,

NAPOLEON A. CHAGNON
useful , an actthat requiresthe "cnuorsement" ofco- villagerswho will
go along \Vitl: the manipulation by altering their own kin usage to
conformto that initiateu by theoriginal manipulator(Chagnon 1988a).
Athird way isfor particul armen to lobbyfor a village fis sion that will
divide the largcr group into smaller ones permitting Ego to surrounJ
himselfwith a mixture ofco-resiuentkin more congenial to his social
and reproducti ve intcrests (Chagnon 1981; 1982). One's ability to
inAuence others, make uemanus, coerce, garner cooperation, etc. is
often adirect function ofthe individual' skinship nexusand the kinds
andnumbersofkin-definedalliesheorshecandrawonto enforcehi sor
herwill. ConAi cts ofinterestemergeand uevelop in akinshipmatrixin
most band and village societies, necessitating an understanding of
genealogical related ness, reproducti ve anu marital hi stories, and othcr
features ofkinship and descent. In addition, high status and esteem
usuall y confer advantage in matrimonial striving and, therefore, in
reprouuctive success. lt thus should be expected that individual swil!
compete over and have conAicts about rclative degrees of esteem,
conAicts that may, on the surface, reveal no obvious rclationship to
either somatic or reproductive resources. Measurements of relative
status and relative uegrees ofreproductive success should be made to
determine if thcrc is a positive conclation between them.
An effective synthesis of the nvo bodi es of theory is only no\V
uevcloping. What appea rs to be clear, however, is that the emerging
synthesisismorecapableofincluuingawidervarietyand largernumber
ofspecificvariablesthatcan beshown to be implicatedin thegenesis of
conAi ct, violence and warfare. lt ca n al so proviue explanations for
them,as well asexplanationsfor theinstitutionalandbehavioral mcans
whereby humanscopewith themanu manage to livein acondition of
relative harmony.
Notes
l .Isubsequentl ylearnedthatthetwogroupstClllporarily,cttledthcir
differences in a club fi ghl.
2. Li zot(1989 1 givcs 1979 as thedate ofthis war in aeriti cism of111\
1988a publi cd ti on where thi s war is bri efl y menti oned in a
footnote. Iwill addrcssLizot's criticislll Sin a future publication.
104
5
Uneasypeace: intertribalrelations in
Brazil's Upper Xingu
TI-l o /\S GREGoR
TeJl the Americans about usoTell thelll we are not \\"ild
lndians who club people. Tcll thelll we are beautlful.
SHU M ION Ml': HI NA KU
1 n 1884 the German cxplorer Karl von uen Steinen descended the
KulescuRiver in Central Brazil to becomethefirst Europeantovisitthe
tTi hes ofthe UppcrXingu basin. Thepcopl es he di scovered spoke four
differentlanguages(Trumai, Tupi , Cariband Arawak),butlongstand-
ingpol ticalandsocial contacthad createdaremarkabl yhomogeneous
Xinguano culture that endures today. The ten Xingu villages remain
separatc and politic.: ally autonomouscommunitics, with astrong sensc
ofthcirown uniquenessandpositi vequalities. Althoughthe are
in sorne respects opposed to and suspiciousofthei rnci ghbors, theyare
intemclyand elaborateJyinvolved \Vith thcm through trade, intermar-
riage and intertribal ritual.
WhatisstrikingabauttheXinguanosisthattheyarepeaceful. During
theonehundred years overwhich \Ve ha verccords there is noevidence
of warfare amon"g the Xingu groups. To be sure there hd\e bccn
instantes ofwitcheraft killings trib;:d dlld rare uefensi\'e
ID,
THOMAS GREGOR
reactions to assaults from the war-Iike tri bes outside of the Xingu basin.
But there is no tradition of violence among the Xingu communities. In
fact , the value systems ofthese communities are "a ntiviolent" in nature.
Supernatural sa nctions inhabit the express ion of aggression, prcstige is
awarded to men who avoid conflict, and methods of socializing children
discourage displays of anger. The entire pattern of intense, peaceful
relationships between communities speaking different languages is rare
if not unique in nati ve South America and other culture areas of the
world. This paper examines the puzzle: Why are Xinguanos peaceful?
AN APPRECIATION OF THE XINCUANO
ACHI EVE ME NT: THE SCARCITY OF PEACE
Ifby war we mean organized violence between separate communities,
then humans are a warlike species. In attempting to compare the Xingu
data with other similar societies I have been frustrated by the mini mal
numbers of peaceful peoples. Other researchers, who have combed the
Iiterature more systematically than myself, have reached the same
conclusion. Thus Richard Sipes notes in hi s study of war and combative
sports: "Relatively peaceful societies are not easy to find . I had to
investigate 130 societies to find eleven, of which five \Vere rejected
because of insufficicnt information" (1973: 68). Similarly, Otterbein
(1970) found only four peaceful cultures among the fifty in his study of
the evolution of war. Turning to advanced, state-Icvel societies the
searcher for peace becomes even more di sheartened. Thus Arthur
vVesting in a study ofhigh fatality wars during the last eighty years, finds
that on the average three such wars were occurring simultaneously, and
that there was only one year in which none was being waged. He
concludes: "The sad, but seemi ngly inescapable conclusion that I dra\\'
is that war remains as a routine, typical , and thus, in faet, normal
human activity" (1982: 263).
In construeting a sample of peaceful cultures I have adapted thc
criteria of David Fabbro (1978), who has written one of the fcw
comparative studies of pcaceful societies. A peaceful society is one that
is not i nvolved in i nternal collective violence; one that exhibits relati\"cly
little interpersonal viol ence; one that provides no special roles for
warriors; and one that has values and sanctions precluding violence as a
means for resolving conflict. lt is di sheartening to realize that very few
societies meet these criteria completely. Even those that do so in a
106
lntertribal relations in Brazil's Upper Xingu
reasonable fashion , such as the KUllg Bushmen (the "harmless people" of
Elizabeth MarshaJl 's popular account) are themselves perpetrators and
victi ms of armed violence. 1 Peaceful societies are rare, and even when
they are found pea ce is seldom, if ever, absolute. As shall beco me clear
fu rther on, thi s same caveat applies to the peoples of the Upper Xingu.
The societies that come c10sest to fi tti ng the model of the trul y
pcaccful culture are small in scale and primarily hunters are foragers.
This conclusion is in keeping with research on war by Wright (1965:
68), Borch and Caltung (1966), Russell (1972), Eckhardt (1975: 56),
Wibcrg (1981: 114) and others who have positively associated war with
community size and cultural development . Peaceful peoples also tend
to be gcographi call y isolated. Otterbein (1970), for example, find s that
societies lacking in military organizations, such as the Copper Eskimo,
the Dorobo and the Tikopians, live on islands, mountain tops, arctic
wastel ands and plateaus surrounded by malaria infcsted jungles. In
some cases thi s isolati on is a strategic adaptation to dealing with more
aggressive societies that sllrround them. In most instances, however,
pcaccful societies appear to achieve their status by evading rather than
soh-ing the problems of intertribal relations. The tri bes of the Upper
Xingu are of special interest precisely beca use they do not conform to
thc profile of the typical peaceful society. Unlike most peaceful peoples
they are sedentary with fairly advanced economies based on slash and
burn IlOrticulture (a system that is often associated with warfare [Vayda
1961)) . The isolation characteristic of many peaceful societies is also not
typical of the Xinguanos. Indeed, the unique feature of the arca is the
intemity and the richness of intertribal relations.
PEA CE IN THE UPPER XINCU
The historical developmental of the Xingu s)'stem
How did the Xinguanos reach the compromises and make the social
changes needed to crcate thc culture of the region? At the moment, in
thc absence of systematic archeological data or an historically accurate
oral tradition in the Xingu, \Ve have onl)' cducated guesses. The first
derives from Calvo's (1953) bel ief that the Xingu basin has served as a
refugc for peoples mCllaced by more aggressive tri bes to the north. We
may speculate that the refugees , unwilling to stand their ground and do
battle, \Vere uniquel y arnenable to assimilation into a peaceful inter-
tri bal system. The geographically isolated Xingu basill may have thus
107
THOMAS GREG O R
filtered out warlike peoples. In opposition to this theory is Carneiro's
(1961 ) intriguing hypothesis that the Xingu tribes \Vere once warlike and
organized at the chiefdom leve!. Malaria , introduced in the sixteenth
century by the Europeans, depopulated the region, reduced competi-
tion for the now abundant land, and eventually produced the stable
intertribal system we see today.
Although the origins of the Xingu peace remain uncertain, we can
occasionally glimpse the past in today's social life. One of the major
puzzles of the Xingu system, for example, is its tendency to assimilatc
other societies. The Xingu region is an area of "native acculturation,"
whi ch has expanded to peacefully inelude tribes from four different
language groups ~ t o a remarkably homogeneous Xingu culture. The
Txicao, a Carib speaking tri be, is in the initial stages of joining the
Xingu system and provides llS with elues to how the process occurs. The
Txicao village, formerly a single house in a clearing, is currently
designed like the Xinguano communities. The Txicao cut their hair like
Xinguanos, wear arm bands and earrings in the Xingu style, and carry
on Xingu rituals. These changes have largeJ y occurred within the last
few years, and are all the more remarkable since the Txicao initiated
contact with the Xinguanos by raiding their villages.
The events that led the Txicao to participate in Xingu culture began
with a series of retaliatory raids by the Xinguanos against the Txicao.
These raids, combined with epidemic disease, left the Txicao in such a
desperate state that it became in their interest to establish peaceful
relations with the Xinguanos. This process was facilitated by thc
tendency of all the Xingu villages to expel men accused of witchcraft.
Two accused witches, using kinship ties to the Txicao established
through previously kidnapped Xinguano children, are now living with
Txicao wives in the Txicao community. They are agents of"Xinguifica-
tion" and are systematically teaching their rituals, music and
mythology. The Txicao do not yet clearly perceive that their OWll
culture is in jeopardy. Rather they view what they are learning as a
beautiful and interesting addition to their traditional culture. But it is
also clear that if the process continues the Txi cao will be full members of
the Xingu system within a relatively short period of time. The Xingu
system thus has a tendency to expand and bring unrelated groups into
its orbit.
108
Intertribal relations in Brazil's Upper Xingu
The structure of the Xingu peace and the system today
Peace in th e Upper Xingu is a bimodal pracess which organizes tri bes
and individuals in enduring relationships, but at the same time sets
lhem apart and keeps them at a social distance. Fundamental to this
process are values that define the identity of individuals and tri bes while
sctti ng expectations for participating in the system. M)' account of these
val ues and institutions is primarily based on field work among the
Mehinaku, but it is supported by research in most of the other
communiti es, and by recent (1985) interviews with informants fram all
of the other language groups on the specific subject of war and peace.
In ancient times, according to one ofthe Xingu origin myths, the Sun
createel three races ofhumankind, including Xinguanos, "wild Indians"
and whites. Recogni zing the warlike nature of whites and wild Indians,
he assigncd them to separate worlds and even separate afterlives, well
away fram the headwaters of the Xingu River and the "Village in the
Sky" aboye it . Idea]]y, according to the villagers, things should have
remai ned thi s way, since the Xingu system was self-contained and
autonomous. AII the villagers' needs, whether for trade goods, spouses,
or ri tual participants, were met within the system. But the wild Indians
and the white man came to look for the Xinguanos, out of a desire for
plull der, trade, and sex with Xingu women. As a result, the system is
more open than it has ever been before, with regular contact with both
Brazi lians and llon-Xingu Indians. But these relationships are a source
of profound ambivalence to the villagers, who look forward to a day
(which may actually come fairly soon) when they can once again seal off
thcir reservation from contact with the outside world .
Scen from within, the Xingu \Vorld is divided into ten single-village
tribcs, each of which is intensel y engaged with its fellows. The total
popllJa tion is appraximately 1200 persons, most of whom can per-
wnally recognizc each other at Ieast to the extent of identifying tribal
and kinship affiliations. To a degree, social interaction is structured by
mcmbership in the four maJor language groups, with the most freqllent
contact occurr!ng between members of the same group. Each language
group is regarded as having a unique culture, and even a unique
h)'sical appearance. From the perspective of the Arawakan tri bes, for
cxample, all of the Carib tribes are described by one term
(uYanapukwa ") and all are said to be alikc in their languagc, thcir
appearance, their pCllchant for rudcncss, and cven thcir prcfcrence for
109
THOMAS GREGOR
barelycooked fish. Similarstereotypeslabel theotherlinguisticgroups,
and are reciprocated by them. At heart , however, the values of this
peaceful intertribal system are cosmopolitan in charaeter. Thusone of
theMehinaku villagers explained thatto him theCariblanguageonce
resembledthebarkingofdogsand theslobberingofpigs. "Butthenlate
theirfood, had sex with theirwomen, and learned theirwords.2 Now I
think their language is beautiful."
The values and institutions o{ peace
In June of1985, Itookthe Mehinakuchief, who was thenvisitingRio
de Janeiro, on a tour of the city zoo. Each time we examined an
unfamiliaranimal, such as agiraffe oran elephant, he posed the same
guestion: is it peaceful ("awujitsi") or is it ferocious ("japujaitsi")?
IntheXingu, thisis theoverridingmoralguestionaboutindividuals,
relationships, andcultures. Aboveall, theconceptofwhatisgood is tied
to peacefulness . The good man is circumscribed in hisbehavior, he
avoids confrontations, and he rarely shows anger. In Mehinaku, his
peaceful comportment is "ketepepei," a word that is also applied to a
well-made craft objeet, suggesting a beauty and balance.
1
The good
citizen is peaceful because he responds to the feelings ofothers. He
refrainsfrom injuringthem becausehewould"feelsad"andsensetheir
painasifitwerehisown. Theviolentmanlackstherequi sitesensitivity,
and kills and maims others, often without motive. He beats and
tonncnts his children, and "itis for this reason thathi schildrendo not
senseothers'feelings whentheyareadults."Thedifferencebet:weenthe
peaceful Xinguanos and the warlike Indians and whites is lack of
empathy.
Ifpressed, informants \ViII offer additional reasonswhy peace makes
good policy: "The peaceful man istreated peacefully by others; after
death, onl ythe peaceful man lives in the lparadisiaeal] Village in the
Sky."Butin faet, peaceas aphilosophyis notsubstantiallydevelopedas
a positively defined concepto The value ofpeacefulness in the Xingu
appearsto be a"terminalvalue," one that is at the apex ofapyramidal
structureofinstrumental and speeific rules ofconducto As is common
with sueh fundamental values, its rightness is taken for granted.
Question avillageraboutthe reasonsfor peaeeful interaetion, and he is
more than likely to leaveaside an abstractdiseussion ofvalucsand tum
to the relationships that conneet the Xingu tribes.
110
lntertribal relations in Brazil's Upper Xingu
lntertribal relations: trade
Theeadiestobserversto the UpperXingunoticed thateaehofthetribes
aftheregianmadegoods thattheyexportedto theothers:shellbeltsand
neckl aees, stone axes, salt, cotton, fish spears, hard-wood bows and
eeramie pots. Sorne ofthe tribes maintained effeetive monopolies in
thescitcms, sothatgoodtradingrelationshipsandpeacefulrelationswas
a necessary part ofsubsistence. Today, the monopolies persist, even
though thesystem has been significantly eroded by the introduction of
steelaxesand aluminum pots.
Sorneofthe specializationsare ecologicallybased. Thehigh guality
clay used by the Waura for theirceramie pots is available primarily on
lhei rtraditional territory. Theshells used for necklaces and beltsby the
Cari bspeaking tribes are mainly aceessible in areas near theirvillages.
Othcr trade speeializations are grounded in speeial knowledge or in
substantial investment oflabor. The Waura ceramic monopoly is the
bestcxampleoftheformer, and theMehinakusaltworks, with its large,
laborintcnsive plantings ofwater hyacinth
4
is an instance ofthe later
type afmonopoly.
Whatis particularlyintriguingaboutthespeeializations is thatnane
of thcm is fully defensible as a monopoly. Knowledge of ceramics
dlffusesacross tribal boundariesas aresultofintermarriageandvisiting.
Cla)'andshellscouldbeobtainedoutsideofCariband Wauraterritory.
Thelaborneeded to makesaltorotherproductsis potentiallyabundant
inalloftheXinguvillages. Thepuzzleofthemonopoliesdeepenswhen
we consider that the trade goods are highly valued by the Xinguanos .
Shell bdts, necklaces and eeramic pots are the measure ofa man's
wcalth. 'fhey are used to pay shamans in medieal treatments and are
public offeri ngs for important rituals. Why not make these things at
home?
When this guestion is put to villagers from different tribes, they
rcspondwithalistofsa nclionsthatprotectthemonopolies.Themendo
no! dare to get shells and make necklaees because they are fearfulof
Caribwitches. Thewomenwouldnotmakepotsbecausetheywould be
ridiculedby Waurawomen whoare mastersofthecraft. No onewould
makesaltfor exportbeeausethat\Vould angertheMehinaku. Thusput,
lhetradesystemseemstodependon institutionswhichseparatevillagers
ratherthan thosewhich bringthem together. But in fact , the villagers
alsovalU: thepositiveimpaetofi ntertribalrelations. Trademeanstrust,
I II
THOMAS GREGOR
since the items offered may not be reciprocated for several months or
more. Trade means mutual appreciation, since craft objects, unlike Our
manufactures, are an extension of the self which the maker hopes will be
admired. Trade is a social relationship that is valued in and or itself. and
is a conscious reason for maintaining the monopoli es. As one of my
informants explained to me: "They have things that are really beautiful,
and we have things that th ey like. And so we trade and that is good."
Intennarriage
AII of my informants agreed that no one likes to live in villages other
than his own. When, due to the extension of the incest taboo and the
small size of the Xingu communities, a man can not find a suitable mate
in his own communit)', he will marry into another vilJage. Approx-
imately thirty-five per cent ofXi ngu marriages are of this type, and given
the reception of a typical in-married spouse it is a wonder that the
percentage is that high. The first problem such a spouse faces is that of
the language. The Xinguanos are proud of their languages which are the
major maker oftribal distinctiveness. They are intensely ashamed when
they try to speak a language they do not know. Consequentl)' the
villagers develop a passive understanding of the language of their
spouses' community long before they speak it themselves . "You should
not," explained one of my Mehinaku informants, "speak the Janguagc
of your father-in-law."
An equally significant barrier to intertribal marriage is the pattern of
hazing the in-married spouse. The villagers maintain that such a person
"robs" them of potential mates, and properly he must paya price by
being made to suffer. Presently, in the Mehinaku, one young man is
being tormented by his wife's lovers. His hammock has been filled with
ashes, his canoe hidden and submerged deep in the swamp outside the
village, and his garden has been vandalized. The level of ridieule and
practical joking to which he is subject has reached such a pitch that he
avoids the men's house and the center of the community: "1 don't dare
go there at all."
Finally, the in-marri ed spouse must overfulfill his obligations to his
in-Iaws. Onerous enough at home, in the host community these cluties
can amount to servitude. As one young man explained to me as 1
interviewed him in his father-in-Iaw's house: "Each day 1 fish for ITl\
fath er-in-Iaw, work in his garden, or build him a canoe. Today's \\'rk is
112
Jntertribal relations in Brazil's Upper Xingu
done, but 1 can not lie in my hammock and seem to be lazy. 1 have to sit
on th is bench even though m)' back and shoulders ache."
The shame regarding language and the pattern ofhazing is evidenee
of the centrifugal, oppositional nature of the Xingu villages. Intertribal
ma rriage transcends the social distance by extending kinship across
triballines and eventually producing bilinguaJ children who can move
faidy freely from one community to the other. These individuals are
cru ciallinks bctween the Xingu villages in establishing new marriages ,
in extending bonds of fictive kinship between the tribes, and in
fa cil itati ng visits, rituals and trade. The res ult is that loyalties are di vided
a/ong lines of kinship and residence and the barriers between the tri bes
are semi-permeable. In a conflict between perso ns from different
communiti es kinship overrides eommon residence. The quarreJ is
simply not perceived as a community affair, and the disputants count
only on the active support of their kinsmen.
he centripetal potential of intertribal marriage for producing one
Common Xingu soeiety is most nearly realized in the Yawalapiti village,
which no\\! has more speakers of Carib and Tupi than Yawalapiti For
some of the villagers in other communities this is an unfortunate
arra ngement , whieh they compare to a noisy flock of birds of different
spee es, none of whom can speak eaeh other's language. But for all of
the Xi nguanos, intertribal marriage is a major souree of peaceful contact
with each of the Xingu villages.
The role of the chief
"One \\'o rd from the chief and everyone does whatever they want" was
}oh n Cooper's summation of the ehiefly role in soeieties of the South
American lowlands. Among the Xingu soci eties th e chid also has
mnimal authority I.see Dole 1966), and yet he has a substantial
expressive role in rituaL" in making public speeches (Gregor 1977: 80-
3), and in organizing trade sessions (Dole 1956-58). The most visible
enactment of his special position Occurs during the major intertribal
fcstivals in the fall of each year. The foeus of the rituals is the inaugural
of ne\V chiefs and the commemoration of ehids who have di ed. The
rituals require the participation of all the tribes, and are perceived by the
"il1agers as an explanati on of the peacefuJ nature of the Xingu system:
"We don't make war; \ve have festivals for the chiefs to whi ch all of th e
viJ]ages come. \Ve sing, dance, traeJ e and wrestle."
113
THOMAS GR EGOR
Within the communities, the chief embodies the values of
generosity, hard work, lawfulness, and peacdulness. Ideall y, he never
di splays anger, never engages in gossip, never makes witchcraft accusa-
tions, and never participates in witch killings, no matter how severe the
provocation. Psychologicall y, among the Mehinaku, he approximates
the idealized father in that he represents prosocial values and the
repression of anti social impulses. The villagers explain that the chid
"takes care of," "provides for ," and "gives food to hi s people," who are
often referred to as "his children." Like thdather of a famil y, the chief is
said to be responsible for the moral tone ofhi s community. He lectures
"his children" in the evening and at dawn , correcting their behavior
"the way a father ledures his son." According to the villagers, com-
munities that split into rival factions (as has recently occurred to the
Carib speaking Kalapalo) or lose ancient traditons (as has happened
among th e Trumai ) havc not heeded th eir chiefs, or have not had a chief
worth heeding.
The role of the eh id in the Xingu is not entirel y limited to ceremony
and speeches. With the advent of the Brazilian administration the tribes
have moved much closer together, and increasingly interact in informal
situations unstructured by the traditional rules of affinal kinship and
ritual. On occasion, th is contact has led to fri ction between groups of
young men in different tri bes who chop down trees on the paths between
the villages and "borrow" other tribesmen' s canoes and bicycles. Thc
chiefs play an acti ve role in negotiating a quick settlement to disturb-
ances of thi s kind. More serious di sagreements, such as allegations of
theft and witchcraft , are beyond the scope of the chief's formal authority
and must be resolved by the disputants themselves.
What strikes the observer of the Xingu eh ief is that despite his role as a
negotiator, the status is elaborated far beyond what might be expected
from societies with an essentiall y unstrat ified political system. The
elevation of the chieftainship, particularl y its ceremonial componcnt,
appears to be a means of celebrating the peaceful intertribal systelll
rather than the chief' s local authority.
As expressed in institutions, the Xingu peace system is based on
mutuall )' rewarding ritual participation, th e exchange of trade goods
and spouses, and a conscious appreciation of the values of a peaccful
culture. But in fact this is onl )' half the picture. The Xingu peace rclics
heavily on institutions that separa te the tri bes and preoccupy the
villagers with thoughts of death and violence. If we question el yillagcr
114
lntertribal relatians in Brazil' s Upper Xingu
about the benefits of peace, he quickl y turns to the dangers of unusual
but soci all y important events, su eh as \vitch killings, battles with wild
Indians, wife-beal ings, rapes, physical puni shment of children, and the
sadi stic killing ancl torture of animals. In an important sense, the Xi ngu
peace is negatively defined, emerging with greatest clarity in the culture
of aggressioll and violence. T his culture stands as a foil against which
the Xinguanos forge a self identity, and aim their beha\ior towards the
ideal of peaee .
)JEGATIVE PEACE: BL OOD, WITCHES , AN D
W JLO Ji\ DI A N S [:\1 T H E X 1 N G U C U L T U R E O F
V IOLEi\ C E
Anger, aggressian, violence
Anger is a di sturbing and even frightening emotion for many of the
Xinguanos. In \l1chinaku, the term for angcr ("japujapai") is applied
mctaphori call y lo things that are compl etcly out of control, such as a
raging fi rc, a thu nderst orm, venomous wasps and a species of pepper so
hot that il lca\'es anyone foolis h enough to eat it writhi ng on the ground
in agony. The vill agers explain that they are not angry people, because
the)" avoid foods that produce aggression: "\Ve don't ea t animal s with
bloocl or hot foocl s. Our food is 'tasteless' and so our bellies are neVC 1l0t
with angcr." ; V/ hen the villagcrs do become angry, they can rclease
thci r fcel ings harmlcssly: "When we are all gry \VC wrestle; and then the
anger disappears. The wild IndiJ1lS don't know how to wrestlc and so
thcy cl ub pcople instead." Finall y, th(; vill agcrs claim t11':) are pt<lccful
beca use of thcir unique culture: "[ n ancient tim es the :)un ga ve us
cera1l1ic pots, hoes and fi shing arrows. He gayc the wild Indians clubs,
spcars and arrows for killing men."
B/oad
Thc \' iJlagers' attitude towards \ iolence finds its expression in d logicall y
developecl set of ideas that are rooted in a theory of h LI man biology, and
shared b\' at least the Ara wakan and T upian tribes. Above all, blood is
defi ling. Most garne animals are considercd incdible beca use they "ha\"('
too mueh blood." F'ish are served \\cll- cookcd so that thcir blood is not
\' isi ble. Persons \Vho are bleeding - from scri ous \\'ounds, ritual inci-
sions or menstrual ion - are rituall v impure. Killing is \Hong, in part ,
11 5
THOMA S G [( E C OH
beca use the killing produces \\founds and blood, and especlally a bloody
corpse, which is an object of fe;,u and revulsion. Such a corpse lS buried
m a shallow graH': outside the vlllage, and its soul wlll nevo ascend to
the Village in the Sky. The killcr is also contam inated. The blood enters
IlIS body where it enlargc:. the fJce and abdomen, darkens the skin and
produces a characteristic foul smell. Xinguanos \Vho kill witches use
special garments and medicines to distance themselves from the task,
and when it is completed, they attclllpt to remove the dcflling blood
from then bodies by consUlling medicines made from anatto. But, like
Lady \Jacbcth \\ho coulclllevcr cleanse her hands of Duncan's blood,
these "fi ends for killing humc.iT1s" cal1not \Vholly wipe away the moral
stain. Considered repulsive, they are deprecated as village leaders and
rejected as lovers.
The "wild" Indians
Each ofthe Xi ngu languages has a word for non-Xingu Indians. In eacn
language the term comeys emotions that rall ge from fear, to distaste and
contempt. The term, \vhich my informants often translated as "wild,"
condenses in a single powerful symhol what the Xinguano is not and
strives not to be. According to my Arawl ka n informants, the wild Indian
is ugly. His skin is black and he rubs it with rancid pig fat. His hair is slick
with grease. He \Vears lip and slobbers whcn he talks. He sleeps on
the ground like ,m animal. He sits on the pro\\' ofhis canoe and defecaks
into the water he dri nks. He nevcr bathes and has a foul smell. 6 He eats
rats, toads, pigs ancl sl1d kes.
Above all, the \Vild lndian lS violent: "ll e beats his children. Il c rapes
his \Vife. He shoots arro\Vs at the \vhite 111an's planes. He spl its peoplcs
heacl s wth clubs. Il e kidnaps children ancl burns \illages. He kills his
o\Vn kin. War for him is a festival." Why is the wild Indian so violent?
The answcr is traclition: "It is because their grandfathers werc violent."
My informants also cited legends in wh ich th e Sun presented tne \VilJ
Indlans with weapons and a violent culture, and othcr origin myths in
which some of the wild Indi ans \Vere the offspring of a bestial union
bctween a \VOlllan and a venomous snakc (sce Gregor 1C)S 5: 56-7).
\\fhen I pressed further, they woul c.l expla in the propensih for viol en ce
as derivi ng from their steady diet of bloady animals, and their inability to
wrcstlc like the Xinguanos.
Thesc answcrs are genuine explanations for thc \ill agers, hut
11 6
Intertribal relations in Brazil's Upper Xingu
ultimately, the wild Indian is violent because that is his nature: "That lS
the \Va\' it is with them. Wild Indians are not people." It is for this rcason
that outrages are described with an almost detached tone, as if they
are so beyond the human pale that one can expect no better of them. In
f3et, in terms of cultural symbolism, the wild Indian approaches the
status of an animal. Like an animal, he sleeps on the ground, defecates
in the water, fornicates in strange positions, crawls with vermin, and
reeks wi th foul odors. Like an animal, he attacks unpredictably, and
\\'l thout provocati on.
7
The Xingu villagers themselves ha ve been the victims of unprovoked
aggresslon from thc wild Indians, most recently the Suya and Txicao.
Thesc raids (in which Xinguanos were shot and kidnapped and com-
mll ni ties ra nsacked) precipitated violent defensive reactions. In one
su eh assall lt approximately thirty years ago, individuals from several
eommuni ties teamed up \vith a Brazilian to kili many Txicao men,
women, and children. In explaining this departure from the normal
pattern of peacefulness, my informants exphasized that the Txicao had
atlacked first and that they would attack again unless destroyed.
Moreover, the moral gulf between the wild Indians and the Xinguano
made it per missible. Nonetheless, the successful warriors took no
trophies and received no special honor upon their return. According to
some of my accounts, they had to consume medicines to rid themselves
of the enerny' s defili ng blood before they \Vcre accepted in their villages.
These dcfensive raids clearly show that the Xinguanos are fully capable
of orga ll ized armed violence. Such warfare, however, is regarded by thc
vill agcrs as a moral exception to thc normal pattern of peaceful
relations, and in fact seems to have been historically unusual.
Nowadays there is little daily contact between most of the Xinguanos
and the wild lndian. Nonethdess, he is in their thoughts, their myths,
and (among the \ J[ehinaku) in their dreams as an image of no\V not to
beha\'e. ' Keep still, " a Ill other \ViII say to her child as she atkmpts to
remove his heacl lice, "or you will crawl with verminlike a \Vild l11c.li;ll1."
\ggressi\ic indi vicluals find themselves compared to a lip-diskecl Kayapo
Indian wbo the find especially repulsive, and \vho has allcgcdly
ki llecl more than one hundrec1 lndians and whites. The wild Indian
thereby provi des a dram,ltic moral counterpoint for the ideal of peaceful
hehavi or. Contrary to appcarances, he plays a role in the Xingu rcace.
11-
THOMAS GREGOR
The witch
In ancient times, claim the Mehinaku, a malignant spirit felJ to the
ground. His body parts and adornment were apportioned among sorne
of the villagers who discovered that they could use them as fetish objects
to cause ilJncss and death. These first witches passed their skills on to
their sons by scarifying them with snake fangs and scorpion stingers.
Graduallya culture of witchcraft developed, which allegedl)' persists in
each of the Xingu vilJages. It is witchcraft, the personal malevolence of
individual villagers, that causes death, disease, crop blights, wind-
storms, and plagues of mosquitoes.
Why are sorne villagers thought to practice witchcraft? Ultimately,
the reasons are the same as those which motivate aggressive animals and
wild lndians: "They don't have feelings for others; they are not people."
But there is also an important difference between witches and wild
Indians. The wild l ndian, like an animal, has no social relationship
with his victims. The witch, on the other hand, is a kinsman , a fellow
villager, or another Xinguano. His motives for attacking others includc
revenge for the theft ofhis possessions, anger at sexual advances towards
his wife, the affront of being left out of a distribution of fish, or almost
any imagined slight. Nursing his grudge, the witch assembles a variety
of fetishes charged with magical power. These include tiny bows and
arrows, splinters of wood from ancient objects, potsherds believed to be
from mythical times, and the hair clippings, clothes or personal effulvia
of the victim. After the appropriate spells and magical acts, the victim
falls ill and soon dies.
Both the literature on the Upper Xingu and my own research
persuade me that witchcraft is seldom if ever actually practiced. The
fear, however, is palpable in all of the communities in which 1 have
worked. Among the Mehinaku, where 1 have lists of witchcraft accusa-
tions from more than half the adults in the villagc (Gregor 1977: 207-9),
every adult male is suspected by a least some members of the COI11-
munity. It is small \Vonder that at night children may be called in so that
they will not be shot with invisible arrows. The doors are locked, and
fearful householders sometimes construct crude alarms of tin pot Iids.
By day, the village gossip network is alive with rumors and accusations of
specific individuals ..\fter a death, highl y paid professional sorcercrs and
witch hunters (see Gregor 1977: 340-344) provide further support for
the system of belief by identifying the witch and exhibiting his fetishcs
118
Intertribal relations in Brazil's Upper Xingu
(which thcy thcmselves have secretly created for the occasion).
lfthe alleged victim of witchcraft is a young man, the accused may be
executed by the kin of the deceased. Such executions occur approx-
mately evcry two to three years in the Xingu (the last instance in 1984),
and are a major limitation on om classification of the Xinguanos as trul)'
peaceful. The kil li ngs themselves are brutal affairs, in which the victim
is taken by surprise and slaughtered with arrows, machetes and riAes.
Such executions are possible in an essentially antiviolent culture
becausc of the killer's absolute conviction in the justice of his action,
and the use of devices that separate him from his daily role as a peaceable
Xinguano. These include a variet)' of magical spells and objects said to
give him courage for the task, weaken his victim, and remove the
polluting effcct of the blood.
8
Thc political impact of witchcraft beliefs and witch killings on
Xinguano political institutions cannot be overemphasized. The most
significant is the veneer of courtesy that marks ordinary relationships.
Persons who despise one another seldom show anger. They are fearful of
their enemy's wi tchcraft, and equall)' frightened of being held respon-
sible as an alleged witch ifhe falls ill. Between tribes (especiall)' from the
perspective of Arawakan and Tupian groups), the pattern of false good
manners is even more marked. The mast dangerous witches are often
said to live in other tribes, so it is well to receive them with courtesy
when thev visi t. Those who are cautious avoid all informal (nonritual)
visits to other tribes, limiting intertribal relations to relativel), predict-
able and ci rcumscribed ceremonial occasions.
9
Perhaps the most interesting effect of witchcraft beliefs on the Xingu
political systcm is to limit the power of the chief. Despite the s)'mbolisl11
ofilie peace-Ioving chief, his role is a threatto the balanced nature ofthe
Xingu intertri bal system. As things stand, he can onl)' command the
loyalty of his dependent ki n. A dispute between him and members of
another tribe is seen as a disagreement between the parties involved, not
as an intert ribal dispute. Were he able to command the loyal tv of an
entire community, an interpersonal quarrel between two loca1 chiefs
could evolve into a full Aedged war. But even a powerful Xingu chief
knows he must tread softly in excrcising his limited authority.
The chief is the point man in the political system, and as such he is
inevitabl y the target of jealous accusa tions by the communit)' at large. In
the midst of his noncontroversial, self-depreciati ng public speeches, he
Imagines hi s rivals' resentment. When he has actual control of individu-
119
T I-l OMAS GRE GOR
als and resources (as is becoming increasingly true of two of the chiefs
who are linked to the Indian agency administration) he li ves in chronic
anxiety: "Let us say that 1di stribute gifts. 1keep nothing for myself or my
family, but thEre is still not enough to go around. 1explain to those who
received nothing that there is no more Idt. They say 'fine, that is alJ
right, \Ve understand.' They say the right things, they speak beautifully.
But at night they sort th eir fetishes and plan to murder me and my
family " Thus he imagined witch curbs the power of the chid, retains
the essentially egalitarian basis of local organizabon, and preserves the
symmetrical and balanced relationships among the different tribes.
But at what a price
l
\Vithin the villages the apparent good fellowship
of everyday life masks murderous intent. Today in the Xingu villagcs
there are individuals who are marked for execution. During my last field
trip in December of 1985 two of them were too frightened to leave their
own houses. Others moved erratically from village to village in the vaio
hope that they would be accepted elsewhere . . \n even larger group nO\\l
live off the reservation among the wild Indians and on backwoods
Brazilian farms and ranches - completely cut off from their kinsmen
and their traditional culture.
The cost of witchcraft beli efs can go beyond openl y accused witchES
to include virtually everyone, no matter how well established as good
citizens, no matter how well protected by large numbers of male kin.
Such is the sys telll among the Mehinaku, and 1 believe in the other
villages as well, that everyone will from time to time be gossiped about
and come under suspicion. As one man explained: "1 am not a witch.
But 1 am frightened that people think 1 amo Have you noticed that
children will not comE close to me? It is beca use their parents have said
that 1 am a witch. "
Witchcraft beliefs have the potential of dragging the Xinguanos into
an abyss of accusatiollS, killings, and ultimately a Hobbesian war of
"c\cryone against everyone. " But even though the villages smolder with
anger, only rarel y does the violence degenerate into a \'engdul feud o 1
have documented a few instanCES in o\'er fifty years of Xi ngu witch
killings where the conflict continued to spiral after the initial executiol1 ,
but in no case \Vas there more than one Jdditional death. In general.
widEr conflict is restrained by the careful choice of the victim. Though
apparentl y scl ected magicall y, he is killed onl y \\fhen he and his kin are
substanti ally outnumbered by the killers . If he is too strong, they nwy
employa sorcerer to kili hi m magicall y, or they \Vill Silllply seek re\'cngc
120
Intertribal relations in Brazil's Upper Xingu
through malicious gossip.
The villagers are conscious that the systelll of revenge is atavistic and
violates the essential Xingu value of peacefulness . One of the most
powerfuI indi vi duals in the Xingu cOlllmunities, aman whose own
Eather was killed as a witch, is presently conducting a personal crusade
against the killi ngs . He belittles false shamans, and uses the gossip
network to discredit their accusations. "1 would not kili a witch even if
he ki lled my sons," he explained. ''There is no way to know with
certainty who is guilty. Most of those who have been killed were
neecJ]cssly kill ed. TIl eir deaths were ugly affairs. " Conscious awareness
oE how the system functions thereby limits the vi olence in the Xingu,
and offers sorne hope for an evoluti on towards a more rationally based
politicaJ system.
WHY ARE XINGUANO S PEACEFUL?
War makes rattling good history; but peace is poor reading.
TIIO MA S HAROY
Peace theory
Comparati ve research on the cause of war and peace is based on the
hiddcn premises that peace is an expectable state of affairs in human
rel ationships vv'ere it not for conflict. Peace is the absence of conAict,
and it is confl ict that needs to be expl ained (cf. Ilaa: ,. this volumc) . My
own perspective is the reverse. Political systems are so volatile and war is
so contagio us that its existence should occasion little surprise. lt is peace
that needs special explanation. But it is here that social science lets us
clown, sin ce peace is seldom studied much less explained. A case in
point is lhe oumal of Peace Research , the foremost interdisciplinary
oumal devoted to peace studies. Despite its titl e, the oumal is mainly
dcdi cated to tll e studv of conAict, conAict resolution and the causes of
war. Thus the editor sadly notes: "It huns out that of the approximately
400 articles [publisned in the ouma!]over seventeen years, a single one
has been devotec1 to the empirical study of peaceful societies with a \i e\\,
to find out what seemed to make thcm peaceful " (Wiberg 1981: 11 3l.
With fcw ethnographic examples to studv. peace theory is a highl y
spcculati vc and essentiall v intuiti ve sector of social science rescarch .
Thc simplcst of the theo'ries are those of "negati vc pea ce" (Galtung
121
THO MAS GR EGO R
1968: 487) . Negative peace in a pure form is based on minima}
relationships: "Good fencesmakegood neighbors."Thecl ass icform of
negati ve peaceis deterrence, aconsciouspoliticalpolicythatgoes back
at leastasfar as ancientGreeceand remains an actin:partofmilitar)'
doctrine. Theeffi cacyofdeterrence in nonWesternsocieti es isattested
to by asubstantial number ofanthropological studies, suchasE\ans-
Pritchard's (1940) examination of t:e "ordered anarchy" of statelcss
societi es.
Positivepeacedepcndsontheexchangeofgoods,scnicesandpeoplcs.
Oneoftheeffects ofexchangeistocreateloyalties whichare di videdby
both territoryand bonds ofinterest, such as kinship and economi cs.
These competingallegi anccsattracta natural constituencyin favorof
maintaini ngpeacefuJ relations(Col son 1953, Gluckmal1 1955,Murphy
1957) MoreO\"Cr, exchangc leads tothccreationofacommonculture.
Parall el institutions in different societics can generate a consensus of
val ucsandstimulatethekindofdiffuseemotionallymeaningful relation-
shipsthat would inhibit violence (Caltung 1968: 491).
Although a]] of these theori es are intuitively reasonable, none of
them survives the test of the cross-cul tural data. Trade and culture
homology, for example,areactuaJJ ypositi velyassociatedwith war(Teft
197 5). In termarriage, whi ch should lead to the most solid kind of
alli ancesbasedonkinshipalsofail sthetestofcross-culturalcompari son:
intermarriage, kinship and warareactuall ypositivel ycorrel ated (Tcft
1975:70 1).Justas intcrpersonalviolenceoften nccursin cl ose relation-
ships, themostintense conflictsseemto occurbetweenpolities thatare
similar in structure ano intenselyengaged with one anot] cr.
Peace theory is at an embryoni c stage of development, but thc
apparent failure of the basic theorems of deterrence, exchange and
cultural homology is surpri sing and counter- intui tive. A partial
explanabon of the fail urc may be that the theorems are to a degree
contradl ctory. Thosewhochampion cultural homolog\and exchangc
maintain that closeness , positive affect anddiffuseattachmentare the
keysto peace. Promotersofthe instituti ons ofnegati ve peace, on the
otherhand, see peaceemcrgingfrom di stance, fear,and evcnhostility.
There is a possibility that both arepartl vcorrecto IntimateassociatioI1
andsimilarvalucsmean thattherearecommon])'yaluedobj ects tofight
over, and that the fights will be intcrneci ne and vici DUS . Cultural
distancealld themutualfear,however,createl highl yvoh:lti lesystemin
whi ch a minor breach ofthe peaee lcadstoa major conflict. We can
122
lntertribal relations in Brazil's Upper Xingu
conceive ofa system - and 1beli eve that the Xinguanos are arare
example- which is balancedat astabl emidpointbetweentheinsbtu-
ti ons ofpositi ve and negative peace . At this point there ismeaningful
homology and exchange, and yet at the same time members of the
systemwill besufficientl yconcernedaboutthe dangersofaggressionto
avoid anybreach ofthe peace.
WhythendotheXinguanosli veatpeace?Theydosobecauseoftheir
uniquehistory in an ecological abundant and isolated region ofSouth
America . They maintain that peace because of the insti tuti ons and
valueswhich bringthem togetherin meaningful relati onships, includ-
ingtrade, intertribalmarriageandtheinstitutionofthechieftainship. In
an ideal world, this would be enough. But political life inevitably
generates resentment, antagoni sm and fear. Among the Mehinaku
these tensions have been built into institutions that mal' actuall y
preserve the peace. Thus lingui sti c ethnocentri sm and hostile
stereotypes keep the Xinguanos athomeand limitinteracti on between
thc tri bes.The compelling images ofviolence and blood are constant
remindersto thevillagersofwhattheymustavoid. Theaggressive "wild
lndians" stand as symbols of \Vho they must not be. The beliefs in
witchcraft are a powerful sa ncti on that enforces courtes)' among the
villagers and limitsthe rol eofthechief. Theirony oftheuneas)' Xingu
peace is that the institutionswhi ch curb conAict areal so those which
painfulll' cxpress fear and anger.
Notes
Thispaperisbasedonficl dresea rchamongthetribcsoftheUpper
Xingu duri ng lhe summer and fall of 19j5. Thi s work was
supported by grantsfrom the HeHT\ Frank Guggc nheim Founcla-
hon and the Nati onal Sciencc Foulldation
1. Richard Lee(1 9-:- 9: 370- 400)documentsalonghi storyof2Z felld-
li kc kilJings among a smalJ population of Kung. Although the
demographi c impact of sll ch battles is hard to qualify, it would
seemto accountforatleastas high apercentagcofthedeathrateas
war in OUT own societv(about 1-2%ofalJ dea ths).'rhcsocialand
psychological impact violencewithin Kungsocie!:) maycvcnbe
higher. IftheseareOllTmeasuresof\Vhatconstitutcs\\ ar,tl lene\en
"harmless pcopl e" qualify.
2. Speech, sexual relati ons, and eati ng are all cl oscly connectecl
activities in the concepti on ofthe human body.
3. Theconcept allc1 importance ofpcacduJncss isrough ly similarlo
123
THOlvlAS GREGOR
thal of"ifutisu,"the tefln used by the tribes in the
Xingu system (sce Basso 1973: 12-1'+,.
4. Foliage from this pblltis burnedand theas his Icaehed \Vith water
lo produce potassium ehloride, the sall ofthe Xinguanos As the
villagcrsdescribe it, the Carib tribes' speeialii'ati ol1S in , hell belts
and necklaces should also be consldered labor intensive
monopolies, sineetheyinvoll'elonganddangeroustripstofind the
shell s, and arduous\Vork to craft them into Ihe final produet.
5. AII oftheXingutribeseatfi sh,monkey,andanumberofspeciesof
birds, most game animals.
6. The\'i1lagersarefar moreaeute1yattunedtoodorsthan\Ve are, and
will oftenspitto reducethesensationofabadsmell. In Jeeountsof
interaetionwith both wild Indiansand for es tan imals(with whom
wild Indians are believed to share a great deal in eommon) the
Xinguanos often remark on thefetid smell.
7. Some of tlIc Xinguanos claim to have seen a tribe eall ed "Thc
PeopleWhoLive Inside Trees" somedi stanee totheeastoftheir
reservati on. The tree people purportedly live in tree trunks, eat
uneooked birds , \Vear no clothes, and spea kin hootsand grunts.
Symbolieally, this eulturelesstribefulfill sanani mal-likepotential
whieh other wild Indians onl yapproximat e.
8. In 1972, 1witnessed thekillingofan Xingu Indianwho\Vas haeked
to death \Vith machetesas we tri ed to hold off hi sJssailants in a
room at the Indian Post. .\ftertheexeeuti on, thekillersexhibitcd
an iey calm. Oneapproaehed me, pointed to thecorpse and thc
wa i)ing relatives and said: "You can take photographs now."
9. Thcse findings are based primaril y on Iny work \Vith Tupi and
Arawak spea king informants.
124
6
Raiding, trading and tribal
autonomy in insular Southeast Asia
THOMAS GIBSON
In this paper, warfare in insularSoutheastAsia is examinedthroughthe
compari sonofthreegroups ofhighlandshiftingcultivators .I Thethesis
is that thecurrent social evaluation ofviolence and aggression within
eaeh group is the result of differing historical experi ences wi thin a
loosely integratedregional political economydominated bytheinstitu-
ti ons of slave raiding and coerced trade. Those groups positively
disposed toward bellicosity are those which pl ayed a predominantly
preda toryrole in theregion,while those which are negati velydi sposed
toward violence in anyform were primaril yprc\. There is, however,
nothingdetermini sti cabout the argument , for eachconcreteexample
rcpresents onl )' oneofanumberoflogicallypossibleresponsestoagiven
sequencc ofhistorical events. All three groups still retaill asignifi cant
degreeofautonomyovertheirinternalpoliticalandideol ogical systems,
and each must be seen as creatively responding to a changing se t of
e>..iernal politi cal and economic forces.
In thefi rstpartofthe papcr, abriefoutlineisprovided ofthegeneral
hi storical context in which these three soci eti es havedeveloped. [nthe
secondpart,anovcrviewofeaehsocich is gi ven. III thethirdpart, their
respecti veattitudes toward\'iolcncc, social rankingandindebtednessare
125