in Los aborigenes de Venezuela, vol. 3, Etnologia contemporanea. Edited by J. Lizot, pp. 479-583. Caracas: Fundacion La SalleIMonte Avila Edi- tores.

-. 1989. A propos de la guerre: Une reponse a N. Chagnon.
Iournal de la SociBte des Americanistes 75 :91-I I 3.

LOPES DE ARAUJO, F. X. 1884. Relator10 da ComissLio Brasileira Demarcadora de Limites Brasil-Venezuela (1879-1884). Rio de Janeiro: Itamaraty.

MC CAULEY, C. 1990. "Conference overview," in The anthropol- ogy of war. Edited by J. Haas, pp. 1-25. Cambridge; Cambridge University Press.

MELANCON, T. F. 1982. Marriage and reproduction among the Ygnomamo Indians of Venezuela. Ph.D. diss., The Pennsylvania State University, University Park, Pa.

MEN GET, P. Editor. 1985-86. Guerre, sociCtC et vision du monde
dans les basses terres dlArnCrique du Sud. Iournal de la Sociktd
des Americanistes 7 I: 129-208; 72: I 33-220.

MICHAUD, Y. 1988. La violence. Paris: Presses Universitaires de

MICHELENA Y ROJAS, F. 1867. Exploracion oficial. . . desde el
norte de la America del Sur . . . en 10s afios 1855 hasta 1859.
Bruxelles: A. Lacroix, Verboeckhoven.

MI GLIAZZA, E. C. 1972. Yanomama grammar and intelligibility. Ph.d. diss., Indiana University, Bloomington, Ind. MO RAN GE, M. 1986. Biologie moleculaire et anthropologie. L'Homme 97-98:125-36.

OTTERBEIN, K. F. 1973. "The anthropology of war," in Handbook of social and cultural anthropology. Edited by J. Honigman, pp. 923-58. Chicago: Rand McNally College Publishing.

PAUL, R. A. 1987. The individual and society in biological and cultural anthropology. Cultural Anthropology 2:80-93. PETERS, J. F. 1987. Yanomama mate selection and marriage. lour- nal of Comparative Family Studies I8:79-98.

POSEY, D. A. 1987. Contact before contact: Typology of post-Co- lumbian interaction with northern Kayapo of the Amazon Ba- sin. Boletim do Museu Paraense Emilio Goeldi, Antropologia,

3:135-54. ROBARCHEK, C. A. 1989. Primitive warfare and the ratomorphic image of mankind. American Anthropologist g1:go3-20. ROGERS, A. R. 1988. Does biology constrain culture? American Anthropologist 90:819-3 I. ROSS, J. B. 1988. A balance of deaths: Revenge feuding among the Achuara Jivaro of the northwest Peruvian Amazon. Ph.D. diss., Columbia University, New York, N.Y. SAHLINS, M. 1g8oa. Critique de la sociobiologie: Aspects an- thropologiques. Paris: Gallimard. -. 1g8ob. Au coeur des socibtes: Raison utilitaire et raison culturelle. Paris: Gallimard. SCHOMBURGK, R. H. 1841. Journey from Esmeralda, on the

Orinoco, to San Carlos and Moura on the Rio Negro . . . in the spring 1839. Iournal of the Royal Geographical Society 10~248-67.

SMOLE, W. J. 1976 The Yanoama Indians: A cultural geography. Austin: University of Texas Press. SPERBER, D. 1987. Les sciences cognitives, les sciences sociales et le materialisme. Le Debat 47:103-15.

SPRUCE, R. 1908. Notes of a botanist on the Amazon and Andes. 2 vols. Edited and condensed by A. R. Wallace. London: Mac- millan.

STRADELLI, E. 1889. Dal Cucuhy a Manaos. Bolletino della Societa Geografica Italians, series 3, 2(23[26]):6-26. WEISS, K. 1976. Demographic theory and anthropological infer- ence. Annual Review of Anthropology 5: 35 1-81. WERNER, D. 1983. Fertility and pacification among the Mekranoti of Central Brazil. Human Ecology I I 227-45.

launching: of anv raid involves an endocannibalistic fu-

tion. Yearbook of Physical Anthropology I 9:95-110.

nerary rGe for victim to be avenged and the ritual

-. 1976. Fission in an Amazonian tribe. The Sciences

On Structure and Entropy: Theoretical Pastiche and the Contradictions of "Structuralism"


Department of Anthropology, University of Chicago,
Chicago, 111, 60637, U.S.A. 9 VIII 90


as a form of anthropological thought preeminently associated with the work of LCvi-Strauss, is generally conceived to be founded upon a concept of li~tru~t~rell

imported in more or less equal parts from linguistics and mathematics. The linguistic ideas from which LCvi-Strauss drew inspiration are by now rela- tively familiar. The mathematical side of the story, how- ever, has been much less well understood. This lack has now been remedied by Almeida's (CA 31:366-77) masterful exposition of the mathematical ideas upon which LCvi-Strauss appears, explicitly or implicitly, to have drawn: group theory with its concepts of group, transfor- mation, and invariance and notions of topology, proba- bility, and entropy derived from other forms of mathe- matics.

The concept of structure in terms of invariant con- straints governing transformations is of course not con- fined to group theory or to mathematics; it is fundamen- tal to any form of structural analysis. Levi-Strauss was not the first to apply it to the analysis of human phe- nomena: Marx's analysis of the transformation of values into prices of production in volume 3 of Capital has exactly this form, and many other examples could be cited, among the more notable from the works of LCvi- Strauss's contemporaries Bateson and Piaget. LCviStrauss's idiosyncratic manner of applying this concept to anthropological data, however, sharply differentiates him from these thinkers and others who have applied it to social, cultural, and psychological data. This idiosyn- cratic approach is the result of his attempt to synthesize the concept of the transformation group with the basic ideas of Saussurean linguistics and Prague phonology.

Almeida's article will be of great value to those wish-

  • Recommend Us