Handbook Steward

SMITHSONIAN INSTITUTION
BUREAU OF AMERICAN ETHNOLOGY

BULLETIN 143
HANDBOOK
OF
SOUTH AMERICAN INDIANS
Julian H. Steward, Editor
Volume 5
THE COMPARATIVE ETHNOLOGY
OF SOUTH AMERICAN INDIANS
Prepared in Cooperation With the United States Department of State as a Project

of the Interdepartmental Committee on Scientific and Cultural Cooperation
UNITED STATES
GOVERNMENT PRINTING OFFICE
WASHINGTON : 1949
For sale by the Superintendent of Documents, U. S. Government Printing Office, Washington 25, D. C.
Price $3.00
Digitalizado pelo Internet Archive.

Capítulos extraídos pela Biblioteca Digital Curt Nimuendajú:
http://www.etnolinguistica.org/hsai
Part 4. South American Cultures: An Interpretative
Summary ^
By Julian H. Steward
introduction
It is the purpose of the present article to provide a basis for classi-

fying South American Indian cultures and to present comparative
summaries of the principal cultural types in terms of their ecological
adaptations and historical development. It endeavors to reduce the
bewildering variety of cultural data to categories which have a real
and historical meaning. In this respect, it differs from the method
of historical particularizing which treats each tribe and culture as
unique, emphasizing their peculiarities and stressing the exceptional
rather than the general. The latter method is valuable in detailed
analyses of individual tribes, but, applied to large areas, it gives the
impression that cultural elements and patterns occur in a random and
fortuitous manner.
Admittedly, an attempt to subsume large numbers of tribal cultures
under general types, as undertaken here, encounters difficulties pre-
sented by borderline cases, by insufficient data, and by possible mis-
interpretation of data. The general character of the main cultural
types will doubtless have to be redescribed in the light of new infor-
mation and more detailed comparisons, and many tribes will quite
probably be found to belong to types other than those to which they
are here assigned. Authors familiar with certain tribes will find that
the generalizations do not do their tribes justice. Science will be best
served, however, by correcting faulty generalizations with better gen-
eralizations carping at these constructs because a few special facts do
;
not fit them can only lead to the impression that culture development
is utterly capricious and haphazard.
The fourfold classification used herein (map 18) corresponds in

—
general to the four volumes of the Handbook Marginal tribes, Trop-
ical Forest peoples, Circum-Caribbean and Sub-Andean peoples, and
Andean civilizations —except that in retrospect it is evident that many
1 1 am grateful to Drs. A. L. Kroeber, Robert H. Lowie, W. D. Strong, John M. Cooper,
and Gordon R. Willey for having read and criticized this summary.
669
;
670 SOUTH AMERICAN INDIANS [B. A. E. Bull. 143
( Solid black, Central Andes

Map 18.—Distribution of aboriginal culture types.
horizontal liachure, Circum-Caribbean ; cross-hachure, Tropical Forest and
Southern Andes; diagonal haehure, Semi-Marginal; stipple, Marginal.)
Vol.5] SOUTH AMERICAN CULTURES STEWARD 671
tribes were improperly classified. In the preceding volumes of the

Handbook, the tribes were classed more or less on the basis of impres-
sions. The classification, like previous ones,^ used principally the
general element content of the cultures rather than a systematic com-
parison of the patterns. Special weight was accorded one or another
feature in each case. The Marginal peoples were distinguished by their
lack of farming and their generally simple cultures the Tropical For-
;
est peoples were identified by their agriculture and various material

traits, which were adapted to the tropical rain forests; the Circum-
Caribbean and Sub- Andean peoples were grouped together because of
their class system and temple cult; and the Andean peoples, from
southern Colombia to Central Chile and Northwest Argentina, were
distinguished by their Central Andean technology, material culture,
and ritual complex.
The present classification is based primarily on sociopolitical and
religious patterns. Culture elements are accorded secondary impor-
tance because too many
of them are independent variables. Their
distributions were dissonant with those of the sociopolitical and
religious patterns, and they occurred in quite different patterns. They
were the building materials of culture and did not greatly affect the
architecture. The bow and example, occurred in widely
atlatl, for
differing patterns of hunting and warfare.
Kitual elements, such as
hair cutting, flagellation, or the scratching stick, served a very dif-
ferent purpose in each local context. Deities, such as the sun or stars,
were variously mythological characters, shamans' spirits, or tribal
gods. Items of adornment, such as face painting, tattooing, and ear-
plugs, were badges of tribal membership, sex, society affiliation, or
class status, according to the tribal patterns they entered. Even ele-
ment complexes, that is, stable groups of elements, were found in dif-
ferent settings. For example, the initiation of boys into a secret
organization, which used a sacred trumpet to represent the voice of
the gods, was part of an ancestor cult in the northwest Amazon and
of a priest-temple cult among the Mojo.
A classification based on culture elements would not at all corre-
spond to one based on sociopolitical patterns. In terms of culture
elements, eastern Bolivia would be classed with the Tropical Forests,
and the Northern and Southern Andes would be included with the Cen-
tral Andes. The Circum-Caribbean peoples would belong with the
Tropical Forests if material elements were emphasized and with the
Andes if social and ritual elements were given more weight. Culture
elements have greatest classifactory significance in the case of the
Marginal peoples, who, though very heterogeneous, differed from all
other South American tribes in certain important absences. They
^'Wissler's fivefold grouping (1922), Cooper's three-fold (1942), and Stout's nine-fold
(1938).
:
loom weaving, twilled and woven

characteristically lacked agriculture,
basketry, dugout canoes, and a considerable list of lesser items found
among their neighbors.
The patterns chosen herein as the basis for classification are those
which integrate the institutions of the sociopolitical unit. The unit
is the cohesive group whose members live in more or less permanent
association with one another, participate in the same economic, social,

and religious activities, and submit to the same in-group sanctions
and political controls. It is the group of persons whose varied and
reciprocal behavior patterns form a self-contained cultural whole. It
is sometimes but not always the tribe the term "tribe" is often applied
;
to a group of units which, though culturally and linguistically similar,

are politically independent of one another. The magnitude of the
sociopolitical unit varies In rare cases, it is the conjugal family more
: ;
often, it is the extended family or lineage ; frequently, it is the multi-

family or multilineage community and, in some areas, it is the multi-
;
community state, federation, or empire.

The pattern or structure of each unit varies not only with its size
and composition but with its special cleavages, which may be based
variously on kinship, sex, age, and status and on military, religious,
and economic activities. The units are thus distinguished by such
institutions as clans, lineages, and other kin groups, division of labor,
secret societies, special associations, warrior classes, a priesthood, a
nobility, and the like. Though every independent sociopolitical unit
differed somewhat from all others in the way it patterned these insti-
tutions, certain broad configurations occurred over wide areas, and
it is possible to group them in four principal types, as follows
(1) The Marginal peoples had sociopolitical units which consisted

either of a single kin group or of several loosely organized kin groups.
Members of the unit were differentiated on the basis of age, sex,
economic activities, and sometimes associations. Behavior was sanc-
tioned by the informal and often unconscious influences of tribal
custom operating through these institutions. These tribes had very
similar crisis rites, shamanism, and magic, and their technology and
material culture, though not homogeneous, was rudimentary and
generally lacked the developed agriculture, building arts, and manu-
facturing processes found among other South American Indians, The
sociopolitical patterns, however, varied with local conditions. Be-
cause the Marginal tribes lived in areas of limited resources and had
elementary exploitative devices, the size and composition of their
groups and many of their institutions had to be adapted to subsistence
needs.
(2) The Tropical Forest and Southern Andean peoples also had
sociopolitical units consisting principally of kin groups and structured
Vol.5] SOUTH AMERICAN CULTURES —STEWARD 673
along lines of age, sex, and associations, but theirs differed from those
of the Marginal tribes in that more developed exploitative devices,
which included farming, and better transportation afforded by the
canoe, permitted larger and more stable units. Social control was
informal, except in a few communal activities such as warfare, which
often had a special chief. They also had a richer technology and
material culture, but their crisis rites and shamanistic patterns were
of the same types as those of the Marginal tribes.
(3) The Sub-Andean and Circum-Caribbean peoples, though sim-
ilar in technology and material culture to the Tropical Forest Tribes,
had a more effective subsistence complex which supported a denser
population and larger and more permanent villages. The villages
were composed of many non-kin groups and were organized on the
basis of classes rather than merely of age, sex, and associations.
Warfare, carried on by the Marginal peoples mainly for revenge and
by the Tropical Forest tribes for revenge and for personal prestige,
became, among the Sub-Andean and Circum-Caribbean peoples, the
principal means of achieving membership in the upper class, and war
captives formed a slave class. Individual behavior was still largely
sanctioned by custom, but governmental regulation through state law
was foreshadowed by special and delimited powers accorded chiefs,
warriors, and shamans in particular contexts and for delimited periods.
Among the Tropical Forest and Marginal peoples, the shaman dealt
principally with his own spirit helper, and his principal function was
to cure and practice magic; among the Sub- Andean and Circum-
Caribbean tribes, he served also as the priest in a temple-idol cult
dedicated to the tribal gods, though his rites tended to be private
oracular sessions rather than public ceremonies in a ritual calendar.
(4) The Central Andean peoples had the most developed agricul-
ture in South America, the most dense population, and very efficient
transportation, the combination of which permitted the growth of
true urban centers and the extension of social interrelations and po-
litical controls over large areas. Patterns implicit in the Circum-
Caribbean and Sub-Andean culture were fully developed in the Central
Andes: a rigidly hereditary class system; war for conquest rather
than for personal advancement through the capture of slaves a temple
;
cult with a hierarchy of gods and priests public ceremonies forming

;
a ritual cycle and regimentation of a considerable portion of the lives

;
of the commoners through political controls enforced by the state more

than by the sanctions of tribal custom. These institutions overshad-
owed and to some degree replaced behavior patterns pertaining essen-
tially to the kin group and the community. Central Andean material
culture differs little technologically from other areas, except in
;
metallurgy and the building arts, but it is outstanding in the quantity,

variety, and excellence of its products, and finer goods were produced
by special craftsmen for the upper classes.
This fourfold classification has developmental implications in that
some institutions and practices were necessarily antecedent to others,
but it is not a unilinear scheme. As will be shown in a concluding
section, a strong historical tradition carried certain sociopolitical in-
stitutions and probably several technologies throughout a considerable
portion of South America, but the acceptance and patterning of such
institutions was always contingent upon local potentialities. The po-
tentialities were a function of the local ecology, that is, the interaction
of environment, exploitative devices, and socioeconomic habits. In
each case, the exigencies of making a living in a given environment with
a specific set of devices and methods for obtaining, transporting, and
preparing food and other essential goods set limits to the dispersal
or grouping of the people and to the composition of settlements, and
it strongly influenced many of their modes of behavior.
TheYahgan, for example, quite obviously had to live in small, widely

spaced groups if they were to obtain sufficient shellfish in their coastal
habitat. The Ona and Tehuelche could live in larger, nomadic hunting
bands, and those of the Tehuelche undoubtedly grew in size after they
obtained the horse. Probably as a function of the hunting economy,
the bands were made up
of patrilineal lineages. Along the rivers and
coasts of the Amazon area, a more abundant subsistence and developed
canoe travel permitted large population concentrations, and the settle-
ment pattern and sociopolitical institutions had wider latitude for
variation. The Circum-Caribbean area probably had still greater
latitude, but it from the Tropical Forests because of the his-
differed
torical influences which brought it a class-structured society. The
Andes had the most efficient subsistence pattern of all South America,
and it developed over a long period within a strong historical tradition.
The ultimate origin of its patterns has not yet been disclosed, but their
influence on non- Andean peoples and their final formulation in the
Central Andes by the Inca can be traced with some certainty.
In the following pages, population densities are related to com-
munity size. South American cultural data are then classified and
summarized according to the four sociopolitical patterns. Under
each, the subsistence activities and settlement patterns are described
the social, religious, and political institutions are analyzed; and the
technology and material culture are briefly sketched. A concluding
section traces basic trends in the development of South American
cultures from the earliest known periods to the present day.
.

POPULATION AND COMMUNITY SIZE
Population densities suggest potentialities for community develop-

ment, but far greater insight into the actual setting of sociopolitical
patterns is afforded if they are analyzed in relation to settlement types
and community size (map 19)
Population density and community size are by no means correlated,
because the number of people who may live together in permanent
association depends not only upon the number of persons per unit of
area but on the quantity and distribution of resources, the transporta-
tional facilities, and the sociological factors, such as warfare and re-
ligion, that determine settlement patterns.
The Marginal tribes of Southern Argentina and Chile had a com-
paratively sparse population, ranging from 2.5 to 9 persons per IOC
km.^ In addition, their reliance on wild foods, which were widely
scattered, and the necessity of transporting their goods and foods on
their own backs (in some place they used bark canoes) precluded large,
permanent settlements. Even where their population density equaled
that of neighboring Tropical Forest tribes, their nomadism prevented
community growth, for example among the Naiiibicuara^ Mura^
/Siriono, and Western Amazon Semi-Marginal tribes. (Cf. maps 16
and 19.)
A striking illustration of the importance of transportational facili-
ties to community size is afforded by the tribes which adopted the
horse after the Conquest. Among the peoples of Patagonia, the
Pampas, and the Chaco, the bands increased several fold after the
adoption of the horse.
Many of the Marginal tribes had groups consisting of only a single,
conjugal family, that is, five or six persons, or slightly more if there
were polygamy. Where a greater number of people remained in perma-
nent association, they tended to form an extended family or lineage
rather than a group of unrelated conjugal families. The limit of the
lineage, however, was usually about 60 persons. When a community
or band exceeded this number, they often lost a sense of being related
to one another. Larger groups, therefore, generally consisted of
several unrelated lineages, as among the somewhat agricultural tribes
of the Chaco, who had communities of some 200 persons, and of eastern
Brazil, where villages had 400 or more.
The had permanent food resources near their
agricultural tribes
villages, but farming did not have the same relationship to com-
munity size in each area. In the Tropical Forests, where the popula-
tion density ranged from 10 to 50 persons per 100 km. 2, cultivation
676 SOUTH AMERICAN INDIANS [B. A. B. Bull. 143
Map 19. — Size of native communities. Tlie average size is given for each area.
.
was by the slash-and-burn method, and the fields, and in some cases
the village, had to be moved periodically. The soil was tilled by
women, who, having many household duties, could go only a limited
distance to the fields. The Tropical Forest villages, which had as
many as 1,000 persons among the riparian and littoral tribes, were
supported more by the combination of water resources and transpor-
tation in efficient, dugout canoes than by farming. Conceivably,
communities could have been larger had they been dispersed, but war-
fare kept them tightly nucleated several large houses, each sheltering
;
a lineage, were clustered inside a palisade. In some areas, however,

as along the upper Amazon, houses were strung along the river with-
out special grouping, though the population was comparatively dense
(at least 30 persons per 100 km.^)
In the CircumCaribbean area, the denser population (several hun-
dred persons per 100 km.^) and the larger communities, which had 1,000
to 3,000 persons each, represent an intensification of the Tropical For-
est ecology: somewhat better agriculture, abundant sea foods, and
excellent canoe transportation. The large villages, however, were
not solely the product of the numerous population. The villages
tended to grow because they had become administrative and religious
centers and because their population was augmented with captive
slaves. The Araucanians, also good farmers and with a population
density about equal to that of the Circum-Caribbean area, lacked these
sociological factors which made for community growth. Instead of
fewer and larger communities, they had numerous small ones, each
constituting a lineage of 100 to 150 persons. Many other tribes simi-
larly failed to achieve the community size which their ecology would
have permitted.
In the Central Andes, the population, which had a density of 720
persons per 100 km.^, was so concentrated in the fertile Coastal and
Highland valleys that the settlement pattern was one of almost con-
tinuous dispersal over areas that were inhabitable. Instead of con-
centrating in fortified villages (the North Coast was probably an
exception), refuge was sought in hilltop forts. Religious and ad-
ministrative centers, however, provided nucleating points for com-
munities of varying sizes. The ayllu village usually had several
hundred inhabitants. The more important centers had several thou-
sand people, and some communities were true cities, for example,
Cuzco, which had some 100,000 persons. The basis for these urban
communities was the combination of the dense population, which was
supported by highly efficient agriculture, coastal transportation by
sea-going balsa rafts and inland transportation by pack llamas travel-
ing on roads, and the strong development of religious and adminis-
trative centers.
These general, comparative data on population densities and com-

munity size will serve as a background to the analyses of sociopolitical
and cultural types which follow.
THE MARGINAL TRIBES
The Marginal tribes were essentially hunters and gatherers, and

they exploited comparatively unproductive environments by means
of a simple technology. Their sociopolitical units were small, some-
what unstable, and frequently nomadic, and the relations of their
members to one another were governed by kinship, age, sex, and as-
sociations. The size, permanency, and composition of their groups
were strongly affected by subsistence patterns, each adapted to a
distinctive environment, and by special local developments, such as
clans, moieties, and associations. Sociopolitically, therefore, the
Marginal peoples differed from one another as much as they differed
from other South American Indians.
In terms of culture elements, the Marginal tribes were also extremely
varied not only because many of the comparatively archaic features
are unevenly distributed among them but because locally they bor-
rowed many items from their Tropical Forest and Andean neighbors.
The following are some of the elements which certain Chaco tribes
borrowed from the Andes Loom, weaving, painted cloth, tie dyeing,
:
ceramic styles, dice games, some farming, use of salt and other condi-
ments, fermented drinks, slings, feather fire fans, sandals, body paint
applied with stamps, drums, flutes, whistles, eyed needles, some myth
themes, and other minor items (Handbook, 1:210-211). From the
Tropical Forest tribes they evidently took mortars, scalping, and use
of tobacco a few tribes also borrowed house types, hammocks, urucu,
;
and arrow types.

The tribes of eastern Brazil acquired a considerable list of items from
their Tropical Forest neighbors, and the Northwest and Central Ge
might even be considered transitional between the Marginal and Trop-
ical Forest tribes. These items include: Farming, large thatched
houses, dugout canoes {Garajd^ Guato) reburial and urn burial, ham-
,
mocks (some Ge^ Garari, MasJiacali^ Carajd), rubber balls (ApinayS,

Sherente) various musical instruments, penis sheath {Bororo^ Gayapo^
,
Gamacdn, '^Tapuya''')^ penis thread {^''Tainiya^'' Patasho), feather-

work {Goytacd^ Botocudo^ Tinibira, Gentrol Ge)^ earplugs, labrets,
use of genipa and urucu as body paint, and others.
The Gharrua and Querandi near the Parana Delta adopted the tak-
ing of trophy heads from the Tropical Forest tribes and, doubtless
under the influence of the neighboring Tupian peoples, they made
canoes and thatched houses and wore ear, nose, and lip plugs.
Stripped of these borrowings, the Marginal cultures would be ex-

tremely simple. They would lack agriculture and domesticated ani-
mals. They would also lack loom weaving and probably finger-
weaving or twining, though some of them would probably know netting
techniques. They would have no basketry (the Fuegian coiled baskets
are probably Andean-derived). They would lack pottery or make
only a very crude ware. Their canoes would be only of bark and their
houses one-family, conical, domed, or lean-to type shelters. They
would prepare their food directly on the coals, in an earth-oven, or
in water made to boil with hot stones, and they would eat it without salt
or condiments. They would use no clothing, except skin robes in the
colder areas, and they would lack ear, nose, and lip ornaments, adorn-
ing their bodies only with necklaces and paint. Their musical instru-
ments would be limited to rhythm beaters, but probably excluding rat-
tles as well as drums, and their only games would be shuttlecock and
athletic contests. They would use no narcotics or stimulants.
THE ECOLOGICAL BASIS
Subsistence. —The subsistence potentialities of the different Mar-

ginal groups are difficult to estimate quantitatively, but there is no
doubt that food could be had only in limited amounts, so that the
general population was extremely sparse. In addition, transporta-
tional facilities were so poor that, unable to assemble food and other
goods at points of concentrated population, people had to disperse so
as to be near their resources.
On the whole, the Marginal peoples inhabited plains or savannas
and relied upon collecting wild seeds and fruits, hunting small game,
and fishing. This required that they split into small groups, fre-
quently of individual families, and move seasonally from place to
place as different foods became available. The Patagonians and the
Ona of Tierra del Fuego, however, were somewhat distinctive for
the emphasis on large-game hunting (the rhea and guanaco), the
yield of which supported cohesive bands of 50 or 60 people. The
maritime peoples of the Chilean Archipelago, on the other hand,
were essentially shellfish gatherers who lived scattered in small
clusters along the seashore. Though the Marginal tribes generally
avoided large streams, many Guato^ Mura^ and Yahgan virtually lived
in canoes, each vessel carrying an encampment of a small, bilateral
and independent family.
It would be incorrect to characterize all the Marginal tribes as
nonfarmers. Although the peoples of southern Argentina and Chile
lived south of the limits of any American Indian crops, many of the
tribes farther north cultivated on a limited scale. It is more im-
portant that this farming was not very productive. Outside the
738931—49 45
Andes, with its potatoes and cereal-crop staples, agriculture was

based on tropical root crops, especially manioc and sweet potatoes.
These require a rain forest environment, but the Marginal tribes
generally occupied plains or savannas, and rarely had access to
gallery forests where they could be grown.
Agriculture was wholly lacking in the Archipelago, and among the
Tehuelche^ Puelche^ Charrua^ Puri-Coroado^ Guiana Marginals, and
Ciboney. Of the Orinoco groups, only the Yaruro farmed a little.
Agriculture occurred mainly among tribes of eastern Brazil, among
some near the Parana Delta, on a small scale in the more favored
portions of the Chaco {Zamueo, Tunierehd^ Lengua^ Ashhislay, Ma-
taco^ KaskiJid, Sanapand^ Pilagd) and among the Huarpe and Come-
chingon, the last two showing Andean influence in their maize,
beans, and quinoa, and llama herds. During the historic period, the
Caingang, Yaruro and probably Guayahi abandoned farming, where-
as the Botocudo^ Goroado, and perhaps the Poya seem to have adopted
it.
The main crop among most tribes was maize

(it was the only Huarpe
crop) but it was eaten green by the Guayahi^ Mashacali, Malali, and
,
Yaruro, presumably a practice of incipient farmers. The tropical

root crops were grown only by some of the Ge tribes, who cultivated
strips of gallery forest.
—
Transportation. Transportational facilities are no less important
than food resources in determining the ability of a population to group
itself in sociopolitical units. They also delimit the quantities of
material goods a nomadic people may possess. In aboriginal times,
the Marginal tribes used either human carriers or canoes, though there
was a doubtful case of dog packing in Patagonia. Perhaps the oldest
device was the skin bag, which was used at the Contact Period in the
—
far south. The netted bag and the carrying net ^the former widely
distributed north to Mexico and the latter found in the Andes and
—
north to western North America were probably also old methods.
These were used in the Chaco, and among the Caingang, Botocudo,
Mashacali, and Yaruro. The woven carrying basket of the Tropical
Forests was adopted in eastern Brazil {Northwest and Central Ge,
Bororo, and Guayahi) and by the Guahibo.
,
The canoe was of major importance to the Tropical Forest peoples,

but among the Marginal tribes its use was limited and its construction
primitive. Many Marginal tribes lacked canoes simply because they
did not live on large streams. At the same time, they often purposely
avoided large rivers because stronger, more aggressive, and definitely
more aquatic-minded Tropical Forest peoples occupied them.
Canoes were absent among the Tehuelche and Puelche of the almost
streamless Pampas, though in the Pampas and among some Chaco
tribes the buUboat was occasionally used. Other tribes lacking canoes
were the Guayaki^ Caingang, Bororo^ Puri-Goroado^ Tapirape^ Siriono,
Nanibicuara, Guiana Marginals, most of the Northwest and Central
Ge, and many of the Chaco tribes. The Bororo and Botocudo adopted
dugouts in the historic period.
Bark canoes were used by tribes living peripheral to the Amazon
Basin on its headwaters, by the Suya, Mura^ and upper Xingu tribes,
and by the Archipelagic peoples. The last, however, adopted the
dugout early in the historic period, and later the plank canoe. Dug-
outs were apparently native to the Guato^ many of whom virtually
lived in them, and to the Gudhibo of the Orinoco, the Carajd^ some of
the Chaco tribes, and the Charrua. Under Andean influence, the
Huavpe made reed balsas.
The importance of transportational facilities to the sociopolitical
group is well illustrated by the Tehuelche^ Puelche^ and Charrua^ who
adopted the horse during the early historic period. Through increas-
ing the yield of the hunt and facilitating the movement of foods to a
central point and of peoples to the sources of food, it enabled the
small, native bands of foot Indians to amalgamate into large, multi-
lineage hunting bands. Comparably large bands elsewhere were only
seasonal and temporary, as in the Chaco during harvests, or in the
Archipelago when a whale was stranded.
—
House types. The nomadic or seminomadic life and the small
social units of these tribes precluded the use of large, permanent,
frame, thatched houses of the Tropical Forest types. Some tribes
lacked houses altogether {Namhicuara^ Guahibo^ Ctboney^ Guamo^
Taparita, Gayon) others made some kind of crude, temporary shelter
;
{Guaja, /Siriono, Mt(/ra, Yaruro). Where construction is described,

one of four types is indicated: (1) A
domed pole hut covered with
leaves, brush, or grass (Archipelago, Mura, Botocudo, Patasho,
Mashacali, Macuni, Chaco) (2) a conical,
;
tipilike pole hut {Yahgan,
Guaranoca in the Chaco) (3) a portable
;
pole toldo covered with mats
{Charrua and some Chaco tribes, especially the horsemen) or skins
(Pampas, Patagonia, Charrua) (4) a single or double lean-to {Ona,
;
Guato, Puri-Goroado, recent Gaingang). The large, frame multi-

family house is reported only among the Bororo and possibly the Mash-
acaU, and it was used by some Northern Ge as a clubhouse. The
Huarpe and Gomechingon, under Andean influence, made slight use of
masonry houses and possibly of semisubterranean houses.
People slept on the ground, except among a few tribes which had
borrowed the hammock from the Tropical forests {Siriono, Charrua,
Apinaye, Eastern Timbira, Sherente, Puri-Goroado, Gudhibo, Yaruro^
Gayon, Mashacali, and a few upper Xingu and Guiana Marginal
tribes) and some of the Ge who used a platform bed. The Pilcomayo
and Berrnejo River tribes of the Chaco used the hammock as a cradle.
SOCIOPOLITICAL AND RELIGIOUS PATTERNS

Three principal sociopolitical types and several subtypes were found
among the Marginal tribes. The principal types were: (1) Groups
consisting of a single bilateral or conjugal family; (2) the unilateral
band, consisting of a single, extended family, either matrilineal or
patrilineal; and (3) the mixed band, consisting of several unrelated
families, lineages, or clans. Where the group was composed of a
members were kin and it was, therefore,
single family or lineage, all its
exogamous. This usually meant local as well as group exogamy. The
relationship of the members to one another was based on kinship, and
the headman was the family elder. Where the band was composed of
several unrelated families, lineages, or clans, local exogamy was not
necessary and social behavior and political controls extended beyond
kinship relations.
—
(1) Family units. The bilateral or conjugal family, consisting
merely of father, mother (or mothers), and children, underlies all
sociopolitical types. Among the Guato^ Mura^ and Namhicuara^ it
was the only permanent sociopolitical unit, and, though several fami-
lies might associate with one another seasonally for special activities,
they lacked permanent cohesion and had no leader or chief.
Comparatively little is known about these tribes. The Guato family
was polygynous, and it lived in its own hut, or, during much of the
year, its canoe. The Gimto had a loose band organization in that sev-
eral families occasionally foregathered for temporary association with
one another, but the band had no political significance. Among the
Mura^ some two to five individual family houses sometimes were close
enough to form a village, but each family retained its independence.
The population of these tribes was usually dispersed, so that there
was little opportunity for social activities, but the Mura had puberty
ceremonies, in which girls were isolated and boys were whipped,
given parica snuff, and subjected to the ant ordeal. The elements
of the boys' rite were borrowed from the neighboring Tropical Forest
peoples. The Namhicuara evidently lacked a true secret cult, but in
ceremonies to the thunder god they used sacred flageolets which were
taboo to women and children. These tribes had little or no warfare
by which individuals could win honors, but among the Guato^ indi-
vidual prestige was won by killing a jaguar.
(2) Unilineal bands. —
The unilineal band consisted of a single ex-
tended family or lineage, i. e., a man, his wife, and their descendants
through either the male or female line. The band was patrilineal
when postmarital residence was consistently partilocal, women going
—
to their husband's family. In this case, it consisted of husband, wife,

unmarried daughters, and married sons whose wives came from other
bands. When the residence rule was matrilocal, the band was matri-
lineal and consisted of parents and their unmarried sons and married
daughters whose husbands came from elsewhere. Extended families
were also common in the Tropical Forest area, where, however, they
were much larger than those of the Marginal peoples, often consisting
of 100 or more persons occupying a large, communal house (map 20).
(a) Patnlineal hands.—Âmong the Archipelagic tribes Alacaluf,
Chono, Tahgan, Ona, and Haush (Handbook, 1) —there were vari-
vol.
ous gradations from the independent conjugal family to the patri-

lineal band. Among the maritime Alacaluf and Haush^ patrilocal
families often lived alone, but among the Chono and Yahgan^ three or
four related families stayed together in shifting villages. Each of
these small sociopolitical units claimed a certain territory within
which it traveled and which it defended against trespass. Warfare
against trespass and in revenge for witchcraft strengthened group
cohesion. Political control was synonymous with kinship control by
the headman. The only divisions within the local group were those
based on sex and age. The former included division of labor, the
Yahgan secret society (which was probably related to that of the
Ona) and the various girls' puberty rites. Age gradations were some-
,
what institutionalized through the initiation of pubescent boys and

girls into the tribal society.
The Ona socioeconomic group, based on the collective hunting of
land game, was larger than that of the Archipelagic peoples, ranging
around 60 persons, but it, too, was based on kinship and consisted of
about a dozen patrilineally related families. It had the characteris-
tics of the typical patrilineal band (Steward, 1936), being patrilineal,
patrilocal, exogamous, and land-owning. Thus, it resembled a local-
ized sib. The lineage head was the headman of the band. The only
formal sexual grouping was the secret tribal society to which boys
were initiated at puberty. Girls were isolated at puberty. The func-
tion of warfare was about the same as among the Archipelagic tribes.
—
(6) Matrilineal hands. The Guayaki^ nomadic hunters and fishers,
were grouped in bands of some 20 persons each (Handbook, 1: 441).
The band had a strict rule of matrilocal residence and so must have
tended strongly to consist of a single matrilineal lineage or extended
family. There was no formal social differentiation within the band
except by sex and age. Pubescent girls were scarified and observed
meat taboos boys had their lips perforated.
;
The Siriono (Handbook, 3: 458) had extended matrilineal families

that averaged perhaps 15 persons each. The entire extended family
SO 75 60 45
\K
15
15
KINSHIP MSIS
OF SOCIETY
Conjugal family
Lineage, patrilineal
llultilineage patri-
,
lineal
45 — Clan, patrilineal
Lineage, matrilineal
Multilineage, niatri-
lineal
Clan, matrilineal
Kultifamily or bi-
lateral groi:p
Map 20.—Distribution of the types of kinship basis of aboriginal sociopolitical

groups.
traveled together and the constituent conjugal families often lived

together in a single hut.
(3) Mixed bands and villages. —The mixed bands consisted of
several unrelated families or lineages. They differed according to
whether the component units were individual conjugal families, matri-
Unless fixed by rigid rules of
lineal lineages, or patrilineal lineages.
postmarital residence, these types intergraded and overlapped.
Nevertheless, a tendency either to matrilocal or patrilocal residence
was usually very strong, and most bands appear to have been aggre-
gates of one or the other kind of lineage. In some cases, mixed bands
were formed during the historic period, when new economic factors
permitted previously independent lineages to band together in perma-
nent association.
—
(a) Mixed hands with a matrilineal base. The foot Indians of the
—
Chaco the Mataco^ Choroti^ Ashluslay^ Macd, Lengua^ Toba^ and
—
Lule-Vilela ^had land-owning bands of 50 to 200 persons. The re-
cent Ashluslay villages of 1,000 are exceptionally large and may repre-
sent a post-Columbian condition. Though the smaller bands or vil-
lages may have consisted of a single lineage, most were made up of
several extended families, which, being strongly matrilocal, must have
been somewhat matrilineal. The band as a whole, however, was en-
dogamous, except among the Pilagd.
The occurrence of matrilineal lineages mainly in the Chaco is proba-
bly to be explained by the relatively great importance of seed gathering,
which placed women in a strong economic and, therefore, social
position.
The band chief was usually an elder who attained leadership through
personal influence. Some areas had a district chief, an office very
possibly resulting in part, at least, from White influence. Hereditary
chieftainship was uncommon.
A special bachelors' house or club was not a typical Chaco feature,
but occurred among the Pilagd and Ashluslay. The ChamxiGoco initi-
ated young men in a secret rite. The Chaco held public ceremonies at
girl's puberty.
Warfare contributed to social cohesion among these tribes, but did
not form the basis for a warrior class. Scalps were the only trophies.
The Tapirape and Carajd, also had bilateral bands, each consisting of
several extended, matrilineal families.
—
(&) Mixed hands with a patrilineal hose. The Tehuelche^ Puelche^
Pehuenche^ and Poya^ of the Pampas and Patagonia (Handbook,
1 127) and the Huarpe (1 169) and Comechingon (2 673) of western
: , : :
Argentina had mixed bands with a patrilineal base. The last two,
though little known, had some Andean features (e. g., farming, llama
raising, and stone construction) that placed them slightly above a
Marginal level, though sociopolitically they probably resembled their

neighbors.
The aboriginal sociopolitical organization of these tribes is not
known, but they were hunters rather than seed gatherers and, there-
fore, were probably predisposed to patrilineal families. Their ecology
and social patterns very likely resembled those of the Ona. When
first described in any detail, they already possessed the horse, which
they obtained early in the 18th century from the Spaniards.

The horse must have revolutionized their ecological adaptations
through its usefulness both in hunting rheas and guanacos on the
plains and as a means of transportation. The Tehuelche^ Puelche^ and
probably the Poya and Pehuenche became organized in bands of 100 or
more persons, which, unlike those of the Archipelagic tribes, consisted
of several tolderias, or households, each comprising four to five patri-
lineally related families. A
certain fluidity of band membership pre-
vented a band from consisting only of relatives, and there was conse-
quently no rule of band exogamy. Each Tehuelche band claimed
hunting territories, but the boundaries appeared to have been somewhat
flexibleand to have depended on the band's ability to defend them.
Warfare was carried on mainly for revenge against witchcraft,
murder, and trespass, as in the Archipelago, but a few captives were
taken as drudge slaves. These slaves together with chiefs, who had
some power though not hereditary status, marked incipient classes. The
classes were far less developed, however, than among the horse tribes
of the Chaco (Ahipon Mocovi, Oaduveo, and Mbayd, who may have
had an oboriginal predisposition in this direction) or the Goajiro of
Venezuela. The bands were essentially democratic, and the chief
ruled and held a following by persuasion rather than through kinship
ties, through hereditary status.
as in the Archipelago, or
The Patagonian tribes may have had
a men's secret society other- ;
wise, the main sexual differences were in division of labor.

In religion, the Patagonian and Pampean tribes were on the whole
like those of the Archipelago, though group ceremonialism was mani-
fest in the Tehuelche girls' puberty ceremony, in which a horse was
sacrificed, and in the Puelche Elel ceremony, in which an evil spirit
was impersonated.
The Comechingon (Handbook, 2: 673; Serrano, 1945, pp. 329-332),
though Andean-influenced in having permanent settlements based on
farming and llama raising, were somewhat Marginal in that each
settlement evidently consisted of a patrilineal lineage occupying 10 to
40 houses. New lineages budded off, but what were evidently groups
of related lineages formed "ayllus," or larger kinship groups, each
owning its land and having a hereditary chief. A
horse period is not
recorded for the Comechingon,
The Huarpe had patrilineal, probably extended families, each with
itsown farm lands and chief. Men had a special drinking house.
The Querandi (Handbook, 1: 180), Charrua^ and some of their
neighbors of Uruguay (1:191) probably resemble the tribes of
Patagonia and the Pampas in sociopolitical type, but differed from
them in possessing certain social traits borrowed from the Tupian
tribes of the —
Parana Delta and certain religious traits finger mutila-
—
tions, placing skewers in the flesh, and vision seeking ^that are unique
in South America and resemble those of the Plains Indians of North
America.
These tribes became horsemen, organized in hunting bands. The
Gharrua band consisted of 40 to 60 persons under a weak chief. Each
tolderia, or household, consisted of a single family, but the family
composition is not known. Tropical Forest traits, some of them
probably of Twpian origin, are war captives killed at their master's
:
death {Gharrua^ Querandi), skull trophies {Gharrua, Querandi), and

distinctive dress and hammock for the chief {Yaro). These suggest
incipient class structure, which, however, cannot have been well
developed in groups so small.
The Guahibo band, averaging 30 members, may have been unilineal,
but data on its composition are not available.
(<?) Mixed hands with exogamous clans and moieties. —
Sibs or clans
and exogamous moieties, which are really pairs of sibs, differ from
lineages or extended families in that their membership is based on
the fiction of common descent, through either the male or female line,
even though the descent cannot actually be traced, whereas lineages
consist only of true relatives. In addition, a particular clan or moiety
is typically distributed over several places and is, therefore, exogamous
by group rather than by locality, whereas a lineage occurs only in one

place and is exogamous by locality. A lineage could become an exoga-
mous sib or moiety if such features as origin myths and totemism rein-
forced the sense of kinship of its members after their genealogical
connection had become untraceable and if it continued to be exogamous
without regard to where the members lived.
We do not here attempt a theory of sib origin, but it should be pointed
out that the difference between the two is not great and that the
lineage or unilineal band organization is very widespread among the
Marginal tribes. It is patrilineal among the Southern Hunters, matri-
lineal among the seed gatherers of the Chaco, and among many farm-
ers. If two lineages were brought together in a single group or com-
munity and continued to be exogamous, they would constitute moieties,
and if three or more were brought together and continued to be exoga-
mous they would constitute sibs. A change in subsistence, for example
from hunting and gathering to farming, might permit the formation of
.
such multilineage groups, and there is perhaps some significance in the

occurrence of clans and moieties mainly in eastern Brazil, where many
of the tribes practiced considerable farming and where villages were
unusually large for Marginal tribes (the Apinaye villages, for example,
had 500 to 1,400 persons). Once the pattern of sibs and moieties had
been established it could readily spread among neighboring tribes, and
a tendency might develop to form nonexogamous tribal or community
divisions copied from the exogamous ones (amp 20, p. 684)
Whatever the origin of sibs, moieties, and various associations in
eastern Brazil, it is of interest that these sociological refinements
developed mainly in the patterns of the Marginal tribes. There was
no tendency whatever toward a class system.
Exogamous matrilineal moieties occurred among the Pau dÂrco^
Canella^ Bororo (each subdivided into sibs) and Ya7'uro.
, Exogamous
patrilineal moieties occurred among the Sherente, where each was
subdivided into sibs, and among the Gaingang, where each was sub-
divided into two groups with reciprocal functions in mortuary cere-
monies and where there were also some kind of preferential marriage
classes. In spite of their patrilineal moieties, the Gaingang practiced
matrilocal residence.
Exogamous patrilineal sibs and matrilineal sibs were found among
various upper Xingii tribes. Possibly the former had some connection
with the patrilineal sibs of the neighboring Tupian tribes of the
Tapajoz-Madeira region.
Eastern Brazil had a considerable proliferation of moieties in addi-
tion to those which regulated marriage and of age classes and other
groupings. The Ganella were divided into exogamous matrilineal
moieties, three sets of nonexogamous moieties based respectively on
divisions of nature, personal names, and sports groups, and four age
classes. The Sherente had not only patrilineal exogamous moieties
with four sibs in each, but seven grades in the bachelor's hut and four
other men's associations. The Apinaye had four exogamous marriage
classes and also nonexogamous matrilineal moieties. The Tapirape^
Tenetehara, and Garajd, though lacking exogamous marriage groups,
had nonexogamous ceremonial moieties.
A striking feature of eastern Brazil which links some of the tribes
with the Tropical Forest rather than Marginal peoples is a men's
club house Sherente^ Northern Gayapo, Pau d'Arco, Bororo.
:
(4) Sociopolitical units of uncertain composition. The Boto- —

cudo bands of 50 to 200 persons consisted of several extended and
polygynous families, but descent is not stated. The shaman was band
chief.
Puri-G oroado bands had up to 40 persons. Some were single line-
ages,and thus of the unilin eal band type, and others consisted of two
lineages, but the type of descent is not known. The family elder was
the chief. Warfare was carried on, and skull and flute trophies were
taken.
The Mashaeali may at one time have had communal houses, per-
haps implying lineages. Unusual social features were the warriors'
council and the men's house, which served as the center of a cult of the
dead and was taboo to women and uninitiated children. The souls
of the deceased were supposed to appear from the sky during cere-
monies in which the bull-roarer and masks were used.
Information on the Guiana Marginals, most of the Orinoco Margin-
als, and on the Giboney of the Antilles permits no appraisal of their
sociopolitical structure. The same is true of the Guahariho^ beyond
the fact that they had bands of about 40 persons.
—
Warfare. Warfare among the Marginal tribes was rudimentary
and contributed to group cohesion rather than to individual status.
It was virtually never aggressive and was defensive only where the
Marginals adjoined predatory tribes or where trespass on hunting
and gathering territories was involved, as in Patagonia, the Chilean
Archipelago, and some of the Chaco. The most frequent cause of war-
fare was revenge for witchcraft against a group member. Typically,
it was unorganized, there was no cannibalism, except perhaps among
the Botocudo^ and the taking of trophies or captives was rare. The
Ooroado, Oharrua^ Querandi, and neighbors took trophy heads; the
Giaco scalped. Exceptional military practices have been mentioned
under Sociopolitical Patterns.
Religion. —Religion characteristically gave great prominence to
shamanism and magic, and it served individual rather than group
objectives. The Marginal tribes lacked a cult religion, a priesthood,
and group ceremonies, except those mentioned above in connection
with initiation rites, and certain Chaco religious feasts and rites to
control sickness and weather. The concepts of the supernatural, the
specific practices and beliefs concerning witchcraft, magic, and omens,
and the ritual elements varied greatly from tribe to tribe.
—
Religious concepts. ^The Marginal tribes fall into three groups,
according to the supernatural beings to which they attached greatest
importance.
(1) Belief in a High God prevailed among the Patagonian, Pam-
pean, and Archipelagic tribes, but the god is known to have been an
object of prayer only among the last group. There is no proof that
belief in a High God was ever universal in America. To the plains
tribes, lesser spirits, especially evil ones, were of greater importance
in daily life than the High God. The Huarpe also worshiped celes-
tial beings and hills.
(2) Among the Namhicuara, upper Xingu tribes, and the Northern
Ge^ the mythological creator was a celestial being, usually the Sun
or Moon. The Northern Ge had true celestial deities, and the Namhi-
cuara held a ceremony to the thunder god, but in both tribes and
among the Siriono and Caingang spirits, especially of nature, were
more concerned with the affairs of men.
(3) The Bororo and eastern Brazilian tribes attached great religi-
ous importance to ghosts or souls of the dead. The Bororo and Mash-
acali impersonated their ancestors in a cult, and the Caingang be-
lieved that their ancestors, though invisible, attended ceremonies.
In the Canella boys' initiation, the spirits of the dead entered the
neophytes' bodies.
—
Shamanism. Shamanism is a near-universal feature of primitive
cultures and it is certainly very ancient, but it was absent among
some Marginal tribes and very undeveloped among others. Often,
both shamanistic and lay curing of disease was connected with ghosts
or ancestors.
The Siriono and Canella lacked shamans. Among the former, ill-
ness was cured by anyone using the bones of the dead, and among the
Canella by direct appeal to one's deceased kin for aid. Shamans were
rare among the Bacdiri. The Caingang shaman merely consulted
spirits; a sick person was cured by his own relatives. Among the
Carajd^ the shaman only mediated between the living and ghosts;
among the Puri-Coroado^ they conjured up souls of the dead to ask
them questions; and among the Botocudo^ they conjured sky spirits to
a post in the village.These conjuring functions seem to foreshadow
the oracular duties of the Circum-Caribbean shaman, but among the
Marginal peoples they were exceptional and did n,ot involve a well-
formulated tribal cult and deity.
The cases just cited seem to have lacked the concept of the shaman's
spirit helper, but the concept was evidently present elsewhere, though
the nature of the spirit and the manner in which the shaman obtained
it are seldom clearly described. The Archipelagic shaman acquired
his spirit helper, often a deceased shaman, in a dream. The Bororo
shaman's helper was also the soul of a dead shaman; among the
Tahgan^ it was a dwarf or female spirit. In the Chaco, the nature of
spirit helper varied.
Shamans performed various services, including prognostication,
officiating at afew public rites, finding lost objects, and weather
making, but their most important task was to cure illness. Regardless
of concepts of their supernatural power, they cured diseases every-
where by blowing, massaging, and sucking out a foreign object believed
to have lodged in the body. The only exception is that among some
Archipelagic tribes and the Caingang^ Sherente^ Apinaye, Northern
;
Cayapo^ and Botocudo shamans recovered a patient's soul, the loss

of which was believed to cause the sickness.
The concept that any individual may have a supernatural guardian
spirit was doubtless more widespread than recorded. It is attributed
only to some of the Archipelagic tribes and to the Caingang and
Charrua; Charrua men fasted and scarified themselves to obtain it.
—
Rituul elements. A number of ritual elements occur in various
religious contexts. Scarification or blood letting, often about the
same thing, was very widespread Charrua, Chaco, Puri-Goroado (who
:
used the venisection bow) Northern Cayapo, '''•Tapuya^'' Botocudo, and

,
Southern Gayapo. Other elements are of very limited distribution and

occur also among some of the more Marginal peoples of North
America: the drink tube (Archipelagic shamans, Yahgan initia-
tion rites) ; bull-roarer {Mashacdli) ; scratching stick {Yahgan initia-
tions) ; cutting off a finger joint {Gharrua, Minuane, and neighbors)
running a wooden skewer under the skin (Gharrua, Minuane, and
neighbors) steam bath for curing {Puri-Goroado, Botocudo) sacri-
; ;
fices ( Ona; Tehuelche, of horses Huarpe, of chicha and maize)

; and ;
blowing tobacco to the four directions {Tehuelche, historic ?). The

ant ordeal {Mura), flagellation {Puelche, Botocudo, Mwra), and
taking parica snuff {Mura) were probably borrowed from the Tropical
Forest tribes.
Disposal of the dead. —
In contrast to the tribes with a class struc-
ture, theMarginal tribes disposed of all their deceased in the same
manner. Simple earth burial was the usual method, but the Yahgan,
the Guaharibo, the Aweicoma Gaingang, and sometimes the Toha
cremated (the Guaharibo preserved the ashes) the Puri-Goroado ;
practiced urn burial, a Tropical Forest trait; the Gharrua, Ghana,

Bororo, Gamacd, some Northern Ge, and Mataco gave reburial (the
Mataco, after tree burial); and the Gaingang {Guayand), mound
burial. The ^''Tapuya^'' were accredited with endocannibalism. The
Bororo and Gaingang held mortuary festivals.
TECHNOLOGY AND MATERIAL CULTURE
All Marginal tribes had a primitive technology which, in both posi-

tive —
and negative traits, represented a very early certainly a prehorti-
cultural —American cultural level. More advanced technologies were
absent to a surprising degree, even among the tribes who adjoined or
formed enclaves within the Tropical Forest peoples and would seem to
have had considerable opportunity for borrowing. Some primitive
features were contingent upon a nomadic life, e. g., small temporary
shelters and the restricted number and size of material objects. Others
reflected the environment, e. g., the lack of canoes in areas without large
streams or the absence of agriculture in the far south. And a few
.
primitive features persisted where tribes, such as those of Tierra del

Fuego, were too remote from the sources of more advanced things.
Many primitive survivals, however, seem to represent a sheer cultural
inertia with respect to certain practical aspects of life, even where it
would have facilitated daily affairs to have borrowed new processes.
For example, simple woven and twilled baskets are easier and faster to
make than coiled or twined ones, loom weaving is more efficient than
netting or finger twining, and in many localities even dugout canoes
would have repaid the labor of constructing them. It is as if these
tribes had not yet reached the threshhold of paying serious attention to
material security; they exhibited a certain recklessness toward the
hazards of existence and the food scarcity that kept their population at
a low density.
On the other hand, the Marginal tribes laid considerable stress on the
more recreational aspects of life. Among the tribes in contact with
Tropical Forest or Andean peoples, personal adornment was richly
developed among certain groups, musical instruments were present in
considerable number, and most tribes had a fairly large number of
dances, songs, and games.
—
Fishing devices. Hunting weapons are mentioned below (p. 695)
Fishing was apparently done by all tribes with nets, harpoons, and
spears, and by all but those of the Archipelago with the bow. Fish-
hooks and fish drugs, though occurring sporadically, are not character-
istic of the Marginal peoples.
—
Food preparation. The Marginal tribes are distinctive even in
their methods of preparing food. A striking feature is their failure to
use salt; it is reported only among the Tehuelche and certain Chaco
tribes, who obtained it from the Andes. Even the Puri-Coroado^ who
had pepper, used no salt.
The tribes without pottery (see below) had no cooking containers
and consequently broiled or roasted their food. The babracot, or grill,
so characteristic of the Tropical Forest tribes, was found only among
the Caingang^ and the Mashacali, and some of the latter's neighbors.
Other tribes placed food directly on the fire, but some used the earth
oven, evidently an old, prepottery trait (the Chaco, Caingang, Ge^
Pwri-Goroada^ Botocudo^ and some Orinoco Marginals). Stone boil-
ing, which, like the earth oven, has North American parallels, is re-
ported for the Northwestern Ge.
—
Pottery. Pottery was absent among the Archipelagic tribes and
the Giboney, and it apparently reached the Tehuelche only recently,
perhaps in the historic period. (See map 3.) Its aboriginal presence
among the Northwest and Central Ge^ eastern Nambicuara, upper
Xingii, Pirahd, Mura, and Botocudo is doubtful even if made by these
;
tribes, it cannot have been of great importance. The remaining tribes

were unquestionably potters, but their ware was plain, incised, stick-
impressed, or cord-marked. Painted ceramics are accredited to only
the Huarp^ and the western Chaco, who borrowed it from the Andes,
and perhaps the Chiaharibo, to whom a "beautiful" ware is attributed.
—
Basketry. Basketry (map 21) had a limited distribution and was
90 75 60 45
^ T T T
15
15
30
45
'/Z/Z\ TmNED
Map 21. —Distribution of basket weaves.

;
probably introduced comparatively late among these tribes. It was

absent among the Puelche and Tehuelche, who used skin and wooden
containers, the Charrmi and most Chaco tribes, who used pots, and
the Botocudo and Oamacdn.
Coiling would seem to be a fairly old technique, to judge by its
occurrence among living tribes at the northern and southern extremes
of the hemisphere and its spotty archeological and ethnographic dis-
tribution between, but in South America it is the characteristic Andean
technique and ithas a ver}'^ restricted distribution elsewhere (map
21 ). In the Central Andes, it appears to be later than twined basketry
in Northern Chile, coiling is older than twining. Coiling was used
by the Archipelagic tribes, the Comechingon^ the Pancararu^ some of
the Northwestern and Central Ge, and possibly the Chaco {Mataco),
Twining, which has a more restricted hemisphere distribution, is re-
ported from Patagonia, the Huavpe^ the Carajd^ and the Shiriand.
Several Chaco tribes used it for mat making.
Baskets and often mats and fans were woven of a single palm frond.
These, evidently a primitive form of true woven and twilled baskets,
are accredited to the Mura, Guato, and Siriono. Twilled and woven
baskets, borrowed from the Tropical Forests, occurred mainly in east-
ern Brazil {Giiayaki^ Caingang^ Puri-Goroado Northwestern and
^
Central Ge, Carajd^ Tapirape) , and among the Taniro, Guafo, Bororo,
Bacdiri^ and CaingvA.
—
Fabrics. Fabrics were finger-made, either netted or twined, rather
than loom-woven. Thread was made of wild bast cotton was grown ;
only by some of the Ge and the Mashacali. Fish nets were made by
many of the tribes, and were the only Archipelagic textile product.
The Chaco, Carajd^ Caingang^ Patasho^ and Mashacali made netted
bags, and the Charrua^ the Orinoco Marginals, and perhaps the
Mashacali^ netted hammocks. The Botocudo^ Guayaki^ Bororo^
Yaruro^ and C-amacdn used also netting for various objects.
Twining was perhaps more common than is known, for textile tech-
niques are seldom described. It is accredited to the Chaco, Gv/ito^
Siriono^ and possibly some of the eastern Brazilian peoples.
Several tribes had adopted the true loom, but, as most Marginal
peoples either went nude, or wore skin garments, textiles were limited
to woven bands and a few skirts, and loom weaving was a minor craft.
The loom and true weaving were probably pre-Columbian among the
Huarpe^ Caingang^ Northern Cayapo^ Camacdn^ the Arawakans of the
upper Xingii, the Namhicuara^ Tapirape, Ghono^ and Siriono. The
Tehuelche seemingly adopted it from the Araucanians in the historic
period. Some western Chaco tribes had learned weaving with cotton
or wild bast from the Andes in prehistoric times, but they intensified
the craft after receiving sheep.
Weapons. —The Marginal peoples differed from the Tropical Forest

tribes in their almost complete lack of the blowgun and in the use
among and sling, though both groups
the southern tribes of the bolas
had bows. Some differences in weapons reflected the environment:
blowguns were used in the jungle, not in the more open plains; slings
and bolas are plains weapons arrowpoints were of stone in the plains
;
of the south {Gharrua to Tierra del Fuego), of wood farther north.

But there are nonenvironmental differences: virtually no Marginal
tribes used arrow poison (only the Guahiho and possibly Puelche) ;
their arrows did not have the Tropical Forest variety of specialized
points; and clubs were absent among the Gharrua and Patagonians.
The bow, though found throughout the hemisphere, is late in the
Fuegian sequence and it may have spread rapidly at the expense of
other weapons, particularly the dart and spear thrower. Outside the
Andes, the spear thrower had a sporadic occurrence, representing local
survivals. Among the Marginal tribes, it was limited to tribes near
the Parana Delta {Ghana, and possibly the Querandi), and to the
upper Xingii and Garajd. The pellet bow, more of a toy and probably
of European origin, occurred in the Chaco and among some eastern
Brazilian tribes.
A spear or lance was general. Among the horsemen {Tehuelche,
Puelche^ Gharrua^ and some Chaco tribes) it was developed to ex-
treme length and became a major weapon, even at the expense of the
bow.
The bolas is fairly old in the Fuegian shell mounds, though at the
Conquest extended from the Chaco south only to the Rio de la Plata.
it
With the horse, it spread farther south.

Clubs occurred from eastern Brazil to the Chaco, but were not
characteristic of the Southern Hunters.
The sling, a characteristic Andean trait, occurred in the Archipelago,
among the Ghari^ta^ and as a toy in the Chaco, but was absent among
other Marginal tribes, except the Giboney of the Antilles.
Defensive devices are obviously late acquisitions from the Andean
peoples. The Tehuelche and Puelche used skin -tunic armor and a
helmet the Chaco, a woven shirt or hide armor.
;
—
Dress and adornment. Except in Patagonia, Tierra del Fuego,
and among some Chaco tribes, where fur and skin cloaks provided
protection during cold winters, the body was adorned rather than
clothed. Adornment included painting, featherwork, ear, nose, and
lip ornaments, coiffure, headdress, necklaces, arm and leg bands, and
tattoo. Ear, nose, and lip ornaments were absent among the Archi-
pelagic peoples, and, therefore, are probably not among the oldest
South American culture traits. Elsewhere, these and other orna-
ments occurred in such variety that their types, distribution, and
738931 —49 46
696 SOUTH AMERICAN INDIANS [B. A. H. BuU. 143
history require special study. For the present purpose, it is of great-

est interest to observe that when these items were not purely orna-
mental, they served to distinguish tribes, sexes, societies, or other
associations rather than to indicate social status. Chieftains' in-
signia were rare, and class insignia were probably unknown.
RECREATIONAL ACTlVlTiJflS
—
Musical instruments. The Marginal peoples once lacked any
musical instruments, except perhaps a stick to beat time, the only in-
strument accredited to the Archipelagic peoples at the Conquest. No
instruments are reported for the Charrua. A rhythm beater in the
form of a bamboo stamping tube became characteristic of the Bororo^
Ge, and other tribes of eastern Brazil, and rattles, often of bark but
of gourd where there was some farming, were widely distributed, ex-
cept in the Archipelago. The Chaco evidently had only the rattle at
one time, but it borrowed a considerable number of other instruments,
especially wind instruments, from the Andes. Similarly, eastern

Brazil borrowed trumpets, panpipes, flutes, and other instruments
from the Tropical Forests. From the same source, the Guahibo ac-
quired rattles, flutes, and panpipes.
Intoxicating beverages. —
Chicha, a fermented alcoholic drink
made of various cultivated plants or wild fruits, was not originally
part of the Marginal culture. At the Conquest it was unknown to
the Archipelagic and plains tribes, except the Puelche and Huarpe.
Its principal occurrence was eastern Brazil: Caingang^ Ge, Tene-
tehara^ N ambicuara, upper Xingu tribes, Mura^ Tajnrape^ Carajd^
probably the PvH-Coroado, and other Marginals near the Brazilian
coast. Here, it was certainly borrowed from the Tropical Forests.
It was also made by the Orinoco Elver Marginals. The Chaco made
chicha of wild fruits, and the Guato and Bororo are accredited with a
palm wine.
—
Games. Games had a limited occurrence. Of special interest are
the Ge log races and the dice games (Andean), hockey, bull-roarer,
stilts, "snow-snake," and shuttlecock found among many of the tribes,
especially in the Chaco. (See p. 503.)

Tobacco and other narcotics. —Tobacco was comparatively recent
among the Marginals. It is reported only for the Bororo^ who smoked
cigars; the Chaco Indians, who chewed it or smoked it in elbow and
tubular pipes; and the Coroado and southern Amazon Marginals,
who usually smoked it in a pipe (map 9). Pipes have been found
archeologically in the Pampas, Patagonia, and Gaingang territory,
but none of these tribes used tobacco in historic times.
Other narcotics are equally rare among the Marginals. The Chaco
used a narcotic snuff, the Mashacali a vision- producing grub, the Mura
and Guahibo parica snuff, and the Pancararu a vision-producing drink,

borrowed from Tropical Forest neighbors.
all
THE TROPICAL FOREST AI^D SOUTHERN ANDEAN TRIBES

The Tropical Forest and southern Andean peoples differed from the
Circum-Caribbean and Central Andean tribes and resembled the Mar-
ginal peoples in having sociopolitical groups vehich were based on
kinship and lacked classes. Their groups, however, were character-
istically large, semipermanent villages, whereas the Marginal groups
were small and more or less nomadic. But each village was independ-
ent; there were no federations or empires, and, political units that
transcended the community were exceptional.
Both the Tropical Forest and Southern Andean peoples had loom-
woven textiles, fairly commodious houses, developed basketry, dugout
canoes or balsa rafts, and other material traits and technologies that
were characteristically absent or little developed among the Marginal
tribes. On the basis of the specific forms of these features and of
certain environmental adaptations, however, the Tropical Forest peo-
ples must be distinguished from those of the southern Andes. The
former, occupying the tropical rain forests of the Guianas, the Amazon,
eastern Brazil, the coast south to the Parana Delta, and the Paraguay
River, were linked with the Circum-Caribbean area. They had trop-
ical root crops, f rame-and-thatched houses, dugout canoes, twilled and
woven basketry, and the same general tradition of ceramics. The
southern Andean peoples were linked with the central Andes in that
they had potatoes, cereal crops, and agricultural techniques adapted
to desertsand semideserts, coiled baskets, some stone houses, balsa rafts,
metallurgy, llama breeding, Andean type clothing, and many minor
items derived from Peru and Bolivia.
Thus, two streams of historical influence are evident in the material
culture and technology of these tribes: one carrying central Andean
traits southward, especially after the Inca conquest, to Chile and
Northwest Argentina, and representing desert and Highland adapta-
tions the other, carrying Circum-Caribbean traits down the Atlantic
;
Coast and up great rivers, and representing coastal, riparian, and rain
forest adaptations. In the Chaco, the farming Arawakan Chane
{Ghana and Guand) in the north, and the natively riparian and later
equestrian AMpon, Mocovi, and perhaps Payagua and Mhayd of the
south and east were evidently influenced by both streams, and it is
possible that considerable Tropical Forest influence even reached the
Araucaniatis of Chile.
Because of the great difference in the element content of their cul-
tures, the Tropical Forest and the southern Andean tribes are treated
,
separately. Their sociopolitical patterns, however, were of the same

general type, though the latter approached sub -Andean types'.
THE TROPICAL FOREST TRIBES
The Tropical Forest culture, though comparatively uniform, had

As volume 3 of the Handbook carried a description
several subtypes.
of the general culture (pp. 886-895) and of the subtypes (pp. 896-899)
the Tropical Forests will be accorded less detail than was given the
Marginal groups.
Some of the subtypes, especially those of the western and southern
Amazon, represented gradations between the Tropical Forest and
Marginal cultures and they are classed as semi-Marginal (map 18).
Particularly, they lacked certain characteristic material and techno-
logical elements of the former and they had several distinctive social
and political features.
The Tropical Forest tribes had a fairly adequate subsistence, but it

was by no means based entirely on farming. These peoples were ripa-
rian and maritime, and water resources were no less important to them
than crops. In fact, population density was greatest on the coasts
and rivers.
After receiving the horse, the riparian Ahipon, Mhayd, and Mocovi
gave up what farming they had practiced to become looters or overlords
of farming tribes (Handbook, 1:250). The Goajiro turned from
farming to cattle raising.
Farming was by slash-and-burn in the rain forests, a method requir-
ing periodic shifting of the fields if not of the village. The most impor-
tant crop was bitter manioc (prepared with the tipiti or manioc
squeezer) which, however, was little grown in eastern Bolivia and
farther south. Also staples were sweet manioc sweetpotatoes {I'po-
: ;
moea hatatas) a species of yam {Dioscorea) Xanthosoina^ the New

; ;
World equivalent of taro arrowroot {Maranta arundinacea) maize,

; ;
often in many varieties; pepper {Capsicum) beans and squash. The

;
Tropical Forest people also cultivated various palms, gourds {Lagen-

aria vulgaris) calabashes {Orescentia cujete) such dyes as bixa (Bixa
, ,
orellana) and genipa {Genipa americana), cotton, toh^Q,Q,o {Nicotiana

tdbacum) and in various localities', arrow reeds, cane, palms, and other
,
special plants. Although some opinion tends to place the original

home of maize and perhaps other crops in eastern Bolivia or Mato
Grosso, the Tropical Forests cannot definitely claim the first domesti-
cation of any but the tropical root crops, which are distinctive of them.
The Tropical Forest peoples characteristically lived along the coasts
and rivers, and they were skilled boatmen and fishermen. The most
.
common craft was the dugout canoe, but the Island Carib^ expert
ocean navigators, used planked canoes some of the tribes on the smaller
;
tributaries of the great rivers, however, made only bark canoes, and
the eastern Chaco had bullboats. Fish, a major food resource, were
taken with drugs (p. 277) , baskets, nets, weirs, multiprong and harpoon
arrows, and, frequently, hooks. Important foods along the Amazon
and its main tributaries were turtles, turtle eggs, and large water mam-
mals (especially manatee {Trichechus inwnguis) and river dolphins),
the last taken with harpoons. On the coast, fish and shellfish were
eaten.
Land resources were also exploited. Wild plants (see Levi-Strauss,
Handbook, were collected with the aid of a great variety of
vol, 6)
baskets and, in many localities, the climbing ring (a loop held between
the feet and placed against a tree trunk to aid climbing) Game was .
taken in various kinds of traps (p. 265) or it was killed with the spear,
the bow, a great variety of special types of arrows (p. 229) and, in the ,
western portion of the area, the blowgun (map 4). Most of these
tribes kept Muscovy ducks.
These food resources not only supported a population that was
much denser than among the Marginal peoples, but, combined with
the transportational advantages of the canoe, they permitted people
to live together in larger and more permanent communities. Whereas
the Marginal band usually had 10 to 60 persons, the Tropical Forest
village consisted of hundreds and even 1,000 or more people welded
into a more or less cohesive unit. The village, moreover, could remain
in one place several years, but it was rarely permanent. Soil exhaus-
tion periodically required that it be moved to the site of a new planta-
tion, and many tribes customarily destroyed the village at the death
of one of its inhabitants (p. 354)
SOCIOPOLITICAI. AND EEUGIOUS PATTEBNS
The Tropical Forest village, like the Marginal band, usually con-
sisted of an extended family or lineage, i. e., several generations of
families related through either the male or the female line. In the
case of the Marginal peoples, a lineage ordinarily could not exceed
50 or 60 people before it had to split, a new lineage budding off to
become an independent band only an unusual ecology, such as that
;
based on the horse in Patagonia, permitted several lineages to remain

together in a single mixed band. Among the Tropical Forest tribes,
the greater resources permitted the lineage to become much larger
before a new lineage separated off, and the new one often set up its
own house not far from the parent lineage. Where several houses
made up a community, the arrangement varied with local circum-
stances. In some areas, they were tightly clustered around the chief's
700 SOUTH AMERICAN INDIANS [B. A.E.Bnll, 143
house and the men's club, and they were enclosed by a palisade the —
coastal Tupi village had up to 700 or 800 persons. In other places,
as along the Amazon River, they were closely spaced for a distance
of many miles. In either case, it would seem that the household
typically consisted of a lineage, had its own chief (the family head),
and retained considerable independence.
More rarely, as among the Apiacd, the maloca, or large house, accom-
modated several hundred individuals who evidently constituted several
kin groups.
Among the Tropical Forest people as among the Marginal tribes,
the distinction between the exogamous sib and the lineage or extended
family is often difficult to make. By definition, a sib or clan is exog-
amous regardless of whether the members live together and whether
their relationship can be traced. The fiction of relationship of its
members is usually supported by legends that trace their descent from
a mythical ancestor, and their cohesion is maintained by ceremonies
and other observances. Among the Tropical Forest peoples, the most
common sociopolitical unit was the extended patrilineal family,
occupying a single large house. Where the house stood alone the
group was exogamous by locality as well as by kin. In the northwest
Amazon, each exogamous household had ceremonies and myths, and
thus might be considered a localized sib. In other regions, however,
several extended households formed a village, so that there was house-
hold exogamy but not local exogamy, nor sibs properly speaking. In
short, there seems to have been a strong predisposition to patrilineal-
ity, which showed every gradation to true sibs, though lack of gene-
alogical and other data prevents their adequate classification.
It is logically possible that matrilineal, extended households might
have occurred as the counterpart of patrilineal ones. Actually, they
and matrilineal sibs are restricted to the Guianas, the Goajiro, and
the Patdngoro, and they occur doubtfully among the Ucayali River
Panoans and perhaps some upper Guapore River tribes.
Strong chieftainship and social classes were not characteristic of
the Tropical Forest tribes, because duties and obligations based on
kinship overshadowed those based on status. Where the village con-
sisted of a single lineage, the chief was merely the family head and
interpersonal relations were controlled by factors of age, sex, and rela-
tionship. In the strongly nucleated, multilineage villages, there was a
village chief, who usually succeeded his father and was often a shaman.
In such villages, the chief controlled some of the activities of people
who were not his relatives. This represents the beginning of civil
government, but the chief had authority only in limited contexts, such
as land use, warfare, or hunting. Among the Circum-Caribbean and
Andean tribes, the chief's status carried many special privileges ; the
ôL6] SOUTH AMERICAN CULTURES —STEWARD 701
Tropical Forest chief, however, merely served the community, and he

gained little personal advantage beyond a few extra wives and some
support from village members.
Common people frequently acquired some status through special
activities, especially warfare and hunting, but this rarely led to the
formation of a special class and never to hereditary rank. Where
villages were usually large and warfare was well developed, as in the
Guianas and among the coastal Tupi^ the chiefs had unusual power
and a semihereditary position, the warriors achieved fame, and the
war captives tended to form a loose slave class. But the classes were
not fixed: the slaves were fitted into the kin groups, usually inter-
marrying into their captive's family and being regarded as relatives,
and their children became freeman. The serflike status of the Macu
with respect to the Tucano was exceptional. The eastern Chaco, how-
ever, evidently had an aboriginal tendency to a class system. Here,
certain tribes carried on considerable warfare and accorded warriors
membership in a special class. After the Conquest, the acquisition
of the horse so intensified warfare and the spirit of conquest that
captives, both individuals and whole tribes, came to form a lower
class while the warriors became a definite upper class. These Chaco
tribes are not classed with the Sub- Andean peoples, however, because
the classes were largely of post-Columbian origin and because the
tribes lacked other diagnostic Sub-Andean features, such as the priest-
temple cult.
—
Types of kinship basis. The lineage seems to have been funda-
mental to most if not all of Tropical Forest society, and even in
multilineage villages each lineage probably retained its identity.
The Tropical Forests were overwhelmingly patrilineal, even where
the sib was the basis of society(map 20). Matrilineal descent is
recorded only in the Guianas and parts of Colombia among tribes
adjoining or near the Circum-Caribbean area. As the latter seems
to have been definitely matrilineal, the peripheral occurrences of
matrilineal descent may have been the result of Circum-Caribbean
influence which is postulated also to have introduced considerable ma-
terial culture and technology to the Tropical Forests. (See p. 726 ff.)
The native patrilineal basis is perhaps correlated with the importance
of hunting and fishing, both of which were masculine activities.
—
The patrilineal tribes. The extended, patrilineal household was
characteristic of the upper Amazon Tropical Forest (Panoan,
Tupian) and the semi-Marginal peoples {Zdparoans^ Petans, Western
Tucanoans, and some Panoans and Arawakans (Handbook, 3, pp. 507-
798) ) . Usually the village consisted of a single communal house, but
the Tupi on the Amazon and Maraiion Eivers had larger, multihouse
villages, and they and the Quip kept captive slaves. Tlie Panoan
tribes of the lower UcayaliKiver are accredited with totemic, matri-

but the data are not conclusive. The Arawakans of the
lineal clans,
neighboring Jurua-Purus region had animal-named patrilineal
households. Among the Tenetehara, the coastal Tupi, and the
Guarani^ the villages were extremely large and each house held 10
or more patrilineally related families (several hundred persons)
under a powerful chief who was usually the shaman. Intensive war-
fare contributed to Tupi village cohesion.
Tribes with patrilineal sibs, except the Palicur and Ghaco^ adjoined
those with the patrilineal lineage or household (map 20). In the
northwest Amazon, the Tucima had patrilineal, nonlocalized sibs de-
spite some matrilineal households. The Tucanoans and possibly the
Arawak and Carib of the Uaupes-Caqueta area and the Witotoans
had patrilineal sibs, each generally localized in a communal house.
The Witotoans were cannibals, and kept some captives as slaves. An
outstanding Tucanoan feature, perhaps shared in some degree by the
Witotoam.^ Tucana^ and Yurimagua, was the ancestor cult into which
boys were initiated with use of sacred trumpets. Each household or
localized sib held its own cult ceremonies. To the north of the north-
west Amazon, the Achagua and Saliva also had patrilineal, totemic
sibs, each perhaps occupying a communal house. Often one such house
constituted a village. Chieftainship was weak, and warfare was purely
defensive, lacking trophies, cannibalism, or slave taking. Sexual
cleavage is manifest in the use of a men's clubhouse, where drinking
bouts were held.
A second main center of patrilineal sibs was among the Twpian
tribes of the lower Madeira-Tapajoz region (Mundurucu, Parinfintin,
and Tupi-Cawahib). The Maue and lower Xingii Tupians, however,
had small villages consisting of a few single-family houses. The
Mundurucu settlements, though large, were dispersed, and each house
had its own chief.
The Choco also had exogamous, patrilineal lineages, which were

perhaps the equivalent of sibs. The chief had little power and warfare
was not developed.
—
The matrilineal trihes. The matrilineal tribes included several in
the Guianas and the Goajiro, Oagahd, Patdngoro, and Antillean Carib.
In the Guianas, nonlocalized matrilineal clans were found among the
coastal Arawak, the Wapishanu, and Aparai. Although society was
based on kinship, war captives formed a loose lower class among both
the Arawak and Carib, and Guiana Carib chiefs enhanced their pres-
tige by attaching a special following of freemen, usually sons-in-law,
to their households. The Aparai and Papishana, however, appear to
have coupled patrilocal and presumably exogamous settlements with
their matrilineal clans.
The Goajiro, whose aboriginal culture is virtually unknown, had

matrilineal sibs. In post-Columbian times, a cattle-horse complex
introduced a secondary series of social cleavages based on wealth. The
ChaJce {''^Motilones'^) south of the Goajiro had small shifting farm
villages. The Cdgaba, despite archeological evidence in their territory
of a more advanced culture, had a simple unstratified, matrilineal
society.
The Patdngoro^ like their Sub- Andean neighbors, had villages of 80
to 90 houses and a ceremonial structure, but the basis of their society
was matrilineal clans rather than classes. Their warfare was for the
purpose of obtaining victims for cannibalism and not slaves.
—
Other sociopolitical groups. Of the remaining tribes of the Tropi-
cal Forests, many were doubtless predominantly unilineal and other.s
may have been bilateral, but the basis of their descent is not now clear.
In eastern Colombia, the Guayupe and Sae lived in palisaded vil-
lages, each with a plaza and ceremonial building in the center. They
had an elective chief, who used a stool, wore feather blankets, and was
cremated at death, but other people acquired status with age. Boys
were initiated into warrior status, but war was waged to take victims
for cannibalism and human trophies, not to procure slaves. The
nearby Otomac and Guamo also had large villages, and the chiefs con-
trolled groups of houses. The Betoi village had one or more houses,
each sheltering an extended family and ruled by the elder. A unique
Airico feature was hired laborers.
The Gayapa and Colorado in western Ecuador had villages consist-
ing of one to several multifamily houses, but whether these sheltered
lineages is not clear.
In eastern Brazil, some of the small lion-Tupian tribes were on a
Tropical Forest level, but little is known of their sociopolitical groups.
The '"''Tapuya''^ had palisaded villages. The Camacdn communal
houses sheltered as many as 20 families. The Tarairiu were governed
by strong shaman-chiefs, who had special announcers when they died,
;
these chiefs were cremated and their ashes drunk, whereas deceased
commoners were cooked and eaten.
Farther south, around the Parana Delta, the warlike Guarani and
Timbu lived in permanent, palisaded villages, in contrast to the no-
madic. Marginal Charrua^ Queiândi, and other neighbors.
In eastern Bolivia, all tribes except the Sub-Andean Mojo, Baure,
Manasi, and Paressi (p. 714 f.) were probably Tropical Forest in their
general culture, though little is known of the composition of their
sociopolitical groups. Some had small communities and perhaps
should be classed with the Marginal peoples: each Atsahuaca and
Chapacuran family lived in its own lean-to the Chimane village con-
;
sisted of a few small houses; the Canichana hamlet was fortified and
contained only 60 persons. At the other extreme were tribes with

large communities and some Sub-Andean features. The Gayuvava
village had 1,800 to 2,000 persons, and as many as 7 villages were con-
trolled by a single chief. Each village had a temple where offerings
were made. The Xaray communities had 1,000 persons; four com-
munities were controlled by one chief. Perhaps the Cayuvava and
Xaray belong with the Sub-Andean tribes. Between these extremes,
the Turacare village consisted of one or several large houses, each
perhaps sheltering a lineage and governed by the family head. Simi-
lar villages are also reported for the Guarayu^ Pausema,^ Mosetene^
Tacana (100 to 200 persons per house), Tiatinagua (2 to 8 families
per house and each house under its own chief). Southeastern Panoans
(2 to 3 communal huts per village), Yamiaca, and Ghiquito (several
small houses and a men's club in each village, and a village chief and
elected council). The Ghiquito and Canichana were somewhat war-
like. Ghiquito captives married into the tribe; Canichana captives
were eaten or else became slaves, but such slavery seems not to have
created a true class system.
In the Gran Chaco, the western tribes belonged with the Marginal
peoples, but theMhayd^ Payagud^ Mocovi^ Abipon, and Guana should
certainly be classed with the Tropical Forest tribes. The kinship
basis of society is obscure, but to judge by postmarital residence, the
Arawahan Guana may have been somewhat patrilineal, like the east-
ern Bolivian tribes, and the others somewhat matrilineal, like the
western Chaco Marginal peoples. These tribes had an aboriginal
basis for a stratijfied society in their special warrior's class.
After the
Conquest, the acquisition of horses intensified their military activities
and true classes emerged. The Mhayd^ who were exceptional for
their wars of conquest, had the following social strata: (1) Several
kinds of nobles and chiefs with special privileges; (2) a warrior
class into which boys were initiated; and (3) serfs, consisting of con-
quered tribes, especially the Guana, and their descendants. Among the
Mocovi and Abipon, the privileged class of warriors became a class of
nobles, distinct from the commoners. The Payagud drew a distinc-
tion between chiefs, nobles, and commoners. The Tereno and Guana,
though victims of their more aggressive and powerful neighbors, had
a somewhat endogamous class of chiefs, a warrior group which
achieved its status, and a class of commoners, or "camp followers."
The Tereno also were divided into two endogamous moieties.
—
Warfare. Warfare was probably a major reason for the strongly
nucleated communities in much of the Tropical Forest area. Houses
were closely grouped and surrounded by palisades, additional protec-
tion being afforded by pitfalls, poisoned stakes planted in trails,
coltrops, and other devices.
;
Vol.51 SOUTH AMERICAN CULTURES —STEWARD 705
Whereas the Marginal tribes fought only in self defense or for

revenge, a large number of Tropical Forest peoples carried on warfare
to obtain human trophies and victims for cannibalistic feasts. In
addition, a few tribes, including several of the western and northwest
Amazon, some in the Guianas, and the horse bands of the Chaco, took
slaves. But true slaving was mainly a post-Columbian development
tribes which had formerly killed and eaten their victims and many
tribes which had never taken captives at all conducted raids to obtain
slaves for sale to the Whites.
Cannibalism, though widespread, was not practiced by all tribes.
The Tupi and Carib, who attached great importance to warfare, killed
and ate their captives, often after keeping them a long time. Unlike
some of the Sub- Andean peoples, they did not sacrifice their victims
in religious rites; ceremonial objectives, though not absent, were
secondary. Cannibalism seems to have been motivated by deep hatred
of enemy peoples, and there were unending reprisals against former
hostilities. Cannibalism was also practiced among the Tupi and some
of the Zaparoans of the upper Amazon, the Guiana Arawdk^ the
Witoto^ and other tribes of the northwest Amazon, the Patdngoro in
Colombia, the Canichana, in eastern Bolivia, and the Amniapa and
Guaratagaja of the Guapore River, but elsewhere it was absent.
Human trophies, which were displayed in order to give their takers
prestige and to insult the enemy, were of many kinds shrunken whole
;
heads (Montaiia) skulls (often made into cups) flayed skins {Arara)
, , ,
scalps, and flutes made of long bones (p. 409) Trophy taking tended
.
to be associated with cannibalism.

Crisis rites. — based essentially on kinship and lacking
În societies
public ceremonialism, crisis rites had great importance in underlining

sex differences, in marking age status, and in providing focal points for
ritual activity.
Birth customs involved many magical observances and taboos, and,
though they were primarily concerned with the mother and child, the
father was often deeply involved. (See The Couvade, p. 369.)
Pubescent girls were always subjected to certain rites, which, though
more or less concerned with their physiological condition, also signaled
a social transition to womanhood. The girl was isolated and, in dif-
ferent areas, whipped, made to dance, scarified, her hair cut, her
clitoris excised {Saliva)^ given drugs, and the like. In many tribes,
boys passed informally into manhood, their status being evidenced not
only by dress and behavior but often by their occupation of the men's
clubhouse. Frequently, however, their hair was cut, their noses pierced,
or some other public rite celebrated, and in some instances, they were
initiated into a secret society or cult, as mentioned below. The Saliva
boy was circumcised.
In contrast to the Marginal tribes, who almost always buried their

dead, the Tropical Forest peoples disposed of the body in many ways,
but they rarely accorded chiefs or leading men special treatment.
Only the Guayupe and Sae^ both doubtless under Andean influence,
gave chiefs burial different from other people.
Direct earth burial was the principal Tropical Forest method, but
many tribes subsequently exhumed the bones and reburied them in
the earth {Mundurucu, Apiacd, Ahipon, Mocovi) in a cave {Otomac)
, ,
or in urns (upper Amazon Tupi, Goajiro, some Island Carib, Guaranty

coastal Tupi^ and probably many other tribes) or they preserved the
,
bones in the house or elsewhere ( Jurua-Puriis Arawak) The Atorai

.
and Rucuyen cremated their dead, and several tribes drank the cre-
mated ashes in chicha (Montana, /Saliva, Island Carib in the case of
chiefs, Arapimn, Guayupe, and Sae). The Montaiia, Tarairiu, and
Araîum also ate the corpse. The Chahe exposed the body and later
made a bundle of the bones.
The practice of mummification, which in the Andes is associated
with strong ancestor worship, is accredited to only the Piaroa of the
upper Orinoco. Many Tropical Forest tribes, however, had what
amounted to a cult of the dead for exam^^le, the northwest Amazon
;
and Guianas, where there were public mourning ceremonies.

—
Religion and shamanism. Tropical Forest religion resembled that
of the Marginal peoples in that it centered in the shaman, who had
his own supernatural helper and whose main functions were to cure,
prognosticate, and perform magic. The Circum-Caribbean and
Andean pattern was slightly foreshadowed in that the shaman some-
times performed oracular functions, but he probably consulted his own
spirit helper rather than a tribal god. In fact, true tribal gods were
rare and there were no temples or idols. Celestial beings, which, in
the Circum-Caribbean and Andean areas, were frequently tribal deities
and the objects of public ceremonialism, were usually myth characters.
Several Tropical Forest tribes, however, had public ceremonials, and
a few of these involved tribal gods and spirits, though most of them
were secret society rites.
There were considerable local differences in religion, owing to the
limited distribution of secret societies, cults, ritual elements, public
and particular concepts of gods and spirits.
festivals,
—
Cult religion. Cult religion is exemplified by the sib ancestor cult
found in the northwest Amazon (Handbook, 3: 885) and perhaps
among the Achagua and Saliva (Handbook, 4: 410). This, how-
ever, was as much a secret society as a tribal cult boys were initiated
:
to membership and to the secret of certain sacred religious parapher-

nalia that were taboo to women and children. The Jurua-Purus feast
of nature spirits, though providing public festivals, also involved sex
distinctions in that the sacred musical instruments were taboo to

women and children. Even the Tacancm priest-temple-idol cult
(Handbook, 3: 447), though somewhat Andean in general features,
seems really to have been a boys' initiation rite, women being excluded.
Public ceremonies. —TheTupian Grandfather cult, with its mes-
sianic features and hope of a better life (Handbook, 3 131) was a true
: ,
tribal religion. The Saliva ceremonies in honor of the creator (Hand-

book, 4: 410), though in a wholly different pattern, were also tribal
rites. So were the harvest ceremonies of the lower Amazon {Tupi-
namha, Tenetehara, Maue, Mundurucu)^ the Guarani, the Tarairiu,
and the Chake, the fertility rites of the Manao and the Jurua-Purus
Panoans^ the first fish ceremonies of the Achagua and Saliva^ the
Pleiades festival of the Chaco, the bush spirit ceremonies of the Tupi-
Mundurucu^ and Guaranty the ghost rituals of the lower Xingii
narriba,
and the mourning ceremonies of the Guianas and Northwest
tribes,
Amazon.
Supernatural heings. —The principal deities tended to be myth char-
acters rather than objects of religious worship, except that among
the Tupian Guarani, Tupinamha, Parintintin, and Apiacd they entered
something of a cult. The main gods or mythological beings were most
often the sun, moon, stars, and thunder, and it is easy to see in them
prototypes of some of the major Andean
tribal gods. Gods of moun-
and ocean, and place spirits generally were much
tains, springs, rocks,
less important than in the Andes. There was, however, belief in a
multitude of bush and other nature spirits, and many of them were
revered and feared. Ghosts, too, were commonly of religious im-
portance, especially as shamans' helpers.
Sham/mism. — Shamanism of the Tropical Forests was in the general
pattern of the Marginal peoples. The shaman obtained a special spirit
helper. Little pertinent informationis available, but it seems usually
to have been a nature spirit or human ghost (for example a tree spirit,
:
Tucuna; a jaguar, Witotoans., Tuxianoans; a dead shaman, Jivaro).

The werejaguar concept was widespread. The shaman's principal
function was to cure disease, which he did with the aid of a drug
or narcotic, especially tobacco but locally with other plants (this
volume, pp. 588-599), and by massaging and sucking out an object-
believed to have caused the sickness but sometimes by exorcising evil
spirits or by recovering the patient's "lost soul." The shaman also
practiced black magic, prognosticated, controlled weather, and, where
there were group ceremonials, he generally officiated. In fact, his
supernatural power was often so great as to give him special social
status, and, in many tribes, great political power. Among the Tupi^
he was often the chief.
Ritual elements. —
^Many ritual elements, such as the ant ordeal, snuff
taking, gashing and bleeding, circumcision, scarification, flogging, use
of sacred trumpets, cutting the hair, use of the scratching stick, and
the like were, very widespread, but they entered different local contexts.
For example, the ant ordeal was used by the Rucuyen and some of the
Tupi south of the Amazon to test young boys, by the Andaqui to try
out warriors before battle, by the Saliva to induct new chiefs, by the
Guiana Arawdk as a feature of death rites, and by other Guiana tribes
in hunting ritual. Wliipping was used by the Macushi and Montaiia
tribes for pubescent girls, by the Chiaywpe and Sae to make boys into
warriors, by the Turimagua while initiating boys to the tribal cult, by
the Manao young boys, and by the Choke {Motilones) for par-
for
The scratching stick was employed
ticipants in the harvest festival.
by Tucanoan and Choco girls during their isolation at puberty.
TECHNOLOGY AND MATEBIAL CTJLTtTRE
The Tropical Forest peoples had a far richer technology and ma-
terial culturethan the Marginal tribes. These features were found
also in the Andean civilization, which, however, elaborated them far
beyond the Tropical Forest. The Tropical Forests did true loom
weaving with cotton, but used no wool and lacked the fine and intri-
cate Andean weaves. They often made good painted or modeled
pottery but fell far short of the Andes in number of colors and in-
tricacy and variety of design. The Tropical Forests entirely lacked
metallurgy, stone construction, and stone sculpture, the last two pre-
cluded by the absence of stone.
On the other hand, the Tropical Forests had a number of traits not
found in the Andes frame houses, hammocks, mosquito shelters, the
:
blowgun, pepper pot, rubber, bark cloth, dugout canoe, climbing ring,
hollow-log drum, babracot, and poisoned arrow. Some of these had
only a partial distribution in the Tropical Forests, but many extended
to the Antilles and Central America.
—
Basketry. The characteristic basketry type is woven, usually
twilled, but in many places it is also hexagonal and of wicker. Some
tribes such as certain of the Tupi south of the Amazon {Parintintin,
Tupi-Cawahib) however, used only a simple twilled basket woven
,
of a single palm leaf. The Guianas had the greatest variety of weaves
and basketry types. Twining is found only in the northwest Amazon
and in one or two east Bolivian tribes (map 21), and coiling, which
is essentially Andean, only among the Choco.
— —
Weaving. True loom weaving a simple technique with the weft
—
passed in and out between the warps was widespread, extending from
the Guianas to the upper Amazon and south to the Tupi-Gawahih^
Guarani^ and Mhayd. Many tribes, however, only wove bands, be-
cause large fabrics hadlittle use in a climate in which people generally

wore few garments. The main yarn was cotton, usually domesticated
though sometimes wild, but some tribes used bast of wild plants.
Designs were largely limited to woven-in stripes, but tribes bordering
the Andes painted designs on their textiles.
The Tucanoans and most of the Tupi tribes had a twined or finger-
—
weave without a loom a technique similar to that of twined basketry
in which a pair of wefts enclose each warp element, being given one
turn or twist before enclosing the next.
In addition, netting occurred in the Guianas, northwest Venezuela,
and eastern Bolivia, and among the Achagua and Saliva (who had no
true loom).
Ceramics. —Pottery was made by all these tribes, but the wares
varied greatly. They were plain on the Jurua-Purus and in north-
western Venezuela. Elaborately modeled wares occurred in the lower
Amazon and may be linked with the Cirum-Caribbean area. Painted
vessels were found in the same area as the last and also up the tribu-
taries of the Amazon and south among the Tupi-Guarani.
—
Houses and household goods. The Tropical Forest house, though
varying greatly in detail, was typically a frame, thatch-covered
dwelling which accommodated several families and often had con-
siderable size. The eastern Chaco horsemen, however, used portable,
mat-covered dwellings. The house was furnished with wooden stools,
hammocks (woven, twined, or netted of various materials, according
to the locality), mosquito shelters (Montaiia, Otomac), babracots,
mortars and and storage baskets. The hammock has spread
pestles,
rapidly in the historic period, and the platform bed, a trait of the
eastern Andes, has survived only in a few tribes bordering the Andes.
—
Weapons. The bow and arrow occurred among practically all
Tropical Forest tribes. The bow often had great length, and arrows
were characteristically equipped with a large variety of wooden
—
points multiprongs, barbed rods, lanceolate blades, knobs, harpoon
heads, and other forms for special purposes (this volume, p. 235 ff.) —
and were frequently poisoned {Achagua^ Saliva, Patdngoro, Lucalia^
Guiana Carih, eastern Bolivia) Stone points, though typical of the
.
Marginal hunting tribes of the south, were not found, owing mainly
to the scarcity or lack of stone in the forests.
—
The spear thrower survived in a few areas ^Montana, Jurua-Purus
—
Arawah (pp. 244-247) but was evidently being replaced by the bow, a
process that continued into the historic period (Handbook, 3: 526).
The blowgun, an early Andean trait and possibly of trans-Pacific
origin, was limited to the western Amazon and extended across the
northern Andes to Central America (pp. 248-252; map 4). It con-
. ;.
tinued to spread during the historic period, a condition of its diffusion

probably being increased intertribal contacts and trade of poison with-
out which it is worthless. Its value lay in providing a silent, highly ef-
fective means of hunting arboreal game. There seem to be few records
of its use in warfare, even by tribes which did not hesitate to kill their
enemies with poisoned arrows.

Shields occur among some Tropical Forest tribes (p. 259 f .
; map 5)
Bark cloth.— Bark cloth, a trait of possible trans-Pacific origin,
had a distribution very similar to that of the blowgun (pp. 67-68).
Among many tribes which practiced loom- weaving, it was used pri-
marily for religious masks and as an alternate for cloth. Among
others, such as the Choco, it was used instead of cloth.
—
Metallurgy. None of the Tropical Forest tribes smelted or cast
metal in pre-Columbian times, but objects of gold, silver, and some-
times copper from the Andes had reached a great number of them,
many a considerable distance away; for example, the Xaray and
Guarani of the Parana and Paraguay Elvers and the tribes of the
middle Amazon.
—
Rubber. Many but not all the Tropical Forests shared a knowledge
of rubber working with Central America and Mexico (p. 227). De-
spite its present-day importance, rubber was of minor value to the
natives, being used only to make balls, syringes, rings, and figurines.
—
Musical instruments. The Tropical Forests had a considerable
number of musical instruments, some of special interest. Gourd rat-
tles, which were very widespread, were generally shamans' instruments.
Trumpets tended to be associated with a secret cult, in which they
represented the voice of the deities and were taboo to the unitiated
(p. 576 f .) but the Otomac used them in lay festivals and the Saliva in
,
funeral ceremonies. Hollow-log signal drums, possibly of trans-

Pacific origin,had a distribution very similar to that of bark cloth
and the blowgun (Central America, Colombia, Witotoans, western
Guiana, the Betoi and their neighbors, Achagua, Choco) The pan-
.
pipes, perhaps also of trans-Pacific origin, were even more widespread

than the hollow-log drum, and they occurred also in all periods of
Andean history. Among other instruments were various flutes,
jingles,and stamping tubes to mark time.
—
Narcotics and stimulants, Chicha was common to all Tropical
Forest tribes, who drank it in quantity on festival occasions. An Ilex
was drunk as mate in Paraguay and eastern Brazil, and another Ileso
as guayusa in the western Amazon (map 10) . Tobacco was the main
narcotic, being used by all tribes except some of the Tupi (p. 525 f .
map 9). Other narcotics had a more restricted distribution, and

occurred especially in the northwest Amazon (map 15)
THE SOUTHERN ANDEAN TRIBES

The principle southern Andean tribes were the Ku/nza-speaking
Atacameno (Handbook 2:41, 599), the ââTi-speaking Diagmta or
Calchaqui (Handbook 2: 39-40, 633-637), who occupied the extreme
deserts, especially those west of the Andes, and the Araucanians
(Handbook, 2 42-44, 687) who lived in the temperate zone of winter
: ,
rainfall in Central Chile.

Many features of the cultures of these tribes were of unquestioned
central Andean origin ; the agricultural complex, llama raising, guinea
pigs, dogs,wool and cotton weaving, metallurgy, ceramic styles, slings,
quipus, garment types, blood sacrifice, stone houses, stone cist graves,
and mummification. The Araucanians, who experienced somewhat
less Andean influence than the Diaguita or Atacameno, had a number
of traits that linked them with the Tropical Forest tribes: frame,
thatched houses, mortars, dugout canoes, cannibalism (including
drinking cremated ashes with chicha), human trophies, sporting
games, and urn burial. In all three tribes, however, the sociopolitical
units were in the Tropical Forest rather than Andean pattern.
These tribes were intensive farmers and their basic crops maize, —
quinoa, beans, and potatoes —
were Andean. In the Chaco Santia-
gueno (Handbook, 2:655) and La Candelaria (Handbook, 2:661),
east of the Andes, flood-plain farming was practiced, but in the ex-
treme deserts to the west irrigation was necessary. Irrigation and
farm terracing, both Central Andean traits, extended to the Quebrada
de Humuhuaca in northwest Argentina (Handbook, 2: 619) and to
the northern Araucanians.
On the coast, fish were taken in some quantity by means of hooks,
drugs, weirs, spears, baskets, and harpoons, but habitation sites were
limited to the few sheltered coves where there was fresh water. The
central coast of Chile was far more favorable for occupation than the
north coast.
Transportation was well developed. In the northern part of the
area, pack llamas were driven over roads and balsa canoes were used
along the coast on the south coast, dugout and planked canoes were
;
employed.
THE SOCIOPOLITICAL AND RELIGIOUS PATTEENS
Sociopolitical patterns. —
Despite this fairly efficient teclinological
population of the Atacameilo deserts was very sparse and the
basis, the
people lived in widely separated, fortified villages, each consisting
only of a few one-family stone houses. The village apparently was
made up of related families under a patrilineal chief and it was
738931—49 47
totemic ; perhaps might be regarded as a patrilineal lineage or sib.

it
The Atacameno seemingly had no warrior class, though they carried on

some warfare, and there is no evidence that they accorded chiefs special
privileges or special burial. They lacked a temple-priest cult. In
the area of the Puna de Atacama and the Quebrada de Humuhuaca,
evidence of temples and idols is also absent, and the archeological
remains of villages and graves are basically like those of the Ata-
caTiieno. Warfare in these areas was carried on for individual honors,
and trophy skulls were kept.
The Argentine Dlaguita^ also living in an arid area, had towns simi-
lar to but probably somewhat larger than those of the Atacameno.
The destruction of a house at the death of an occupant, however, is
evidence of some settlement instability. The Diaguita may have been
divided into lineages, but political consolidation probably extended
beyond kin groups and incipient class structure is evident. Village
chiefs, whose succession was patrilineal, sometimes contracted war
alliances with other chiefs. Chiefs may have received special burial
in Chile, certain especially elaborate archeological graves contain the
bodies of several persons. The Diaguita used large armies in their
wars against the Spaniards. Their warriors achieved special status
and may have formed a distinct class. They took and tortured cap-
tives and made head trophies, but they were not cannibals. The
Diaguita had no temple cult and probably no class of priests, though
special persons seem to have conducted agricultural rites.
The Araucanians had a much more fertile habitat and a much denser
population than the Atacameno or Diaguita. The communities were
not larger, however; instead, the population was simply divided into a
greater number of comparative small communities, which were struc-
tured on a kinship basis in the Tropical Forest pattern. Each settle-
ment consisted of a single house, which, reputedly for fear of witch-
craft, was located some distance from its neighbor and was often
palisaded or fortified. It was occupied by 3 to 8 individual families
—
15 to 40 persons who apparently constituted a totemic, patrilocal, and
probably patrilineal lineage, i. e., what is sometimes called a localized
sib. The lineage had its own chief and apparently owned its farm land.
The lineage or settlement seems to have been comparatively inde-
pendent, but several related lineages sometimes constituted a larger,
and the several units of a district in turn may have
less cohesive unit,
been united into loose tribelets under chiefs with only slight power
and nominal titles.
Araucanian society, though based on kinship and essentially demo-
cratic, had incipient classes ( 1 ) wealthy men and chiefs, who, how-
:
ever, had no great privileges except that of polygyny and who did not
receive special burial; (2) commoners; and (3) slaves and captives,
who did not constitute an important labor class. Wealth and military
exploits were the means of acquiring status. Today, with an economy
based on sheep, horses, and mules as well as on cultivation, the rank is
more strictly economic: the wealthy, the commoners, and the poor
peons who work on the estates.
The Araiuianians^ always outstanding warriors, not only took slaves,
but they captured sacrificial victims. The victims were tortured and
killed for cannibalistic but not ritualistic purposes. Their flesh was
cooked and eaten, their hearts sucked by headmen, and their bones
burned and the ashes drunk in chicha. Sometimes their skulls were
made into trophy cups, their long bones into flutes, and their facial
skin into masks.
—
Religion and shamanism. There is little evidence of a true temple-
idol-priestcomplex in the southern Andes, but central Andean influ-
ence is seen in public ceremonialism that was connected with tribal
gods and that was held probably according to a ritual, agricultural
calendar.
The Diaguita held agricultural ceremonies, and the Atacwmeno
worshiped the sun and had fertility rites for their fields. The most
important public Araucanian ceremony was conducted by priests who
supplicated the Supreme Being and Creator with prayers and blood
sacrifice of animals to send good crops and health. The Araucanians
also had a cult connected with a deity responsible for natural catas-
trophes, and they appealed to various spirits, including the sun.
There is some question whether all these public ceremonies were
entirely Andean in origin, but there is little doubt that many of the
ritual elements were. The latter included: llama, and later sheep,
sacrifice,and extraction of the heart blood; aspersions of blood and
chicha and four sacred directions.
;
Shamans obtained their supernatural power from deceased shamans

and from the Supreme Being, to whom they prayed. The Araucanian
shaman, like that of the northwest of North America, was often
a transvestite and used a tambourine. He prognosticated, made rain,
and cured disease. Disease was believed to be caused by the intrusion
of foreign objects, by soul loss, and by magic.
—
Disposal of the dead. Interment of the dead was the rule, but
mummification and stone-cist graves, both Andean traits, were not
uncommon. Urn burial occurs archeologically in Araucanian terri-
tory and in the area of La Candelaria and Chaco Santiagueiio.
TECHNOLOGY AND MATERIAL CTJLTUBE
Pottery. — Pottery was made by

southern Andean tribes but that
all
of the Araucanians was mostly undecorated and crude whereas the
Diaguita wares were bichrome and the Atacamefio were trichrome.
—
Basketry. The earliest basketry in archeological sites was coiled,
but twining was introduced at a later period and a twilled weave was
finally adopted by the Diaguita and Araucanians.
—
Weaving. ^Weaving was done on the true loom M'ith llama and
alpaca wool as well as cotton. The Diaguita also used vicuna wool.
Textiles had woven-in designs in color.
—
Metallurgy. Copper and some gold and silver were smelted and
worked by the Atacameno and Diaguita^ the latter also making bronze
and doing some casting. All three metals were used in native times
by the Araucanians^ but it is doubtful whether they were mined or
smelted. To the east, metallurgy did not extend to the Chaco San-
tiagueno.
Weapons. —Andean weapons present among these tribes were
slings, lances, bronze knuckle-dusters {Diagidta^ Atacameno)^ stone-
head clubs {Atacameno^ Humuhuaca, Diaguita) bronze axes, star-head
,
copper clubs {Diaguita) shields, helmets, hide-cloak armor, and spear

,
throwers {Araucanians, Diaguita). More characteristic, however,

were bows and arrows (poisoned only in the Chaco Santiagueno area)
and bolas {Atacameno, Diaguita doubtfully, Araucanian, post-Colum-
bian). Lassos among the Araucanians were post-Columbian.
—
Dress. Andean-type shirts, dresses, and mantles were common,
especially after the Inca conquest.
—
Musical instruments. The musical instruments were panpipes,
end-flutes, rattles, trumpets, drums, bells, and ocarinas.
—
Stimulants. All these tribes probably made chicha.
THE SUB-ANDEAN AND CIRCUM-CARIBBEAN TRIBES

The fourth volume of the Handbook describes the tribes whose dis-
tinctive sociopolitical and religious patterns were called Circum-
Caribbean or Sub-Andean. These tribes were distributed from the An-
tilles across the northern coasts and northern Andes of South America
and through Central America to the Maya frontier in Honduras.

With the data of all volumes of the Handbook available for a more de-
tailed comparison of the cultures, it is now evident that the Circum-
Caribbean tribes of Volume 4 were somewhat interspersed with Tropi-
cal Forest peoples, while several tribes, especially those of the Ecua-
dorian and Colombian Andes belong with the Sub-Andean rather than
Central Andean cultures. It is also clear that several tribes of eastern
Bolivia, though included with the Tropical Forest peoples in Volume
3, really had the Sub-Andean patterns.
In the following classification, all tribes are called "Circum-Carib-
bean," but the designation of "Northern Andean" is used to dis-
tinguish certain tribes of Ecuador and Colombia from the other Cir-
cum-Caribbean peoples, because they had a considerable number of
. ,
specifically Central Andean culture elements (below). Classed as

Circum-Caribbean are the following: The peoples of Central America
south of the Maya frontier (Handbook, 4:26-28), of the northern
Colombian Lowlands (4 329-338) of the Highlands of western Vene-
: ,
zuela (4: 469) of portions of northern Venezuela (4: 22-23) and the
, ,
Arawak of the Antilles (4: 507-546). The additional name, "Sub-

Andean," is applied to the Chibcha (Handbook, 2: 887-909), the
peoples of the southern Colombia Highlands (2 915-960) of Coast
: ,
and Highlands tribes of Ecuador (2 785-821) and, in eastern Bolivia,

: ,
the Mojo and Baure (Handbook, 3 408-424) the Paressi (3 349-360)

: , :
the MaTMsi (3 389-393) and the Paunaca (3 396-397)

: , :
The subsistence of the Sub-Andean and Circum-Caribbean tribes

differed considerably in the different areas. In general, it was based
on fairly intensive farming, but on the sea coasts and in the Antilles,
ocean resources were added to agricultural produce. The population
was fairly dense, and villages were generally large and permanent.
The Circum-Caribbean and Sub-Andean village resembled some
of the Tropical Forest communities in consisting of several houses,
each with an extended family or lineage, but the bonds between kin
groups were reinforced by stronger political controls and society was
characteristically divided into somewhat fluid classes. A group of
hereditary chiefs constituted the upper class. The nobility seems to
have developed from a who attained rank through
class of warriors
the display of human trophies and the capture of women for wives and
men for cannibalistic feasts and for occasional religious sacrifice.
Tribal gods were represented by idols kept in special temples, and they
were served by priests. Few tribes had a special hereditary class of
priests, however instead, shamans performed priestly functions, thus
;
acquiring social prestige and political power. The temple rites tended
to be oracular seances for the benefit of individuals rather than com-
munal ceremonies for the public good.
Politically, many tribes had achieved multicommunity states, both
through federation and conquest, but they never succeeded in consoli-
dating them against revolt in a stable empire like that of the Inca.
Political consolidation and economic exploitation of conquered masses
of people had not advanced sufficiently to supersede a slave class made
up of captured individuals.
Technologies and material culture, like those of the Tropical Forests,
included pottery, loom weaving, domesticated cotton, woven basketry,
dugout canoes, frame and thatch houses, palisaded villages, hammocks,
and various other items adapted to tropical or subtropical regions.
They also resembled those of the Andes in the use of metals, salt, the
litter, guinea pigs, and other elements which were more numerous in
Colombia and Ecuador than in Central America or the Antilles. (See

below.) Material goods were produced in sufficient quantity to allow

local specialization and trade, especially in such items as metals, salt,
and shell beads.
cloth, pearls,
The Sub-Andean or Northern Andean tribes of Colombia differed
from the Circum-Caribbean peoples of Venezuela and the Antilles in
possessing such characteristically Andean culture traits as copper
smelting and alloying of copper with gold; construction of roads,
bridges, hilltop forts, and stone houses; working of salt; balance
scales; quipus; maces; gold pincers; coca; war banners; the practice
of brother-sister marriage; mountain worship; and wrapping of
funeral bundles. Some of these features extend through Central
America to Mexico.
The Ghibcha of Colombia were previously classed with the Central
Andean tribes because it has always been supposed that they loomed
above their neighbors in political and religious achievements. Now
that the Conquest Period culture of the other Colombian tribes is bet-
ter known, it is clear that the Chibcha were not very distinctive.
They had no urban centers, and their political expansion was little
if any advanced beyond that of several other Northern Andean tribes.
Their temple cult was patterned on oracular performances for indi-
vidual ends rather than public worship according to a ritual cycle,
and their hereditary priesthood, though unusual, was not unique.
Their technology and material culture was not outstanding, even by
Sub- Andean standards.
Ecuador was incorporated into the Inca Empire about 1495, some
40 years before the Spanish Conquest, but in pre-7nca times it was
definitely Sub- Andean. The effect of the Inca conquest was to make
the native social stratification more rigid and more hereditary, to im-
pose strong political controls which stemmed downward from the Inca,
Emperor, to establish a number of truly urban centers, and to intro-
duce the Quechua language and such material items as new crops, more
extensive irrigation works, llama herding, weaving with wool, in-
creased building in stone, cultivation of coca, and new ceramic styles.
The Sub- Andean tribes of eastern Bolivia form enclaves within
more backward peoples, as if they had been compressed within migrat-
ing groups of Tropical Forest and Marginal tribes. Their material
culture, moreover, is much like that of the Tropical Forests, and it
lacks metallurgy entirely. Even the Arawukan language, which all
but the Manasi speak, has affiliation with the Amazon, Guianas, and
Antilles rather than with the Highlands, and but for their class system
and temple-idol cult, they should be classed with the Tropical Forest
peoples.
Because the Sub-Andean tribes of eastern Bolivia are widely sep-
:
arated from those of the Circum- Caribbean and Northern Andean

areas, they will be described separately.
In the following pages, the culture of these tribes is projected against
that of the Tropical Forests and Central Andes, The historical sig-
nificance of the many features found also in Mexico is discussed sub-
sequently.
THE CIRCUM-CARIBBEAN AND NORTHERN ANDEAN TRIBES

Intensive farming and favorable environments helped support some-

what larger and more permanent villages than in the Tropical Forests.
Whereas the Tropical Forest peoples cultivated migratory slash-and-
burn plantations in the rain forests or the gallery forests, the Sub-
Andean tribes of the Highland farmed in more open country, and even
the lowland peoples frequently inhabited and cultivated savannas.
Irrigation, an essential and characteristic practice in the Peruvian and
Chilean deserts, though unnecessary in the rain forests, occurred
sporadically : Ccmari^ the Chibcha or their neighbors, the Timoteoms
(who had storage tanks), the tribes near Barquisimeto in Vene-
also
zuela, Cumanogoto^ the Island Arawak, and probably other tribes, the
determination of which will depend upon archeology.
Some indication of the development of horticulture in the Northern
Andes of Ecuador and Colombia is afforded by the number of crops
in the high altitudes, above 8,000 feet, potatoes, quinoa, ullucos {Ul-
lucus tuherosus), and cubios {Troyaeolwrn tuberosum) ; middle alti-
tudes, 5-8,000 feet, maize, arracacha (Arracacia esculenta), quinoa
{Chenopodium quinoa), achira {Canna eduUs), and auyama {Gucur-
and, somewhat in the middle altitudes but more in
iita verracosa) ;
the lowlands, sweetpotatoes, maize, yuca or sweet (but not bitter)

manioc, beans, peanuts, and a considerable number of fruits including
pineapples, avocados, papayas, custard apples {Amiona squamosa),
chirimoyas {AnTWna cherimolia) and capuli {Physails peruviana).
,
Additional crops were peppers (aji) cotton and/or maguey and cabuya
,
grown for fibers, altramuces {Lupinus sp.), and various grasses.

Most of these crops also occurred to the north, throughout Central
America, but the number dropped off sharply in Venezuela, where
bitter manioc appeared. Bitter manioc apparently spread in post-
Colombian times to the Antillean Arawak and to Central America.
The Antillean Arawak grew only potatoes, peanuts, beans, arrow-
root, and a primitive variety of maize.
An indication of the relative importance of farming is that it was
mainly woman's chore in the Tropical Forests but among many Cir-
cum-Caribbean and Sub-Andean tribes, as in the Central Andes, men
also were cultivators.
In the lowland and coastal areas of both the continent and the
Antilles, however, seaand river resources compensated for the smaller
number of crops, and the population density was extremely great.
Llamas became important in Ecuador only after the Inca conquest
and they never reached Colombia, but the guinea pig extended
throughout the Northern Andes, probably to most of Colombia. The
Muscovy duck and perhaps the dog were found in all Northern Andean
and Circum-Caribbean tribes. The chroniclers mention a curious
"mute dog" among the Ahurrd of Columbia and the Antillean Arawak,
but it may have been a tame bush dog {Icticyon venaticus) or a fox.
The domesticated turkey may have been kept by some Central Amer-
ican and Venezuelan tribes. The Antillean Aratoak and perhaps the
Columbian tribes had "domesticated" parrots. (See Gilmore, Hand-
book, vol. 6.)
—
Settlement pattern and composition. The population density was
very great, being exceeded only by that of the Central Andes, but the
settlement pattern, though imperfectly known, seems to have varied.
Among the Highland tribes of Ecuador and southern Colombia, the
—
dwellings one-family pole-and-mud houses (Ecuador and probably
Uhibcha) or frame houses (Pdes, Moguex, Pijao) were somewhat —
scattered, and their occupants affiliated with one or another religious
or political center. Compact, or nucleated, towns were exceptional.
For Examples of these dis-
defense, people retreated to hilltop forts.
persed settlements are the Cara, who had but few large towns; the
Puruhd, whose villages consisted of only 10 to 12 houses the Palta, ;
who lived in isolated houses but assembled on occasion at a ceremonial

center or at a chief's house; and the Pdez, whose village consisted of
several polygynous families living in houses clustered around the
residence of the village chief.
In the lowland area and in the north Colombian and Venezuelan
Highlands, the pattern was one of tightly nucleated towns or vil-
lages. The houses were carefully arranged around a central plaza
which contained the chief's and nobles' houses or palaces and the tem-
ple, and the village was surrounded by a palisade (western Colom-
bia, Ohihcha, Timoteans, Oorgaho, Lache, northern Venezuela, Pan-
ama and Talamanca Division). Data on village size are few: the
Laclie village had up to 4,000 persons; the Aruacay^ 1,000; the Antil-
lean Arawak^ 3,000 ; tribes of Panama, 1,500 ; and tribes of the Carib-
bean Lowlands in Central America, 100 to 500.
A feature of the Circum-Caribbean village that relates it to the
Tropical Forest village both in settlement pattern and social struc-
ture is the occurrence, in some cases at least and perhaps quite com-
monly, of the large, communal frame house which sheltered an ex-
tended family or lineage. The main difference is that the Tropical

Forest community often consisted of only one such house, whereas the
Circum-Caribbean village had many, so that there was no village exog-
amy and an individual's social and political affiliations and obligations
extended beyond his kin to the many fellow villagers who were unre-
lated to him. Where the bilateral or conjugal family was the basis
of society, as was evidently the case in Ecuador, sociopolitical affilia-
tions also extended to nonkin.

Among most of the Circum-Caribbean tribes, the lineage seems to
have been matrilineal and there are some cases of matrilineal clans,
but evidence bearing directly on this point is scant. The Talamanca
Division and the Antillean Arawak seem clearly to have had matri-
lineal extended families or lineages, and the Brihri and GuayTni had
matrilineal clans. The Mosquito^ the Southern and Western Colom-
bian Sub- Andean tribes, and the Ghibcha probably had matrilineal
descent of chieftainship. The Fincenu even had female chiefs. Fe-
male warriors were also commonly reported in this area. further A
suggestion of matrilineal descent in Colombia is the common occur-
rence of avuncular marriage, but this, like the brother-sister marriage
practiced by iTica Emperors and by some Colombian chiefs, including
the Ghibcha^ may have been merely a form of family endogamy de-
signed to keep descent within the family. Kirchhoff believes that
northern Peru also tended to be matrilineal (this volume, p. 295),
though the Central Andes seem to have been very definitely patri-
lineal.
In the social classes, thesystem of warfare, and the priest-temple
cult, these tribes found a new set of factors for governing interpersonal
relations. In contrast to the Tropical Forests, the kinship basis was
thrust into the background status came to pertain to class membership
;
rather than to age, and political control fell to the village chief instead
of the kinship head. Individualism was greatly curtailed, special as-
sociations gave way to warrior groups (the only recorded men's secret
society is that of the Guaymi) and shamans became functionaries of
,
the temple cult as well as private practitioners.

—
The political pattern. The Northern Andean and Circum-
Carribbean political unit was a group of houses or lineages, and its
head was a chief rather than family elder. The office usually was quite
strictly hereditary, though in some tribes, for example in Ecuador, a
chief was theoretically supposed to have exceptional qualities of leader-
ship, and in some tribes {Talamanca Division), chiefs, especially war
chiefs, were elected.^
* A problem that would repay careful comparative study is the political power and social
status of the priests or shamans, the hereditary secular chiefs, and the war leaders. The
place of all three in the sociopolitical hierarchy would throw much light on the origin
of the social classes.
The village usually constituted the state, but there was considerable
tendency to tribal federation and even to the formation of empires or
kingdoms. Though political consolidation of groups of villages fell
far short of that of the Central Andes, the Gara are alleged to have
dominated most of Ecuador in pre-Inca times. Apart from the Cara
reahn, all the Puruhd were under one chief and the Ganari were fed-
erated under a hierarchy of chiefs. On the coast, the seven Pwnd
divisions had an over-all chief, but each Esmeralda town and probably
each Manta town was independent. In Colombia, federations of sub-
tribes to form the tribal political units had made some progress among
the Pciez^ Lile, Ancerma^ Gatio and other Highland tribes, and among
the Quimbaya, Tolu^ Genu^ and Momfox of the North Lowlands.
There were also incipient empires that were economic if not political.
They were established by conquest and tribute was exacted from the
conquered peoples. Among the Ghibcha and their neighbors, at least
two important realms were established (Handbook, 2: 887-897).
Northern Venezuela is inadequately known, but considerable inter-
village seems indicated for the Timoteans^ Gaquetio^
solidarity
Gumanagoto, and Aruacay. The Arawak of the Antilles and the
Guaymi, the Talamanca Division, and the Giletar of Central America
had also achieved loose states.
—
Social classes. The class structure really stemmed downward from
the chiefs, nobles, or lords, who constituted the most clearly defined
class. Usually, several chiefs formed a privileged and more or less
hereditary aristocracy of considerable power. Among the Tropical
Forest tribes, the main privilege of the chief was that of practicing
polygyny. The Northern Andean and Circum-Caribbean tribes ac-
corded chiefs this privilege in greater degree, and the Puna and the
peoples of the Unare Eiver in Venezuela are alleged to have had
eunuchs to guard the chief's harem. A
chief's wives or concubines
were so bound to him that they were often interred with him at death.
A chief or lord was also distinguished from commoners by occupying
the largest and most central house of the village by having his land
;
cultivated and many other services performed by common folk; by

bearing a special title; by wearing distinctive garments, insignia,
and ornaments of precious metals, gems, and sometimes pearls; by
sitting on a special stool, usually of carved wood; by being carried
in a hammock or gold-adorned litter {Ghibcha and other Colombia
tribes, Gaquetio^ Unare Kiver in Venezuela, Antillean Arawak) by ;
receiving obeisance from his subjects by marrying his sister {Ghibcha,

;
G array a, Picara, Paucura) or his sister's daughters; and by elaborate

disposal of his body at death. As many of the special funerary prac-
tices can be inferred from archeological data, they afford an excellent
clue to the presence of a class system among prehistoric peoples.
Vol.51 SOUTH AMERICAN CULTURES —STEWARD 721
Among the Northern Andean and Circum- Caribbean people, the

chief's body was: Put in a tomb, seated on a stool {Pwruhd) and
accompanied by his wife {Panzaleo) buried with his wives {Canari)
; ;
buried with wives and slaves {Ghibcha) buried with wives in a deep-
;
shaft grave {Pasto, Quillacinga, Caramanta, and perhaps several

Ecuadorian tribes) desiccated or cremated and buried {Qimribaya)
; ;
mummified and kept or else buried {Popaydn, Giietar) buried in a ;
mound-enclosed vault (North Colombia Lowlands; Cauca-Atrato re-

gion) desiccated, placed in a hammock, later cremated, and the ashes
;
drunk {Caquetio) desiccated and hung in a house (Chiribichi)

; ;
desiccated, buried, and later roasted or reburied {Piritu) disem- ;
boweled, desiccated, and kept as an idol or buried with several wives

(Antillean Araivak) desiccated and kept in a house or buried with
;
wives and retainers (Panama) or sewed in a mat and buried with

;
retainers {Mosquito) Mound burial was probably more common than

.
reported, but it seems not to have been characteristically Circum-

Caribbean, and it was not a typical Andean trait. The custom of
cremating the ashes and drinking them with chicha occurs among a
few Venezuela tribes, but is more characteristic of the Tropical For-
ests, especially of the western Amazon.
Colombian tribes have been accredited with matrilineal succession of
chieftainship, and in view of their strong matrilineal tendency (see
above) this , is perhaps expectable. The evidence for matrilineal suc-
cession is usually that a chief was succeeded by his sister's son, but
where brother-sister marriage was practiced, the sister's son was
also the chief's son. A chief's marriage with his sister's daughter
would also give the superficial appearance of matrilineal succession.
True matrilineal succession, therefore, can only be assured where, as
in the case of the Ghibcha, it is specifically stated that the title passed
to anephew rather than to a son.
The aristocratic or noble class was ill-defined and less fixed by
heredity than that of the chiefs or lords. It seemingly consisted
—
of lesser chiefs certainly the chiefs of villages within a federation
and perhaps the headmen of lineages within a village and probably —
of famed warriors and wealthy persons who achieved their status.
Shaman-priests are probably also to be counted in this social stratum.
The nobles were neither so fixed by heredity, so absolute in power,
nor so neatly subdivided as in the Inca Empire. Their privileges
—
resembled those of the overlords in kind a Chihcha noble might
—
have 100 wives but not in degree.
The commoners must often have been the kin of chiefs and nobles,
but they lacked privileges, were servile before their betters, and had
to supply cloth and other produce and labor to the upper classes.
They probably had some potential upward mobility, however, and
could improve their lot through achieving war fame and capturing
slaves.
The war captives, but little is known of their true
slaves were
place in society. Wliether they formed an important group of
drudge laborers, helping their captors to rise economically, is an
open problem. It appears that male captives added to their captor's
prestige mainly through affording them victims for sacrificial rites,
cannibalistic feasts, and human trophies, and that female captives
augmented their masters' polygynous households. Women may also
have performed useful economic tasks, but farming in these cultures
was the work of both sexes.
—
Warfare. Circum-Caribbean and Northern Andean warfare was
of vital importance to the sociopolitical patterns. It was highly
developed, the armies numbering in the thousands and even including
women. A
few tribes, notably the Ohihcha^ conquered neighboring
peoples and exacted tribute, but probably none wholly abandoned
the pattern whereby individual warriors gained fame through the
capture of slaves and sacrificial victims. The armies were not
wholly subservient to the state, as in the Central Andes, where alien
masses were incorporated into the empire and became commoners
and soldiers.
Warfare afforded an individual four means of attaining rank: (1)
The honor of taking war captives for cannibalistic feasts; (2) aug-
mentation of his polygynous household with captive women; (3) the
use of titles and the display of insignia and human trophies; and
(4) probably the acquisition of some drudge slaves, male and female,
who increased his wealth. A
Cuna warrior received a title for killing
20 victims; Mosquito and Sumo warriors gained rank and insignia;
among the Bribri^ Cdbecdr^ and Terrdba^ they were accorded special
burial and among the Caracas^ they had a graded military class, with
;
special insignia.
Cannibalism was strongly developed in Colombia (South Colombia
Highlands, except the Pdez; tribes east of the Cauca and of the Cauca-
Atrato region), and in Venezuela {Cumanagoto^ Marcapana^ Pa-
lenque) . Among the Caramanta and tribes east of the Cauca River,
it amounted human flesh. The Piriiu, though
to a definite appetite for
not cannibals, drank powdered human hearts in chicha. The Chibcha
and certain Central American tribes sacrificed victims for ritual rather
than for cannibalistic purposes, except that the Meso-American tribes
were strongly cannibalistic, the Sumo were probably cannibals, and
Lenca warriors ate their victims' hearts. The Antillean Arawak did
not eat human flesh, though the Island and Mainland Carib, both of
them Tropical Forest peoples, were fierce cannibals.
Correlated with cannibalism was the display of human trophies:
Vol. 5] SOUTH AMERICAN CULTURES —STEWARD 723
skulls (Panama, South Colombia Highlands, Chibcha^ Western Co-

lombia, tribes east of the Cauca) the flayed skin stuffed with ashes
;
or straw, and the hands, feet, and other parts of the body (South
Colombia Highlands, Western Colombia, tribes east of the Cauca,
Manta^ doubtfully the Corbago). Except for the Piritu, who made
flutes of human bone, a Tropical Forest trait, the Timoteans, the tribes
of northern Venezuela, and the Antillean peoples used no trophies.
In Central America, the Sumo used human teeth.
Warfare was an adjunct to religion in that it supplied victims for
sacrificial rites among the tribes east of the Cauca River, the Cauca-
Atrato area, the Chibcha^ the Cuna, the Guetar and perhaps others
of the Talamanca Division, and some of the Meso-American tribes.
(See also below.)
A little-understood phenomenon is the occurrence of female warriors
among many of the Colombian tribes, the St. Croix Island Arawuk^
and probably the Guna. It seemingly had some connection with trans-
vestitism. Male homosexuality, on the other hand, was common on
the coast of Ecuador.
EEI.IGION AND SHAMANISM
The priest-temple-idol cult of the Central Andean peoples centered

around the public worship of tribal gods, who were represented by
idols kept in temples, and the worship was directed by a special class
of priests. Few Circum-Caribbean and Northern Andean tribes had
precisely this pattern. Some tribes lacked almost all features of it,
whereas others had a cult that was transitional between the Tropical
Forest and Central Andean patterns. The priest-temple-idol cult
seems to have been entirely absent from Northern Venezuela and at-
tenuated among the Island Arawak. In Central America, some tribes
may never have had it, but post-Contact deculturation probably eradi-
cated it among many others, so that its aboriginal presence can only
be ascertained by further perusal of the early chroniclers. Certain
archeological remains that appear to have been ceremonial centers and
many fragmentary beliefs and customs that survive among the modern
tribes strongly suggest that a temple-idol cult was formerly present.
In the Circum-Caribbean and Northern Andean tribes, the main
religious functionary was the shaman. Like the Tropical Forest
shaman, he cured disease, divined, and practiced magic with the aid
of his personal spirit helper, but he also assumed the duties of inter-
mediating between the community and its gods. That the same per-
son filled the dual roles of priest and shaman seems clearly evident
for many tribes {Evegico^ Panzaleo, Pijao, North Colombia Lowlands,
Northern and Northwestern Venezuela, and probably Popaydn and
Ancerma). In these tribes, membership in the priesthood was not
hereditary but depended upon the supernatural experience by which

an individual obtained shamanistic power. Through his community
services the shaman attained considerable status, and often he had
political power equal to that of the chief or he was the chief (North-
ern Venezuela, Caquetio^ Island Arawak). Thus, priests, like war-
riors, had a potential upward mobility, and they were not fixed in the
social structure by heredity, as in the Central Andes. A few tribes,
however, apparently had a special, hereditary class of priests {Chibcha^

Meso-American tribes, Bribri and Giletar).
The religious rites were largely private, oracular seances and served
the needs of individuals rather than of the community. Public cere-
—
monials were not unknown the Chihcha had a harvest and New
Year's rite and an annual pilgrimage — ^but there was no fixed ritual
cycle, associated eitherwith a calendrical system or with agricultural
phenomena. The oracular performances resembled the public seances
held by Tropical Forest shamans but differed from them in that they
dealt with tribal gods rather than with the shaman's spirit helper.
The distinction, however, was not always clear-cut: shamanistic,
fetish, and guardian-spirit concepts are interwoven with those of
tribal gods, and spirits pertained to groups of different magnitudes
tribe, community, family, and individual. The zemis of the Island
Arawak^ for example, were household fetishes of different kinds, some
of them derived from dream experiences; only the major zemis, which
were kept in the chief's house and were worshiped in communal rites
under the chief's direction, can be regarded as idols in a temple cult.
Many tribes had household idols, which, though frequently worshiped
with supplication and sacrifice, were family, not community deities.
There was great variation in the nature of the community deities,
and they must represent the results of several streams of historical
influence combined with diverse local patterning. Most common were
celestial beings, especially the sun and moon (Manta, Canari, Chihcha,
Coconuco, Carrapa, Oatio, northwestern and northern Venezuela,
Caribbean Lowlands, Lenca). In the Tropical Forests, celestial de-
ities frequently were myth characters and often creators and High
Gods, but they were rarely the objects of community worship. The
jaguar, another widespread South American supernatural being, at-
tained cult status in some tribes {Nutihara, Fincenu) The worship
.
of and myth of emergence from sacred places, especially mountains

{Gara, Canari, Puruhd, Timoteans) springs {3Ianta), a cave (Antil-
,
lean Arawak), and rocks and trees {Manta, Canari) are probably
Andean-derived. The Lenca pilgrimages to sacred hills may be a
vestige of the same idea, which extends to Mexico. The Chihcha had
shrines to lakes, rivers, mountains, and caves. The nature of many
.
deities, however, is unknown, though the idols representing them are

described in detail.
Idols took many forms, such as clay (Puruhd), pottery, wood, stone,
and thread (Timoteans), various materials including three-cornered
stones {Island Arawak), and most elaborate of all, painted and gold-
sheathed images {Fincenu, Ancei-Tna)
Ancestor worship was probably very general, though the function
of the ancestor of the common family is not reported. Deceased
nobles and lords often became tribal deities. In the case of the
Chibcha and Antillean Arawak^ it is fairly clear that they were wor-
shiped, and in many other tribes it is probable that their remains were
mummified, as in the Central Andes, so as to preserve them as com-
munity or tribal fetishes or gods.
In the Tropical Forests, the men's club was usually the scene of
religious activities, such as initiation rites, seances, and other cere-
monies, and it, or a similar, early inter-American men's house, may
have been the prototype of the temple. The Northern Andean and
Circum-Caribbean peoples seemingly had no men's clubhouses, and,
though minor rites might be practiced in family dwellings or before
various lesser shrines and major ones sometimes in the chief's house, as
among the Antillean Arawak^ many tribes had a special structure in
which the idols were kept and the priest-shaman conducted the orac-
ular sessions and other rites. Offerings were made to idols only in
family dwellings among the Pdez^ Moguez^ Caramianta^ Pozo^ and
Arma, and on the upper Cauca and in parts of Central America. There
was a special temple among the Pasto^ Timoteans^ Chibcha^ the Meso-
American tribes, and the tribes of northwestern Venezuela and western
Colombia. In Colombia and among the Meso-American groups, the
temples were sometimes placed on mounds, but they seem rarely, if
ever, to have been built of stone, as in the Central Andes. Chibcha
temples, like houses and palaces, were of poles and mud. Among the
tribes of northern Venezuela, the Antillean Arawak and probably the
Timoteans^ shamans performed oracular functions in a cave.
Human sacrifice was practiced by many though not all of these
—
tribes the Timoteans and Island Arawak had none. It was closely
associated with warfare and, in some cases, with cannibalism. The
victims seem everywhere to have been war captives, except among the
Manta and the Chibcha, who also sacrificed fellow tribesmen. The
purpose of sacrifice was extremely varied to honor idols and sacred
;
mountains {Puruhd) to insure war success {Pozo) to effect cures,

; ;
to insure fertility by offering blood to fields, and to provide hearts to

be worshiped as gods {Huancavilca) to appease the Sun, especially
;
during droughts, by tying-up the victim, killing him with darts, and
cutting his heart out {Chibcha, who also sacrificed animals) ; to obtain
ghosts to guard family idols {Pijao) to control the weather by offer-

;
ings of hearts {Caramanta) to supplicate the sun for rain with the
;
blood of young girls (northwestern Venezuela) to appease temple

;
idols, the skins of the victims including animals being filled with straw
and ashes {Manta) to honor burial feasts and each moon {Giletar)
; ;
and to accomplish unspecified purposes {Arma and Quimhaya, who

perf ormd the rite on special platforms Picara; Ancerma; and Meso-
;
American tribes)
Kitual elements link the tribes with various areas. Taking salt
(Ecuador) and coca (Ecuador and Colombia) are Andean; burning
copal and the steam bath, are Mayan and Mexican; flagellation
(Puruhd) is Tropical Forest; fasting (Pciez), hair cutting {Puruhd,
at girl's puberty; Pansaleo, as punishment; Puna, as a men's style),
and use of tobacco are widespread.
The strictly shamanistic practices of the Northern Andean and
Circum-Caribbean tribes are in the general South American pattern,
representing beliefs that had considerable antiquity and persisted with
great stability regardless of the cultural context they entered. Under
the influence of a stimulant or narcotic (tobacco. Datura, coca in the
Andes, or other drugs according to local species), the shaman used
his supernaturalpower to cure disease by sucking out the object caus-
ing illness, and he divined, prognosticated, and worked various
kinds of magic. There was also considerable herbal curing.
TECHNOLOGY AND MATERIAL CULTUBE
The basic technologies in handling wood, stone, ceramics, fibers, and

metals were like those of the Central Andes, but the manipulative
skills were definitely inferior. Despite the quantities of wood avail-
able, its stylistic handling was stilted and no vigorous wood carver's
art is reported. Stone was little used. In ceramics, where unlimited
scope in molding, modeling, and painting would have been possible,
the art was confined to narrow plastic traditions and to restricted
painting. Few woven products have been preserved, but the textile
art seems to have fallen far short of the craft potentialities realized
in Peril, the products having a limited range of techniques and designs
being painted on rather than woven in. Metallurgy achieved smelting
and alloying of gold and copper, but smelting and alloying of other
metals and probably many of the Central Andes metallurgical proc-
esses were not employed.
—
Metals and gems. Copper and gold were smelted and alloyed
among the Pasto, Pdez, Moguex, Pijao, and the tribes of the upper
Cauca, east of the Cauca, and in Panama, but copper occurs only rarely
again in Central America until the Maya frontier is reached. In
Ecuador, the Canari had gold, gilded copper, and silver, but only
gold and silver are ascribed to the Esmeralda and Manta. Metal
were traded as far as the Antilles,
objects, including gold-copper alloys,
but smelting seems not to have been practiced by the Ghibcha^ the
Timoteans^ and the other tribes of northwestern and northern Vene-
zuela,perhaps for environmental reasons, though some of them may,
like theIsland Arawak^ have worked placer gold.
Several precious stones including emeralds were used from Ecuador
to Panama. Pearls were probably used by most Coastal tribes, includ-
ing the Manta J Cuna^ and the peoples of northwestern and northern
Venezuela.
—
Ceramics. Pottery reached a high development, mainly in the
modeled, incised, and applique tradition, wares of this type being
found throughout the Sub-Andean and Circum-Caribbean area.
Three-color polychrome is mainly Central America and is probably
of Meso-American origin.
—
Weaving. True loom weaving characterized all these tribes, except
perhaps a few in Central America and the Antillean Arawah, who
netted their fabrics. All fabrics were of cotton and other vegetable
fibers. Use of llama wool, except that obtained in trade, was post-Zîâ
in Ecuador, and use of sheep wool was post-Columbian everywhere.
The cotton was generally cultivated, but the Talamanca Division and
the tribes of northern Venezuela and perhaps the Island Arawah used
a wild species. Ornamentation of cloth was usually with painted
rather than woven-in designs.
—
Basketry. Basketry appears to have been general, and, to judge
by a very few descriptions of the weave, was of the Tropical Forest
twilled and woven varieties.
—
Bark cloth. Though also a western Amazon trait, bark cloth was
made in Ecuador (Pasto), western Colombia, and Central America.
Transportation. —
Dugout canoes were used by all coastal and river
tribes.The Antillean and Panamanian peoples built huge, highly
ornamented craft. The Manta used large, sea-going balsas, as well as
dugouts.
The carrying basket is ascribed to some tribes, but the netted carry-
ing bag appears to have been more common (Pdez, Central America,
Northwestern Venezuela, Island Araioah).
—
Gourds. The manufacture of decorated gourd containers was a
characteristic feature, and in Colombia {Pasto, Moguex, Pdez) the
gourds were ornamented with a special '"''Pasto varnish."
Household furniture. —
Like the Tropical Forest peoples, these
tribes used hammocks and wooden stools. The platform bed is mainly
Andean, but occurs sporadically: Chibcha, the Central American
peoples, and the Ancenna. Among the Island Arawak, connnoners
used platform beds while chiefs used hammocks.
738931—49 48
Stone metates and stone stools occurred especially in the Andes and
Central America.
—
Weapons. The bow was used by a few tribes in western Colombia
(Ahurrd, North Colombia Lowlands), the Timoteans, some peoples of
northwest Venezuela, and most of the Central American tribes. As
elsewhere in South America, it seems to have spread at the expense of
the spear and spear thrower, with which it had a negative correlation.
The spear thrower remained typically Highland in this area it instead
;
of the bow was found in Ecuador, southern Colombia, among the

Chihcha, and in parts of western Colombia. It also survived among
the Antillean Armoah and some peoples of the Venezuelan Andes.
The sling, another typically Andean weapon, was widespread in
Ecuador and Colombia, and also survived among the Ahurrd and on
the Island of Trinidad. Stones were also thrown by hand among the
Pasto and in the Antilles.
Clubs, especially of the flat or macana type, were general. Shields
were used in southern Colombia and the Venezuelan Andes (map 5).
Axes seem to have had a wide distribution, but they were generally
of stone, being made of copper only in Ecuador.
The blowgun was common to the Sub- Andean and Central American
tribes, but the latter shot pellets; the poisoned blowgun dart only
reached the Cuna during historic times.
—
Esthetic and recreational features. A variety of musical instru-
ments are recorded, the most characteristic and widespread being shell
trumpets, flutes, drums, and rattles.
Of stimulants and narcotics, chicha was probably general. Tobacco
had a wide distribution (map 9), but it and other narcotics were
consumed principally by shamans. Coca, too, was widespread, though
its occurrence in Central America and the Antilles is questionable.
Natively, it had ritual and shamanistic rather than popular use.
THE STJB-ANDEAN TRIBES OF EASTERN BOLIVIA

Considerable deculturation of much of eastern Bolivia during the
historic periodand inadequate data on aboriginal culture leaves in
doubt just how many native tribes might be considered Sub-Andean.
The Mojo (Handbook, 3: 408), Baiire (3: 409), Manasi (3: 388),
Paressi (3: 349), and Paunaca (3: 396) were Sub- Andean from a
sociological rather than material point of view. They formed enclaves
among Tropical Forest and Marginal tribes, but, as the archeology of
eastern Bolivia suggests considerable cultural development and some
Andean linkage at approximately the Tiahuanaco Period of the High-
it is evident that the Sub- Andean culture had been partially
lands,
swamped out and compressed by tribal movements bringing more
primitive peoples into the area. All these tribes are Araioakan^
except the Manasi, and they probably moved into the area from the
Tropical Forests to the north.
SOCIOPOUTICAL AND EELIGIOUS PATTEBNS
Intensive cultivation of tropical root crops as well as of many domes-

ticated plants derived from the Andes supported a though not
stable
very dense population. Villages, except among the Mojo, were large
and comparatively permanent. Mojo villages averaged only 60 to 80
persons, though some were larger, but those of the Paressi had 10 to 30
large houses, each with 30 to 40 persons, a total of 300 to 1,200 persons
per village. Villages were very compact and carefully planned, the
chiefs' houses and ceremonial hall standing in the center. Baure
villages were palisaded.
—
Social classes. All these tribes were socially stratified, some more
than others. Mojo chiefs were probably hereditary, except that sha-
mans often became chiefs. War captives, though allowed to marry
into the tribe, were held in contempt and probably ranked below the
commoners, but the classes were obviously ill-defined and quite fluid.
The other tribes had the characteristic Sub-Andean class pattern.
Among the Baure were ( 1 ) a somewhat endogamous group of chiefs,
:
who were furnished food by commoners; (2) commoners; and (3) a

servile class, presumably war prisoners. The Paressi classes were:
(1) hereditary chiefs, who lived in special houses and were furnished
food and other necessities by their subjects (2) nobles, who were heads
;
of independent families (probably of the communal houses made up

of many captive wives) and who controlled drudge slaves; (3) war
captives who were serfs or slaves. The Manad had: (1) powerful
hereditary chiefs, whose special houses in the village plaza were used
as temples and public meeting places, who received homage from their
subjects, and who were accorded special burial in a stone- or wood-lined
grave; (2) shamans and a distinct class of priests (3) a nobility made
;
up of headmen (of households?) and (4) commoners. The chiefs of

;
these tribes were not carried in litters, did not receive elaborate burial,
and lacked many of the other evidences of status of the Andean and
Circum-Caribbean area.
The extent ,of the political group is unclear. In all tribes, each
multilineage or multifamily village had a chief, and among the Mojo,
two or three villages constituted a subtribe under a chief. The Paressi
waged wars of conquest, but it appears that these wars incorporated
alien groups into the Paressi class system rather than created incipient
empires or even federations.
—
Warfare. There was some warfare, especially among the Paressi.
The capture of men for slaves and women for wives enhanced indi-
730 SOUTH AMERICAN INDIANS LB. A. E. Bull. 143
vidual prestige. The cannibalism-human trophy complex of the

Circum-Caribbean peoples is not reported, but a kind of substitute
is
found in the Mojo jaguar cult. Jaguar hunting was the, or one of the,
main avenues to fame, and jaguar trophies were kept along with other
religious objects in the hall which served as men's clubhouse and
temple. Human skulls were kept, but these may not have been enemy
troj)hies.
The priest-temple cult. —These tribes had temples— in some cases,
the chief's house ; in others, the men's club — dedicated to cult worship,
but in some cases the cult had vestiges of a men's secret society.
Shamans, who gained power through individual experiences, served as
oracles and priests, consulting and making offerings to the gods. Ap-
parently only the Baure and Manasi had a special, hereditary class of
priests, but the deities were true tribal or village gods, not mere
shaman's spirit helpers.
Among the Paressi, the men's club was the temple, and it was dedi-
cated to the serpent god and perhaps other deities. It contained the
sacred paraphernalia, including a trumpet, which represented the voice
—
of the gods and was taboo to women a feature of the Tropical Forest
men's secret society. The Mojo also used their men's drinking hall,
a special structure, as a temple. Taboo to women, it was the shrine of
jaguar trophies and perhaps of sacred musical instruments. Here,
special shamans consulted jaguar spirits and made offerings to the gods,
among them certain celestial and nature deities. Both men and women
were shamans, but the women may not have served as priestesses.
Manasi priests were initiated into their class. They alone entered
the sanctuary, which was in the chief's house, to consult and make offer-
ings to the gods, who carried them away. The gods seem to have had
some association with the thunder.
TECHNOLOGY AND MATEBIAL CULTUBB
The technology of these tribes was preponderantly of the Tropical

Forest type, and it lacked such Sub-Andean elements as metallurgy.
Cultural elements shared with the Tropical Forests include cotton ham-
mocks, wooden benches, dugout canoes, bark cloth, twilled and hex-
agonal basketry, good painted pottery, cotton weaving on the vertical
loom, Muscovy ducks, bows and arrows, blowguns, spears, rubber-ball
game, hollow-log drums, and pole-and-thatch houses. More specifi-
cally Andean, perhaps, were trade in salt the Mojo causeways, which
: ;
had lateral canals that served as canoe channels (it is quite possible
that these antedate the Mojo by a long period) balsa rafts; primitive
;
bridges and slings.

; The Mojo also used bull boats, bolas, spear throw-
ers, panpipes, trumpets, and clarinets.
THE CENTRAL ANDEAN PEOPLES
The high civilization found in the Central Andes at the time of the
Spanish Conquest was that of the /nca, and it was coterminous with
the area in which the Inca culture largely replaced that of the local
tribes. One hundred years before the Spanish Conquest, the iTwa
occupied only the immediate vicinity of Cuzco, and the remainder of
Peru, Ecuador, Bolivia and North Chile was divided into more than
100 independent states, some of them small tribes, others, groups of
tribes forming kingdoms. Each had its own culture (Handbook,
2: maps, 3, 4, 5, 7). The Chirmi Kingdom of the north coast of
Peru was probably an empire, but it did not compare with the final
Inca achievement. The Inca ultimately subjugated all peoples from
the Colombian border of Central Chile and incorporated them in a
single state or empire. They deeply affected the cultures of Peru and
Bolivia, especially the sociopolitical patterns transcending the com-
munities, but they only superficially influenced Ecuador and North
Chile. The Uru^ an insignificant enclave in the Rio Desaguadero-
Lake Poopo region of Bolivia (Handbook, 2: 575), remained quite
primitive. The Inca culture, in other words, was imposed on the
peoples of the continuous mountain mass of the Central Andes, while
beyond, to the north and south, where the Cordillera begins to fork
and breaks into smaller blocks, considerable cultural diversity
persisted.
The Central Andean culture brought into sharp focus the patterns
foreshadowed among the Sub-Andean and Circum Caribbean tribes.
Intensive farming and herding supported the most dense population
and the largest and most permanent communities in South America,
and it yielded a surplus which released large portions of the population
for manufacturing and for governmental, religious, and other non-
survival assignments. Social classes became hereditary, endogamous
castes, but the nobility rested on a large base of wealth-producing
commoners rather than on a slave class. Warfare was an implement
of imperialism instead of a source of slaves and ritual victims, as
among the Sub-Andean and Circum-Caribbean peoples, and, though
commoners could improve their condition through war exploits or by
becoming special artisans, they could not rise to a higher caste.
Religion was organized around a state temple cult and a hierarchy of
state and local gods, and it was served by a graded priesthood, the
priests being clearly differentiated from shamans, who were curers
and magicians. The whole society was highly regimented through a
code of laws and regulations which were enforced through a judicial
system, and little was left to tribal custom with its less overt sanctions.
In agriculture, herding, road building, megalithic construction.
732 SOUTH AMERICAN INDIANS [B, A. E. Bull. 143
metallurgy, weaving, and in achievements in government, conquest,

and class organization, the Inca surpassed anything in the New World.
In intellectual attainments, however, and in esthetic achievements,
except in pottery and textiles, they fell far short of the Maya. They
had no system for writing words or sounds, and they recorded num-
bers only with knotted strings, or quipus. Their graded hierarchy
of deities lacked the refinements of the Mexican pantheon, with its
esthetic concomitants in ritual, architecture, and religious art. Their

ceremonial calendar, based on a simple adjusted lunar count and ig-
noring even solstices and equinoxes, did not compare with the carefully
calculated day counts and other means by which the Maya reckoned
time.
The main difference between the Inca and the Maya is that the
former intensified warfare and conquest to create an empire dominated
by political and social organization, whereas the Maya., a compara-
tively peaceful people, developed their temple cult so that an esoteric
priesthood became the leading class and life revolved around cere-
monial centers. Highly developed religious art, architecture, and
ceremonialism and a refined calendar marked the peak of the Maya
achievement. Probably both trends were implicit in the Circum-
Caribbean culture.
Andean farming was the most intensive in the Hemisphere. It em-

ployed vast irrigation works, terracing on steep hillsides, and use
of fertilizers. More than 30 different species (Handbook, 2:5; 6
Sauer) and innumerable varieties of plants were cultivated of potatoes
;
alone there were scores of varieties. Llamas, alpacas, guinea pigs, and
ducks were tended. Local specialization according to the environ-
mental zones of the Andes (about one-third of the crops were High-
land specialties) coupled with considerable exchange of produce gave
a varied as well as abundant food supply. Wild foods were of minor
importance, except for fish on the coast and fish and birds in the Lake
Titicaca-Rio Desaguadero region.
The efficiency of Andean agriculture is evidenced by the large popu-
lation,which was the most dense in South America, and by the surplus
production. The latter was insurance against want and released
people for other work. Although the land was worked by commoners,
only one-third of the produce went for their keep, the other two-thirds
going respectively to the state and church, whose members did not
engage directly in material production. Moreover, vast numbers of
commoners were taken from their fields for military service, special
craft production, household service, and labor on roads, bridges, build-
ings, and other public works. It is said that 30,000 persons were used
in the construction of the fort atSacsahuaman and that the emperors

sometimes had difficulty in inventing projects to keep people busy.
Extensive movement of produce and goods was made possible by the
use of large, seagoing balsa rafts on the coast and of pack trains of
llamas traveling over an excellent system of roads in the interior
(Handbook, 2:229). This efficient transportation, an important
factor in the development of urban centers, surpassed anything in the
remainder of South America, where goods were moved in pre-Colum-
bian times by human carrier, or by dugout canoes, as along the Amazon-
ian waterways and on some of the coasts. The post-Columbian horse
of the Pampas was probably of great efficiency, but the population and
economic surplus of this area was insufficient to support large and
elaborately structured aggregates of people.

Settlement pattern and composition. —The combination of pro-
ductive and transportational efficiency in the Central Andes gave con-
siderable latitude to the settlement pattern : people could live either
dispersed over the countryside or concentrated in urban centers.
Archeology has shed very little light on settlement patterns, not for
want of data but because of preoccupation with other problems.
Chanchan on the north coast (Handbook, 2: pi. 51) is an extraordi-
narily large, compact, planned urban center, covering some 11 square
miles. Possibly the restriction of settlement area.s on the coast to the
narrow confines of irrigated valleys, which are bounded by stark
deserts, predisposed valley and coastal communities to strongly nu-
cleated types. The Highlands, dependent on rainfall more than irri-
gation and given over to considerable herding, allowed more dispersed
settlements.
The cultural factors which operated within the ecological limits to
determine the size, composition, and distribution of communities are
partially evident from ethnographic data. The unit settlement of the
Highlands was a compound enclosing as many as six houses of stone,
stone-and-adobe, or adobe, grouped around a court. Each house
sheltered a biological family and the compound apparently accommo-
dated an extended, patrilineal family (Handbook, 2:223). In pre-
Inca times, the compounds were irregularly clustered around cere-
monial and administrative centers to form unplanned villages, which
rarely included more than 300 families, or about 1,500 persons. The
villages, located near but not on arable land or grazing areas, were not
fortified, refuge being sought in hilltop forts.
The Inca set out on a deliberate program of (1) creating new admin-
istrative and ceremonial centers and (2) moving villages away from the
old hilltop forts to sites nearer the fields. The new villages and cities,
planned by government architects, were ideally laid out in squares,

each with one to four extended-family, unit-type compounds or en-
closures. Except for capital towns, however, the new settlements were
little larger than formerly. Cuzco attained great size, but a consider-
able portion of the 100,000 inhabitants of the valley were scattered in
small villages, separated from one another by farm lands.
The ayllu, a much-discussed social and territorial unit (Handbook,
2 253, 483, 539 this vol., p. 293) seems to have been essentially a land-
: ; ,
owning group consisting of one or more villages and having a myth of

common origin for its members. The land, at least under the Inca,
was distributed annually to the constituent families. The ayllu can-
not have been a clan as defined herein, because it consisted of many
lineages or extended families which were probably related only through
marriage, because marriage was endogamous within it, and because
kinship terms reflect a bilateral kinship structure. The ayllu was
patrilineal only in that office was so inherited group affiliation in an
;
endogamous group is necessarily bilateral. In -pre-Inca times, the

ayllu often seems to have constituted a politically independent tribelet,
which was constantly at war with its neighbors, but some tribes con-
sisted of several ayllus. Under the Inca, and perhaps before, the
ayllus were grouped into twos (moieties) and sometimes threes for
administrative purposes (Handbook, 2:255, 262). The AyTnara
moiety, however, is now and perhaps was in Conquest time an endo-
gamous territorial division of a community within the ayllu (Hand-
book, 2:541).
—
These population groupings of different magnitudes community,
—
ayllu, moiety, and state gave the individual a set of ever-larger
spheres of interpersonal relations that extended far beyond his kin.
Within this enlarged framework, his behavior was governed by a com-
plex set of behavior patterns, which were adjusted to the vertically
graded positions in the hereditary class system as well as to his kin.
Many of these patterns were sanctioned by law. The graded system,
which patterned military and religious as well as civil institutions, is
reflected among the Sub-Andean and Circum-Caribbean tribes and it
was foreshadowed among some -pre-Inca Andean peoples, especially
the Chimu, but it developed in full strength under the Inca^ who im-
posed it on all the peoples from North Peru to southern Bolivia and
who somewhat modified the peoples even farther to the north and
south.
—
Classes and government. Before the Inca conquest, the Andes
surely had social classes with a hereditary tendency, but status, at least
in the smaller groups, could also be achieved through warfare. The
expansion of the Inea state created an elaborate and rigidly fixed
society throughout the Empire. Starting in the vicinity of Cuzco with
a sociopolitical system probably not very different from that of their

neighbors, the iTwa expanded during the century before the Spanish
Conquest and subjugated tribe after tribe, incorporating them into
their own scheme. The Inca^ especially the direct descendants of the
Emperor, became the dominant group and highest caste, but the chiefs
and nobles of conquered peoples were also given high status in the
realm. Commoners generally remained such. Unlike the Sub- Andean
and Circum-Caribbean peoples, among whom the status of the nobles
seems to have depended partly upon a captive slave class, the Inca
nobility was supported by the commoners, who tilled the third of the
land set aside for the state and performed other services. In other
words, conquest incorporated wealth-producing commoners within the
realm rather than adding a captured slave class to the bottom of
society, where there was really no place for it.
(1) The state was headed by the Inca Emperor, who had absolute
authority in all civic affairs and was considered a divinity. The later
emperors were so exalted that they married only their full sisters. The
Emperor lived in a splendid palace, wore special garments and orna-
ments, bore a title, was carried in a litter, and sat on a stool atop a
throne while attendants sheltered him with a parasol and fanned him.
People paid him obeisance according to elaborate rules. He had many
wives and servants. At his death, the members of his retinue were
made drunk and killed by strangulation. The Emperor's body was
mummified and placed in a shrine. Deceased Aymara chiefs and nobles
were placed with their wives, llamas, and other property in stone towers
(chullpas).
(2) The em-
nobility consisted of {a) the descendants of past Inca
perors and special persons called by privilege" and (5) the local
'''Inca
curacas or chiefs. The Inca nobles were divided into 11 ayllus, each
the patrilineal lineage of an emperor, and they had the highest rank,
held the most responsible positions, and were accorded many honors
comparable to those claimed by the Emperor. The curacas were made
up of local chiefs and nobles, who had held power before the Inca
conquest, and of certain provincial rulers who, though appointed origi-
nally for their ability, were given permanent and hereditary status.
The nobility became an elaborately graded hierarchy, the upper levels
approaching the Emperor in status and the lower standing little above
the commoners. There was little possibility of upward mobility be-
cause the system became fixed, but degradation of rank could occur as
generations became more remote from high-ranking ancestors. This
degradation was somewhat checked, however, not only by class endog-
amy but by intrafamily marriage, the Emperor taking his full sister
as his principal wife and the nobles their half-sisters, although both
took additional wives or concubines from the commoners. These mar-
riages in violation of incest regulations may have been a fairly old

Andean custom, designed to protect a hereditary chief's class, for such
marriages occurred north to Colombia.
(3) The commoners formed a fairly homogeneous group. They
could improve their status only through warfare, in which they cap-
tured additional wives or for which they were rewarded with property
and women, thus distinguishing them from the ordinarily monogamous
commoners and giving them benefits nearly equal to those of the lower
nobles (this vol., p. 298).
The commoners produced their own sustenance on the third of the
land allotted them and cared for their own wants, but they were not
permitted to make luxury goods. The extended family lived in its
compound, preferably near its ancestor's mummy, but, unlike the
nobles, the commoners rarely traced their forebears beyond the grand-
father. Their daily life centered in the activities of the ayllu and
community (Handbook, 2: 483).
The commoners were obliged to supply produce for the Inca church
and state from the lands set aside for them, to do "mita" labor on
public works, to serve in the army, to provide domestic help for the
nobles, to supply the "Chosen Women" as wives or concubines for
the nobles and as nuns for the convents, and to work as special artisans
for the Emperor. In the absence of an important slave class, they
performed all the basic and menial labor as well as a few of the more
skilled tasks.
Warfare. —In pre-/nca times, most Andean warfare probably re-
sembled that of the Sub- Andean and Circum-Caribbean tribes in that
its purpose was plunder, revenge, and the taking of women, slaves,
sacrificial victims, and human trophies. Cannibalism had probably

been abandoned, had ever been practiced it is accredited to only
if it ;
the Aymara^ who roasted and ate their captives and drank their blood
(Handbook, 2: 548).
The objective of Inca warfare was to incorporate conquered terri-
tories into the Empire. For purposes of exploitation, it was more
expedient to leave the populations intact and exact labor and tribute
from them than to bring them into the Inca's already crowded lands
as a slave population. The Inca governmental and class system was
imposed upon the provinces, and, though whole populations might
be moved from one region to another to prevent rebellion and to speed
acculturation as far as possible, the pre-existing class structure was
retained.
The early phases of the conquest quickly elevated the Inca proper
to the upper nobility. As the conquest progressed, however, the com-
moners, who served in the great armies, profited little. Their role as
the main wealth producers was fixed by law and their rank was fixed
by heredity, so that their main rewards consisted only of extra wives,

a small amount of property, and a few favors.
After the Empire had become consolidated and the socioeconomic
system functioned to supply the nobles with all the goods and services
they desired, warfare was carried on more and more for the sheer
power and glory of conquest and as a religious crusade to impose the
Inca cult on the Andean peoples.
Human war trophies, apparently a very old Andean trait, were still
—
displayed under the Inca skull cups, ash- and straw-stuffed skins,
leg-bone flutes, skin drum-heads, and tooth necklaces. Presumably
these were evidences of individual prowess. There was no cannibal-
ism, and animals rather than human beings were ritually sacrificed.
The few human sacrificial victims were taken from the Empire popu-
lation rather than from war captives; the sacrifice of war captives
seems to have been restricted to victory celebrations.
—
The government and society. The Inca Empire introduced to
the Central Andes a state culture over and above the local or folk
culture, and a system of control through state law which was probably
only slightly evident among the Sub-Andean and Circum-Caribbean
peoples and was entirelyunknown among tribes whose sociopolitical
unit This system resulted from the need to weld
was the kin group.
together a tremendous and farflung population and to make it serve
the conquerors.
Under the Inca^ the conjugal and extended family remained, but
age became a basis for assignment to government tasks and the kin
head was superseded politically by a government appointed ruler.
The ayllu and moiety became administrative units. Men's clubs and
other social groups apparently were unknown. Birth customs prob-
ably were not greatly altered, and girls were isolated at their first
puberty, but boys' adolescence was celebrated by what appears, at least
in the case of the nobles, to have been an initiation into a warrior status.
At puberty, all boys received a breechclout and new name in the case
;
of noble youths, llamas were sacrificed and the boys were whipped,
made and given earplugs and weapons. Games were a char-
to race,
acteristics feature of mourning ceremonies. Local worship, with its
shrines, deities, spirits, magicians, and curers, was not abolished, but
the state religion was forced on all communities.
In the greater part of their behavior, the commoners were regi-
mented by government decree. They were counted, classified, graded,
and assigned to groups ranging from 10 upward in a decimal system.
They had fixed obligations to serve in the army, work for the noble
class, and lend their services to innumerable government projects.
They learned to speak Quechua^ the Inca language. They were for-
bidden use of the distinctive garments, insignia, objects of precious
metals, and other privileges reserved for their betters. Even their
marriage was regulated, and their very lives were at the disposal of
the Emperor. A large portion of their behavior was governed by
the legal code, which applied different principles to the different
classes.
The Empire was kept under control by large armies and garrisons,
a system of couriers who carried messages and supplies over the roads,
large numbers of census-takers, overseers, and the like, and a judicial
system.
—
Religion and the temple cult. Prior to the Inca conquest, local
religion had been based on belief in various tribal gods and the spirits
of sacred places. There were shrines and perhaps true temples, and
probably there was some kind of priesthood. The conquest imposed
Inca religion, with its temples, idols, priesthood, and system of ritual
on the local groups, but did not entirely destroy the local worship,
which remained in the same general pattern as that of pre-/nca times.
It servedmainly to create a hierarchy of gods, to introduce more elabo-
rate temples and more ambitious rituals, and to establish a large,
graded priesthood.
Under the Inca, the deities and other objects of worship were ranked
in order of importance: Viracocha, who was the Creator, Supreme
Being, and culture hero, and whose servants were the other gods the ;
Sun, ancestor of the Emperor Thunder, the weather god the Moon
; ;
various stars; the Earth and Sea; huacas or shrines, such as tombs,
buildings, and sacred places, especially springs, stones, and mountain
peaks and amulets and images, probably including those representing
;
individual guardian spirits. These range from national gods down

through ayllu and community deities to household and personal spirits.
There were also various evil spirits.
The temples housed images of the gods, religious paraphernalia,
priests, and the sacred women, but the ceremonies themselves were
held out-of-doors.
The priesthood, headed by a close relative of the Emperor, at times
may have rivaled the civil authorities in power (Handbook, 2 298 ff). :
Under the high priest were the priests in charge of the various temples,
their importance varying with that of the temple, and under them
were a great number of assistants.
Ceremonies were held according to a ritual calendar (Handbook,
2: 471). Most common rites were agricultural, but some honored the
various gods and others were held against sickness. Special rites
were performed in times of crisis. Under the leadership of the priests
and their assistants, the gods and supernatural powers were propitiated
with blood sacrifice, prayers, fasting, and offerings of coca, sea shells,
chicha, and other things. Sacrifice was usually of llamas and guinea
pigs, but in times of crisis children were strangled and their hearts
offered to the god. Similar ritual was used in a smaller degree for
the lesser spirits and shrines, to which anyone could make prayers,
offerings,and the like.
Personal sins were believed responsible for sickness and misfor-
tune, and a task of the priests was to confess sinners and impose pen-
ances.
In addition to the use of shrines and idols as oracles, divination was
practiced by sorcerers, who communicated with their spirit helpers,
made spirits speak from fire, examined lungs, seeds, or coca, observed
the movements of birds or animals, and interpreted omens and dreams.
It would seem that shamans, that is curers and diviners, were distinct
from priests. Among the Inca^ curers obtained power from a vision
of a spirit helper or through rapid recovery from disease. Disease was
thought to be caused in several ways punishment by the gods for sin,
:
sorcery, winds, evil forces, and soul loss brought on by fright. The
concept that sin causes disease represents a socialization of more primi-
tive taboo transgression theories. The state was a socioreligous struc-
ture, with a divine Emperor at its head, and social offenses became
also religious, disease-causing transgressions. (See Ackerknecht, this
volume, p. 633 ff ) Disease was manifest as a foreign object in the body,
.
displaced organs, or a toxic condition. The curer sacrificed to his

visionary power, divined the cause of the illness, prayed and sacrificed
to the Sun, heard confessions, prescribed the penance or procedure the
patient should follow, and finally sucked out the object, rubbed the
body, or used other appropriate means.
Modern Aymara curers do not have spirit helpers they are persons
;
who have been struck by lightning or who have learned their art.
They practice both black and white magic.
Andean ritual elements included fasting, praying, offerings, blood
sacrifice, washing (AyTnara), use of chicha and coca, the concept of
sacred directions, and three as the ceremonial number.
TECHNOLOGY AND MATERIAL CULTURE
The Central Andean peoples were distinguished from the remainder

of South America by the number of processes they employed in manu-
factures and by the quantity and quality of their production rather
than by essential technologies. They are also distinguished, especially
from the Marginal and Tropical Forest peoples, by the specialization
of production and consumption according to the social and political
system. Simple goods and buildings were used by everyone, but most
of the finer constructions and craft products were destined for the
government, priesthood, and nobility.
Building arts. —The finest construction of the Central Andes was

devoted to public buildings — ^palaces, temples, warehouses, and forts
built of carefully dressed and skillfully laid stones. Construction em-
ployed lintels, small corbelled arches, columns, and occasionally two
and three stories. Commoners' homes were small, one- or two-room,
stone or adobe houses. The stone-faced agricultural terraces, the
tremendous system of irrigation canals and ditches, and the canalized
rivers were all community works. So were the paved and graded
roads which ran throughout the Empire, crossing chasms on suspension
bridges and swamps and rivers on pontoons.
—
Metallurgy. The Central Andean peoples differed from those of
the Sub-Andean and Circum-Caribbean area in that they not only
smelted and alloyed gold, silver, and copper but they used platinum,
mercury, lead, and arsenic and made various alloys, including bronze.
They also employed a larger number of metallurgical processes (Hand-
book, 2 246-248). The art products and ornaments of precious met-
:
als were intended mainly for the nobility, but, in contrast to other
American Indians, there were many metal utility objects which anyone
might use, for example, knives, axes, bolas weights, clubheads, mirrors,
and tweezers.
needles, bells, crowbars,
—
Weaving. The people of the Central Andes had three types of
— —
true loom horizontal, vertical, and belt and they were the only
aboriginal people in the New World to use wool on an important
scale. In addition to domesticated cotton, they wove with llama, al-
paca, vicuna, and bat wool. Their textiles are unequalled for their
fineness, number of techniques, and variety of woven-in designs. The
finer ornamental products were made by special artisans for the
nobility.
—
Ceramics and containers. The pottery art was no less outstanding
than the textile craft, for the Central Andes have yielded wares of all
—
known American techniques ^molding, modeling, and other plastic
treatment, negative painting in three colors and polychromes in as
many as 11 colors. The Inca ware, reflecting standardized mass-pro-
duction, fell short artistically of earlier ceramics, but technologically
it equalled any of them.
The Central Andes is also notable for its carved and lacquered
wooden cups and for its painted gourds.
—
Basketry. Basketry is little known; apparently the Indians de-
voted the greater part of their attention to ceramics. Twined, twilled,
and coiled baskets are all known archeologically and may have been in
use at the Conquest. Coiling is still used by the Aymara. Mats were
also woven.
—
Weapons. Central Andean weapons seem to have changed little
throughout the archeological history of the area, perhaps because they
;
were adapted to open country. Hunting weapons included slings,

bolas, nets, snares, nooses, and The bow and blowgun were
clubs.
absent at the Conquest. In warfare, spears and spear throwers were
used for preliminary attacks, but they were superseded in the High-
lands by the sling and bolas. For close fighting, star-headed clubs of
stone and metal, macanas, and copper or bronze axes were wielded.
Protective devices included quilted-tunic armor, slat shields, and hel-
mets. It is difficult to see why the bow was not used. Suitable bow
wood cannot be had in the Andes, but presumably it, like many other
products, could have been obtained in the Montana.
—
Transportation. Êffective transportation was provided on the
north coast by sea-going balsa wood rafts that carried 50 men
and were propelled by sails and paddles, and on the south coast
by inflated seal-skin rafts. In the interior, human carriers, who used
square clothes slung on their backs, and llamas, which could bear 100
pounds, transported goods over the roads.
—
Household furniture. Houses were equipped not only with cook-
ing utensils and other articles, but they had wooden stools, those of
the nobles being elaborately carved. People usually slept on the floor,
though the platform bed may have been used in some regions. The
hammock was unknown except as an occasional substitute for the
litter in which to transport nobles.
Clothing. —The basic garments for men were a breechclout, sleeve-
less tunic, cloak, leather sandals, and head band, and for women a long
dress, a mantle pinned at the chest, sandals, head band, and head cloth.
Details of dress indicated class status. Bracelets and necklaces were
worn, but lip and nose ornaments seem not to have been common;
earplugs were used by the nobles. There was some head deformation
but it seems to have been a trait of local custom, not a mark of class
it also occurred in early times.
ESTHETIC AND RECREATIONAL ACTIVITIES
In esthetic and recreational activities it appears that many of the

refinements were associated mainly with the nobility and the church.
Art manifestations were principally on containers, ornaments, and
textiles, which embellished the persons or buildings of the noble and
priestly classes. Chicha was a libation and a ceremonial drink as well
as a beverage of popular consumption, coca was used especially by the
nobles, priests, and diviners, and even tobacco and Piptadenia colwm-
hrina, another narcotic, were probably employed exclusively for reli-
gious purposes.
Musical instruments, though perhaps used by commoners for their
own enjoyment, were employed mainly for military affairs and
religious festivals. These included end-flutes; bone flutes; plug

flutes; panpipes; trumpets of shell, of wood, and of gourds; tam-
bourines; two-headed drums; metal bells; and shell rattles.
Games were little developed, except dice games, which were the
occasion for betting. A
ball game and various athletic sports and
contests were common, and children played with a number of toys,
including tops, which were whipped.
Literature in the form of ballads, created by bards serving the
nobles and priests, and secular plays contributed to the entertain-
ment of the upper classes.
THE HISTORY OF SOUTH AMERICAN CULTURES

THEORETICAL CONSIDERATIONS
In the theoretical approach to problems of American culture

history there are important differences between those who postulate
that the principal Indian cultures originated in the Old World and
those that they were indigenous, New World products.
who hold
Theories deriving Indian cultures from outside the Western Hemi-
sphere range from sensational accounts of the wholly imaginary
"lost continents" of "Mu" in the Pacific and "Atlantis" in the Atlantic
to more plausible treatises claiming that basic culturesand civiliza-
tions were carried by trans-Pacific migrations from China, Indonesia,
India, or the Near East. Some writers claim only that individual
culture elements reached the shores of America from overseas.
The "lost continent" theories do not merit consideration. Some
of the other theories have perhaps received more attention than they
deserve, but interest in them is kept alive by our still very insufficient
knowledge of early American developmental stages, which clial-
lengingly has been used as proof that they are to be found outside
America. Among the more influential and serious writers are
Elliott Smith (1929) and Perry (1926), who derive New World
civilizations from Egypt; Graebner, Schmidt, and their followers,
who postulate that a series of kulturkreise, or cultural strata, each a
more advanced cultural stage, were brought from the Old to the New
World; and Kivet, who has attempted to show the trans-Pacific
affiliation of certain Indian races, languages, and cultures. (See
Schmidt, 1913; Rivet, 1943.)
These theories, especially the kulturkreise, assume: (1) that cul-
tural classification based on Old World data are valid for America, and
(2) that each culture consisted of a complex of political, social, and
religious institutions, economic activities, and material elements, which
diffused as a whole though the cultures interblended subsequently. A
corollary though unstated assumption is that the American Indian
lacked the capacity to create what are assumed to be fundamental

features of culture. They pay comparatively little attention to the
relative chronology of Old and New World culture history, or else
they assume dates for the latter which cannot be proved, and they
ignore geographical and cultural barriers to culture diffusion.
American scholars find that these theories are too deductive and a
priori.^ Concerned for two generations with intensive field work,
they have preferred to classify Indian cultures in terms of New
World, not Old World, data. Schmidt's primitive cultural stratum
of Food Gatherers, for example, is meaningless for America because,
as we have seen, the Marginal peoples are highly diversified and have
little in common with one another. His farming, mother-right stage,
when woman were cultivators, simply does not fit the Tropical Forest
peoples, who were overwhelmingly patrilineal. The assumption that
herding was associated with male dominance is not valid for the
Andes, for it would be a complete misunderstanding of Andean cul-
ture to suppose that llama herding was at all comparable to Old
World cattle raising and that it accounted for patrilineal tenden-
cies in the Central Andes. Similarly, any theory that classifies New
World civilizations with those of the Near East ignores the vital
differences in culture patterns and element content.
The New World cultures not only differed from those of Africa
and Asia, but they differed from one another in complex ways. Most
particularly, the elements and patterns had very different distributions
and, inferentially, different histories, so that the problem is not to
trace total cultures through time and space but to ascertain the history
of each element, element complex, and pattern. American anthro-
pologists, therefore, deal with each on its own merits. Approaching
the problem empirically, they study the manifestation of each cultural
feature in specific situations before theorizing about its broader time-
space relationships, and they prefer to reconstruct local segments of
history before constructing world schemes. They have been unwilling
to concede that the American Indian was wholly devoid of inventive-
ness, and they insist on examining the data of archeology for evidence
both of local development and of the relative Old and New World
chronology of features shared by both hemispheres. They ask for
concrete proof of how migrations could have occurred and how geo-
graphical barriers were overcome, especially in the period before
ocean travel developed.
Few American anthropologists deny the possibility of transoceanic
influence on New World cultures, though most of them repudiate the
theories that bring total cultures from overseas. It is conceded that
^ For further methodological analyses see Cooper (1941), Lowie (1937), and Dixon
(1928).
738931 —49 49
individual elements and even groups of elements may have been im-
ported, though absolute proof is difficult to produce in any single
case. At one extreme are features such as matrilineal descent, moon
worship, moieties, megalithic monuments, agriculture, or writing,
which are not really culture elements or patterns but classificatory
labels. To derive phenomena with such labels in America from simi-.
larly tagged Old World phenomena would be to ignore the realities
of culture. At the other extreme are certain American domesticated
plants, whose identity and genetic connection with Old World species
can be established beyond reasonable doubt. On botanical evidence,
the pre-Columbian occurrence in American and the Old World of
sweet potatoes (Maori varieties), an edible root {Pachyrhizus) cala- ^
bashes {Lagenaria) north Peruvian cotton {Gossypium) plaintains

,
^
{Musa paradisiaca normaUs) and perhaps peanuts and coconuts in-

,
dicates a diffusion from one hemisphere to the other. Between these

extremes are a very large number of dubious cases, which involve sub-
jective judgments and which have entailed considerable controversy.
(See Dixon, 1928.) Wliether these elements are attributed to a single
world origin or to multiple origins depends upon one's opinion con-
cerning their uniqueness and man's inventive capacity. Some ele-
ments, such as thatched, frame houses, penis covers, or crutch-shaped
paddles, would seem simple enough to have been invented several
times. Others, such as the blowgun, bark cloth, panpipes, chewing
lime with a narcotic, star-headed clubs, trepanning, the venesection
bow, and the ikat weaving technique, might be argued with some
plausibility to have been invented only once, but it would still be
necessary to prove that they were earlier in the Old World than in
the New World and to show how they could have reached America.
A good many of them might have been introduced to America by
individual boatloads of voyagers in the course of settlement of the
Polynesian Islands during the Christian Era. It is somewhat more
doubtful that voyagers reached the shores of America in the earlier
milleniums, when the American civilizations were taking form.
The history of these many moot items, however, is not very impor-
tant to an understanding of the development of American cultures,
for, as has been shown previously, they occurred largely irrespective
of the main cultural patterns, all of which developed in distinctive
American traditions. The patterns found among the Marginal peoples
clearly stemmed from local adaptations beginning in early post-
Glacial times, before trans-Pacific navigation was conceivably pos-
sible. Those of the remaining tribes, including the civilized peoples
of Mexico and Peru, acquired distinctive American patterns at a
very early time. They developed along with the cultivation of the
more than 100 native American crops on which they were based, and
these crops required centuries if not milleniums of plant breeding.

None of them, except the few plants mentioned above, came from the
Old World, and it can hardly be conceived that the idea of agriculture
lay dormant in Indians during hundreds or thousands of years while
they migrated from the agricultural areas of the Eastern Hemisphere
across Siberia, Alaska, and southward into temperate America. Our
present ignorance of the formative beginnings of these civilizations
is not proof of their Eastern Hemisphere origin ; it simply means that,
until recently, attention has been accorded their elaborate manifesta-
tions, not their primitive origins.
American anthropologists, though in general accord in minimizing
the importance of transoceanic influences on American Indian cul-
tures, disagree somewhat on problems of the time and place of cultural
origins within America and the directions of cultural diffusion. Their
principal methodological tools have been the culture-area concept, the
age-area hypothesis, and archeological sequences. All the tribes in a
culturally homogenous area are held to have derived their culture
from the same source, which is assumed to be the place with the greatest
complexity of structure or with the greatest element content and which
is known as the culture center. The age-area hypothesis holds that,
other things equal, the oldest cultural features are those having widest
distribution. Where a distribution is interrupted, it is assumed that
where the features are absent they were superseded by other features.
Archeology affords verifiable culture sequences, but these are usually
in terms of culture elements rather than patterns, which are more
difficult to abstract from its data.
For the New World as a whole, the following history is generally
accepted by American anthropologists. As the oldest and more elab-
orate cultures based on farming are found in Mexico and the Central
Andes, it is believed that one of these areas was the center of origin
of an early, ill-defined. Formative Period culture from which the
main complexes and patterns found amon^ all farming tribes were
derived. During subsequent development, both Mexico and Peru be-
came specialized as secondary centers. What might be called tertiary
centers developed in many places, such as the Southeastern and South-
western United States, the Circum-Caribbean area, the Southern
Andes, Northwest Argentina, and the Amazon Basin. When similar
culture elements and patterns are found among farmers at the north-
ern extremity of the farming area in North America and the southern
extremity in South America but not among intervening tribes, the
age-area hypothesis would hold that they are survivals of what was
present during earlier periods in the culture centers and throughout
the intervening areas. The hunting and gathering, or Marginal, peo-
ples located beyond the agricultural tribes in both continents are as-
sumed to represent survivals of an earlier, preagricultural period that

was also once pan- American.
The difficulty with this methodology is that sociopolitical patterns
are treated as if they were comparable to culture elements and as if
they occurred irrespective of the cultural ecology of each area. Certain
ornaments, ritual details, items of material culture, social usages, and
many other elements did occur in very different cultures and environ-
ments and their antiquity may be roughly estimated by the age-area
hypothesis. The history of sociopolitical patterns, however, cannot be
reconstructed by any such reasoning. Many of these patterns were the
result of ecological adaptions to local environments as much as of
historical influences and tradition. At one extreme, the sociopolitical
structure of the group seems to have grown very directly out of the
interaction of subsistence habits, settlement type, settlement composi-
tion, and socioeconomic activities. At the other, where the ecology
afforded potentialities for a variety of patterns, the tribes were more
susceptible to historical factors and the age-area method may be used
within limits to ascertain the succession of sociopolitical patterns.
Applying the age-area hypothesis to American data, it can be shown
with reasonable certainty that the Marginal tribes of both North and
South America had certain elements, such as flint chipping, and certain
institutions, such as crisis rites and shamanism, which were the heritage
of a primordial period. Some of these retained a pan-American dis-
tribution others survived only among the Marginal tribes and were
;
superseded elsewhere. None of the dozens of different sociopolitical

patterns of the Marginal peoples, however, was the survival of a
primordial pattern and, therefore, antecedent to the Tropical Forest,
Sub-Andean, or Central Andean patterns. Simple patterns, such as
those based on kin groups, unquestionably preceded more complex ones,
such as those including both kin groups and social classes, but the spe-
cific nature of the earlier forms cannot now be reconstructed with cer-
tainty by any method. Thus, in terms of sociopolitical patterns, the

Marginal tribes represent neither a cultural type nor a cultural stage,
but rather a class of tribes which had in common only the lack of the
distinctive patterns of the remaining peoples. It is only in terms
of certain technologies, material elements, and isolated ritual and
social usages that they can be said to retain a primordial culture.
Where the combination of environmental potentialities and sub-
sistence techniques allowed considerable latitude in the development
of sociopolitical patterns, the age-area hypothesis is more useful in
reconstructing the history of these patterns. Both the Andes and
the Circum-Caribbean area, for example, had an efficient economy, a
dense population, and large population centers. As certain patterns
are found in both, it is reasonable to assume that the simpler forms
of the Circum-Caribbean area were derived from the more complex

ones, like those of the Andes, and, though the former were somewhat
adapted to local environments, they retained many features found in
an earlier period of the Andes. This earlier period culture, in fact,
probably extended from the Andes to the Mexican and Mayan areas.
The following pages will sketch the outlines of South American cul-
ture history from the earliest periods to the Spanish Conquest. Evi-
dence on the antiquity of man in America is drawn principally from
North America, where the record is better known. The earliest culture
will be reconstructed from archeology and, so far as the age-area
hypothesis is from ethnography. The development and
applicable,
spread of the cultures based on farming will then be traced. A subse-
quent section will outline post-Conquest changes among the Indians.
ORIGINS OF NEW WORLD CULTURES

RACIAL ORIGINS
American Indian is predominantly Mongoloid. He

Racially, the
is, most closely related to the peoples of Asia and was
therefore,
probably derived from Asia by way of Bering Strait. Some anthro-
pologists have seen Mediterranean, Melanesian, and other non-Mongo-
loid strains in the Indian. This thesis cannot be disproved, but,
unless it be postulated that these strains came from comparatively
recent arrivals, it involves a genetic assumption that is difficult to
defend that several racial prototypes could intermix for hundreds
:
if not thousands of years and the genes subsequently segregate out in

the original combinations, a chance of one in thousands.
Though predominantly if not entirely Mongoloid, the Indian was
extremely variable in head form, stature, and other characteristics.
(See McCown, Stewart, and others in volume 6 of the Handbook.)
ANCIENT MAN
Many authors claim to have found evidence of ancient man in the
Western Hemisphere, but few claims have withstood all criticism.
The evidence is usually of three types: (1) typological. New World
artifacts being accorded an antiquity comparable to their Old World
homologues; (2) association of human remains with locally extinct
fauna or flora; and (3) human remains dated in terms of geological,
especially glacial, chronology.
Mere typological evidence has been generally rejected. Lithic types,
which have been accredited with an antiquity comparable to their
first appearance in the Old World sequence —
ûsually a period of the
Paleolithic but according to some, even the "Eolithic" —often survived
in America in comparatively recent cultures. Some early American
periods have certain features in common with Eastern Hemisphere

periods, but none of the Old World Paleolithic or Neolithic cultures
are present in typical form in the New World, and there is no reason
to suppose that America had a cultural sequence that corresponded
precisely in either content or chronology to that of the Old World.
That human remains are associated with extinct fauna and flora is
not per se evidence that they had great antiquity, for the extinct species
may have survived locally until quite recently. Such associations have
greater significance when the extinct species is evidence of important
climatological change.
Attempts to place human remains in major geological periods have
yielded no generally accepted results because, first, the time calibra-
tions are too gross to be of value in cultural sequences, and second,
the presence of human remains in older geological strata often has
been found to be a later, secondary intrusion. Satisfactory field tech-
niques have not always been utilized to demonstrate the contempora-
neity of cultural and geological remains. The correlations of human
remains with glacial periods, however, is susceptible to finer calibra-
tion, and certain correlations of cultural materials with periods of
great climatological and physiographic changes in North America are
now generally accepted. Though opinion concerning the absolute
dates of these changes is not yet unanimous, the margin of disagree-
ment being steadily narrowed.

is
To judge by the correlations of human remains with glacial and

postglacial phenomena in North America, man unquestionably was
present in the New World at the close of or during the last phases of
the final glacial period, which was certainly 15,000 and possibly 35,000
years ago. The Folsom bison hunters, represented by a lithic industry
with some Upper Paleolithic features, are definitely known to have
lived in the western United States during the final phase of the Wis-
consin glaciation ( Roberts, 1945 ) . The oldest remains in Sandia Cave,
New Mexico (Hibben, 1941) are probably earlier than Folsom. It
is presumed that man reached America from Asia via the land bridge
which linked Siberia and Alaska during the glacial period. (See also
Sauer, 1944.) Ocean navigation was certainly not developed at this
time the Polynesian migrations in the Pacific occurred mainly during
;
the Christian Era, especially the last millenium.

No remains of Central or South America have yet been shown to
have antiquity comparable to those of North America. It should
be remembered, however, that there has been comparatively little
search for ancient man in South America and that, before the Folsom
remains were found 20 years ago, orthodox opinion assigned the Indian
a maximum antiquity in North America of only 5,000 to 8,000 years.
Central and South America, however, have remains of pre-agricul-
tural periods. The human footprints found in the volcanic mud in

Nicaragua almost certainly antedate the earliest Mayan civilizations.
The oldest remains of southern Patagonia and at Magellan Strait
—
may plausibly be assigned an antiquity of 5,000 years 2,000 is the
—
minimum and correlations of these with glacial periods may re-
quire that their age be extended (Handbook, 1:21-22). Human re-
mains have been found associated with extinct fauna at various sites,
but the fauna is rarely dated with any certainty. The Lagoa Santa
or Confins man, however, seems to have been relatively primitive
(Handbook, 1: 399), and the associated human and extinct sloth
remains in Tierra del Fuego (Handbook, 1 22) may have fairly
:
great age.
None of these primitive South American remains resembles the
oceanic cultures, but they are generally similar to preagricultural
materials in North America. As there were no serious obstacles to
man's gradual migration southward from North America, it is far
simpler to derive the Indian of South America from this source than
to postulate that he came by trans-Pacific migrations.
EARLY AMERICAN CULTURES
The data of both archeology and ethnology now leave no doubt

that man entered America from northeastern Asia, long before
first
agriculture was known anywhere in the world, and that he possessed
a Stone Age culture. He spread slowly throughout the hemisphere,
from Alaska to Tierra del Fuego, adapting his mode of life to local
environments with the aid of a simple technology and a limited num-
ber of material aids. Subsequent to the original migrations from
Siberia and prior to the voyage of Columbus, Old World influence
on the New World was very limited. Asiatic influence is evident
on the northwest coast of British Columbia and Alaska; some ele-
ments have a circumpolar distribution around the subarctic, and a
few traits, such as the bow, spread very rapidly and widely in Amer-
ica. Within the last two or three thousand years, other traits (p. 744)
may have spread directly across the Pacific to the western shores of
America.
Certainly no less than 3,000 years ago, and probably much more, the
Indian began to bring native American plants under domestication.
The cultivation of these plants spread widely, but did not reach the
tribes in the northern and western jDarts of North America and in the
southern and some of the eastern parts of South America. The peoples
beyond the limits of agriculture remained Marginal, and they retained
fairly primitive cultures, which, though in no case precisely like that
of the first immigrants to America, were comparable to it in general
simplicity. A comparison of these cultures together with the archeo-
logical evidences of early man gives a general picture of what once

was probably pan- American in distribution. But it is impossible to
reconstruct the original sociopolitical patterns, except in terms of ab-
sences, such as social classes. The early culture must be reconstructed
largely in terms of isolated elements, and many of these, though very
old, may
date from anywhere in the thousands of years of prehorticul-
tural times for which there is scarcely any archeological record.
—
Technology and material culture. To judge by the Period I
remains of the middens of Tierra del Fuego (Handbook, 1 17 ff.) and :
by the Folsom jfinds, the earliest Indians knew the use of fire, pressure
flaking of stone, and the process of cutting and grinding bone. They
—
used some kind of projectiles possibly but not certainly spears
which were tipped with stemless, chipped-stone points, and they
worked skins with stone scrapers and bone awls.
On the other hand, the very considerable list of traits found among
the Tropical Forest, Circum-Caribbean, and Andean tribes but not
among the Marginal peoples, were almost certainly absent from the
early American cultures. These include: domesticated crops; traits
of food preparation, such as use of salt, babracots, metates, and mor-
tars; technologies, such as ceramics, loom weaving, basketry, true
tanning of skins, and metallurgy and esthetic and recreational traits,
;
such as musical instruments, chicha, tobacco, and other narcotics and

stimulants; many elements of warfare, including cannibalism and
human trophies; and temples, priests, idols, human sacrifice, and a
ritual cycle.
Sociopolitical —
and religious patterns. No special sociopolitical
patterns can be shown to have characterized the early American cul-
tures. We have seen that at the Conquest the Marginal tribes had a
wide variety of patterns which reflected their ecological adaptations
no less than their cultural heritage from the past. An effort to
reconstruct the early patterns on a split-distribution basis by drawing
parallels between the Marginal peoples of North and South America
would be very misleading, for the similarities are quite superficial
in most cases.
Among the Shoshonean tribes of the Great Basin of western North
America and among the Namhiciuira, society was based on the indi-
vidual family, but the socioeconomic activities of the two areas were
very different and so were the larger, looser, multifamily aggregates
that occasionally assembled. The patrilineal bands of the Ona (60 to
100 persons), each with a hunting territory for big game, were only
superficially similar to the conjugal family groups of the Algonquians
of Canada, each of which owned a territory for trapping fur-bearing
animals. The large, caribou-hunting Athabaskan bands (200 to 250)
of Canada had no parallel in South America, and the matrilineal vil-
,
lages of the Chaco have no precise North American counterpart. The

main between North and South American Marginal tribes,
similarities
in fact, occur where the ecological adaptations were similar. For
example, patrilineal bands were found in southern California and in
—
Tierra del Fuego ^both hunting areas but without large, migratory
game herds. Mixed bands, i. e., multilineage or multifamily bands,
of some size were found in the Great Plains of North America and
Patagonia, and in both areas they were post-Columbian and were
based on use of the horse for hunting and traveling.
The most that can be claimed for the earliest American tribes is
that the people lived in small, generally somewhat migratory, groups
which probably consisted of kin, that they lacked developed chieftain-
ship and social classes, and that they were structured on the basis of
sex, age, and socioeconomic activities.
There are, however, a number of socioreligious features which are
virtually universal among primitive peoples and may be accredited
with great antiquity fasting by the mother of a newborn child the
: ;
isolation and fasting of a pubescent girl earth burial belief in a soul

; ;
and a life after death belief in supernatural beings belief in witch-

; ;
craft and omens; shamanism (but see p. 588) and shamanistic curing
;
by massage, blowing, and sucking.

Some individual culture elements. —A considerable number of
culture elements which were frequently found among the Marginal
peoples of North and South America but less often among the farm-
ing tribes have been interpreted as possible ingredients of an early,
preagricultural period. (See Nordenskiold (1931), Krickeberg
(1934) Ploetz and Metraux (1930) references cited by Cooper (1942)
, ,
and Handbook, 2: 213-214.)

This application of the split-distribution method can be accepted
only with certain qualifications. First, as Cooper (1942) points out,
some of these elements may well have been invented independently in
North and South America in response to environmental needs, so that
their distribution is no evidence of their antiquity. Second, some ele-
ments may have developed as part of the agricultural complex and
spread beyond its limits, later being superseded among many farming
tribes. Such, for example, appears to have been the case with coiled
basketry and netting, though older than loom weaving, is not neces-
;
sarily pref arming. Third, terms such as "archaic," "preagricultural,"

and "early" are altogether too general. Presumably, the preagricul-
tural period in America lasted some 10,000 to 15,000 years, during
which culture became considerably specialized in the different areas.
Finally, few of the traits accredited with great antiquity in America
occurred among all Marginal peoples. Rather, they seemed to cluster
in certain tribes and groups of tribes, and it would be wholly unwar-
ranted to assume that they were once pan-American in distribution and

have since inexplicably disappeared among many Marginal tribes.
Certain games, for example, link the Gran Chaco and North America,
but they are wholly absent in the Chilean Archipelago.
The most fruitful working hypothesis is that American culture his-
tory was very complex. To understand this history we need evidence
of period and regional differences derived from detailed archeological
sequences and ethnographic comparisons. A
certain clustering of
these early traits, for example, suggests that there were at least two
streams of influence in South America, one carrying certain complexes
down the Andes and the other entering the eastern part of the
continent. Wliere archeological sequences are known, the culture
appears to have arrived in a succession of waves. The following list
shows the clustering of some of the elements that occurred also among
North American Marginal tribes
—
Araucanians. Shaman as transvestite (also Puelche) use of tamborine; and
;
pubescent girls race at dawn and carry fire wood (also West coast of North
America).
—
Parand River. Finger mutilation as evidence of mourning; placing skewers
through the skin and fasting for a guardian spirit (all Charrua and especially
;
Plains in North America).

—
Chaco. Such myth motifs as the trickster, the vagina dentata, and the theft of
fire, the last also Montaila, eastern Brazil (Metraux, 1939 see also Handbook,
;
1: 3G9) ; such games as ring-and-pin (also Montaiia), dice (also Andean),

"snow-snake" (AMpon, Mocovi, also eastern Brazil), gambling (also
Andean), the musical bow {Mataco, Tola, Lengua, Guand), and scalping
(also Guianas).
—
Elements of a s'potty distribution. Some elements were fairly wide-
spread among North and South American Marginal tribes, though
not common to all of them harpoons, spear throwers, nets, traps, fish-
:
hooks, body painting, depilation, some kind of secret society, usually

for men {Ona., Chamacoco., eastern Bolivian and Jurua-Purus Arawak,
Northwest Amazon, Bororo), and the belief that disease was caused
by the intrusion of a foreign object into the body.
Features of more limited distribution include such elements as the
ritual drinking tube (Yahgan), head-scratcher {Yahgan, Tucanoans,
C/ioco, Apinaye, Oanella and other Ge) steam bath {Puri-Ooroado,
,
Botocudo, Araucanians), fasting for a guardian spirit (Charrua),

arrow swallowing {Tehuelche, Tereno), certain musical styles {Yah-
gan), deer-hoof rattles (Chaco, eastern Brazil), the soul-loss theory
of disease (Andes, Chaco, eastern Brazil) the bull-roarer {Sherente,
,
Apinaye, and Mataco as a toy; Choco, Aymara, Bororo, Mashacali,

Bolivia, Antilles, Araucanians, Central America, Guianas, Apinaye) ,
hoop-and-pole game {Sherente and some Guiana, Amazon, Chaco, and

Fuegian tribes), hockey (Chaco, Tehuelche, Araucanians, Northern
Cayapo)^ snow-snake (Chaco, eastern Brazil),^ boiling water with

hot stones (Aweicoma, Chono, Caingang^ Shokleng^ Botocudo^ Pv/ri),
heating water with hot stones {Yahgan, Ona) earth oven {^^Tapmja^''
;
Puri-Coroado^ Gaingang^ ShoMeng, Aweicoina, Chaco, Argentina,

Northwestern Ge, Southern CayapS)^ and strike-a-light
(Archipelago).
THE CULTURES BASED ON AGRICULTURE

ANDEAN CULTURE DEVELOPMENT^
On the basis of present evidence, American plant domestication ap-
pears to have been undertaken independently in many places and at
different periods. Some such as maize, beans, and squash,
species,
attained wide pan-American distribution, but on the whole, each
geograj)hical area had, in addition to these plants, its characteristic
complex of staple crops the intermediate and low Andes, maize and
:
other cereals; the high Andes, potatoes and quinoa; the Tropical
Forests, root crops and the Northern Andes, many fruits. The oldest
;
evidence of agriculture is in the Central Andes, where it is conserva-

tively assigned an antiquity of about 500 A. D. (Handbook, 2: 80).
Some authorities believe that it should be dated 1,000 years or more
earlier; in first evidence of farming is probably
Mexico the
much Thirty-one domesticated plant species are known from
later.
the Early Periods of Peru, and probably most of them were cultivated
during the preceding Chavin Periods. Several of these plants, such
as lima beans, had already reached the limits of their genetic vari-
ability in the Early Periods, implying a long antecedent period of
crop breeding. The domesticated llama and alpaca were also present
in the Early Periods.
Botanical data suggest that some area to the east of the Andes may
have been the locale of the first cultivation of the principal Andean
crops, especially maize. Eastern Bolivia, with its milder and more
varied climatic zones, is a likely place, and its archeology has rich
remains that may include those of very early agricultural peoples.
The archeology of eastern Bolivia, however, is scarcely known, and
early sequences have not yet been established. It is necessary, there-
fore, to look to Peru with its long chronology, for a picture of the ear-
liest cultures based on farming. The oldest Peruvian culture yet
identified is that of the Chavin Periods, which may have covered a
" For games and gambling, see Cooper, this volume, p. 503.
^ For the sequence of Andean Periods, see Handbook, vol. 2, p. 80. Research since 1945,
especially the Virfl Valley Project in Perti, has extended cultural sequences back to a
preceramic, agricultural period, corrected the order of the periods as given in the Handbook,
and thrown new light on the development of social, religious, and military patterns. (See
Steward, 1947, 1948 Strong, 1947 Willey, MS.)
; ;
considerable time span. The remains of the Early Periods, which

followed the Chavin, give a more complete picture of life at that time.
In the Early Periods, agriculture was already well established in
Andean patterns, even extensive irrigation works being used. Farm
productivity evidenced not only by the fairly dense population but
is
by the development of manufactures, the building arts, and religion

and warfare, which took many persons from subsistence activities.
Transportation was no less efficient than in later periods, for huge
balsas were used for coastal travel and llamas packed goods over roads
on the land. The precise settlement pattern has yet to be ascertained,
but communities appear to have been small and dispersed, though
closely spaced, and if they were nucleated at all, it was around religious
centers, such as the classical Highland site, Chavin de Huantar.
There was certainly no urban planning, as at later sites like Chanchan.
Habitations of the Early Periods occur either as isolated dwellings
or as loose clusters of dwellings forming small villages any planning
;
Mâs of religious rather than civil centers.

Evidence of a stratified society, developed warfare, and a priest-
temple cult is unmistakable in the ceramic decorations and representa-
tions (Handbook, 2 149-182) especially of the Mochica Period.
: ,
These patterns must have resembled those of the later Andean peo-
ples, and they appear to have been even more elaborate than those of
the Sub- Andean and Circum-Caribbean areas.
Pottery designs show chiefs or nobles with characteristic evidences
of rank wearing special dress and insignia, sitting on thrones, being
:
carried in litters, receiving obeisance from kneeling or bowing sub-

ordinates, being served by retainers, and meting out punishment.
Considerable governmental control is implied by the construction of
great mounds, buildings, irrigation systems, roads, and other public
works, by the use of runners or messengers for communication, and
by the public imposition of punishment.
Warfare was strongly developed. Armies wearing special dress
and using all the principal weapons of later periods fought under
powerful war chiefs. Captives were sacrificed, probably in religious
rites, by being thrown from mountain peaks, or they were made into
human trophies, especially head trophies. Cannibalism is not shown,
but the quartering of prisoners is suggestive of the Tupian cannibalistic
festivals.
The three-story stone edifice at Chavin de Huantar and the temple
mounds of the Early Periods were almost certainly religious centers.
The feline, serpent, and condor deities which are represented in so
much of Peruvian art were undoubtedly among the principal tribal
gods. Various other nature gods are also portrayed. An important
class of priests officiated over public worship, but it is not known
whether they were also shamans and whether they had civil power.
There was also ancestor worship, or a cult of the dead, as evidenced
by the elaborate mummy burials. Shamanism was probably present
with its familiar patterns, for shamans (possibly priests) are shown
shaking rattles and, less often, massaging and sucking their patients.
On the technological and material side, the essential processes and
complexes were thoroughly established by the Early Periods, and
craft production seems to have been specialized. Painted, plastic, and
modeled ceramics equalled those of later periods. All weaving tech-
niques were known, and both cotton and wool were used. Basketry
was made, evidently at first twined and twilled (Early Ancon and
Supe), and later coiled; if correct, this sequence reverses the usual
one elsewhere in America. The weapons used in Peru at the Spanish
Conquest were already present: darts, spear throwers, clubs, and
slings. The Mochica also had the blowgun and dart. The only in-
dustries to show technological advancement subsequent to the Early
Periods are masonry and adobe construction and metallurgy. Chavin
used only gold; Mochica added silver, platinum, and copper; only
the Middle or Late Periods achieved bronze.
The Chavin culture, though the earliest known in Peru, is far from
primitive or even formative, and it probably had the basic features of
the succeeding Early Periods. Its subsistence was adequate, its
technological and material culture contained virtually all essential
elements and complexes of later periods of the Andes, its sociopolitical
and religious patterns were clearly developed far beyond those of the
Marginal or Tropical Forest peoples, and its esthetic attainments show
considerable sophistication. Although its ultimate origins are still
obscure, it is difficult to imagine that it was suddenly introduced full-
blown from across the Pacific. Instead, it must have developed along
with the indigenous American crops, on the cultivation of which it
was based.
Cultures of the Chavin and Early Periods would seem to have con-
tained the generalized patterns that were later to take more definite
form. Undoubtedly there was some local variation, but the main
features were more or less pan-Andean. If San Agustin has antiquity
comparable to that of Chavin, a similar pattern probably existed in
Colombia. In fact, it probably extended northward to Mexico, and
the basic Maycm pattern of a population loosely aggregated around
great mound and temple sites seems to represent a special emphasis on
the socioreligious components of this culture.
Subsequent Andean trends bring out the potentialities implicit in
the early patterns. Population evidently increased considerably, for
in the Middle Periods hillsides were terraced for farming purposes.
After the Early Periods, civil and military affairs were developed on
a larger scale, a development no doubt partly explainable by the pres-

sure of the growing population. An increasing proportion of con-
structionwas devoted to building forts, and, to judge by the i^re-Inca
or Late Periods, each of the kingdoms and many smaller states was
constantly at war with its neighbors. Civil developments are evi-
denced by the construction of planned cities, especially on the North
Coast of Peru, where large urban centers were laid out with protective
walls, gardens, reservoirs, ceremonial mounds, cemetaries, streets, and
blocks of houses.
Attainments in political organization and conquest, however, were
probably not the same throughout the Central Andes. The Tiahua-
naco culture, which widely affected local areas throughout Peru, may
have represented a military conquest comparable to that of the Inca,
which had the same cultural effect. Even the classical site of Tiahua-
naco, however, seems to have been a predominantly religious center,
and the Tiahuanaco influence may have been more religious than mili-
tary. Subsequent to the Middle Periods, a considerable local auton-
omy seems to have returned. In the Highlands, the population
remained dispersed, being only loosely nucleated around religious
and administrative centers, and political control did not extend beyond
the ayllu or group of ayllus which formed the independent state.
Each Coastal valley, too, probably retained considerable independence.
The North Coast, however, had a long tradition of centralized plan-
ning. This planning may have been only within the framework of
inclejêndent city states, which began at sites like Chanchan, in the
Middle Periods, but large irrigation systems evidence the extension of
centralized control over considerable areas. By and large, it was in
advance of the Highlands. The Inca, relative newcomers to empire
building, were the first people who are known to have incorporated
all of the Central Andes in a single state. The Inca (p. 734 ff.)
developed government at the expense of other features of the
earlier patterns. A somewhat fluid class structure, in which status
could be attained by the capture of slaves and concubines, was replaced
by a hereditary caste system, which was based on wealth-jDroducing
commoners and in which the slave class became unimportant. Even
the priests came to form a stratified, hereditary class, and so did the
very gods. All religious, economic, social, and military activities of
state concern were fitted into a system of governmental control, though
otherwise the folk culture was left intact (p. 737).
SPREAD OF THE ANDEAN PATTERNS

Outside the Central Andes, South American archeology has not yet
disclosed with certainty the cultures that were contemporary *with the
Chavin and Early Periods of Peru. Ethnographic data and frag-
;
mentary archeological evidence, however, show that most of the basic

technologies, many items of material culture, and even the sociopoliti-
cal patterns of the Central Andes attained a wide distribution.
—
The Marginal areas. Eastern Brazil, the Pampas, Patagonia, and
the Chilean Archipelago felt little influence from the Andes. None
of these areas contains archeological evidence of anything more ad-
vanced than the Marginal tribes which occupied them at the Conquest,
except those portions of eastern Brazil where a Tropical Forest type of
culture attained limited distribution.
The Southern Andes. —The Southern Andes, to judge by the
archeological sequences on the coast of North Chile, had a Marginal
type of technology prior to the introduction of agriculture. The
Early Period traits include: spear throwers, bolas, harpoons, fish-
hooks, and earth burial. With the introduction of agricul-
fire drills,
ture came pottery, especially painted wares, coiled basketry, slings,

wool and cotton weaving, sandals, dogs, and balsas (Handbook, 2 fig. :
49) . The southward from the Central Andes was intensified

diffusion
toward the end of pre-Conquest times by the development of overland
trade, which was carried on by means of pack llamas traveling on
roads. Under the Inca Empire, the Southern Andes received metal-
lurgy. Central Andean type garments, and other traits, many of which
reached the Araucaniaiis.
The Atamameno and Diaguita had a rich Andean-type material cul-
ture, some of it probably acquired in Tiahuanaco times, but they re-
tained a primitive sociopolitical pattern, because the forbidding deserts,
even though exploited by Central Andean techniques, could not support
sufficient populations to maintain a class system. The communities
remained comparatively small and seem to have consisted mainly of
lineages or kin groups. The Araucanians occupied a much more fertile
area and had a denser population, but itwas divided into a greater
number of units rather than into larger, more close-knit ones. These
units were of the Tropical Forest type, based on the lineage or kin
social classes had scarcely emerged and intervillage federation hardly
begun.
The Araucanian culture has resemblances to areas other than the
Andes. Frame, thatched houses, palisades, dugout canoes, urn burial,
and cannibalism link it with the Tropical Forests. The association of
transvestitism and the tamborine with shamanism are characteristic
of Siberia and the northwest coast of America. The distribution of
this peculiar shamanistic complex is difficult to explain possibly trans-
;
vestitism in Colombia, which had some ritual significance, is another

survival of an early widespread complex.
The Northern Andes and Circum-Caribbean —
area. At the Con-
quest, Ecuador, Colombia, and the area around the Caribbean Sea had
a pattern which contained all the essential features of the Chavin and
Early Periods of the Central Andes, though these features were de-
veloped in differing degrees. An ample subsistence complex sup-
ported a very dense population, though crops adapted to the rain
forests took the place of certain Andean species, and, along the coast,
sea resources were added to farming. Society was stratified into
chiefs or nobles, commoners, and slaves or war captives. There was
a temple cult dedicated to tribal gods and served by priests or by
shaman-priests. Warfare was strongly developed, and it afforded
*
individuals a means of attaining rank. Technological processes in-

cluded ceramics, woven basketry, and metallurgy in gold and tum-
baga. In the Central Andes, metallurgy was the only technology to
advance radically after the Early Periods processes of the latter pe-
;
riods, such as copper smelting and bronze making, are not found
among the Circum-Caribbean peoples.
The similarity in general patterns as well as in specific elements of
theAndean Early Periods and the Circum-Caribbean culture of the
Conquest indicates some connection between the two. This is not to
say that the latter was derived entirely from the former. In a com-
parative analysis of the Circum-Caribbean culture in volume 4 (pp.
6-11) it was shown that the general patterns and traits of this culture
,
extended from the Andes to Mexico and Yucatan. It was postulated

that an early, inter- American Formative Period culture underlay the
more special later developments of all these areas and that this cul-
ture had the following features Subsistence based on fairly intensive
:
farming, especially of maize; dispersed settlements that tended to

nucleate around religious centers ;
political units limited to the village
or small state; a class tendency, status being accorded warriors
and religious leaders, who probably exercised considerable civil power
but whose rank was achieved rather than inherited religious mounds, ;
altars, idols, offeratories, and shrines; priests, probably becoming

distinct from shamans considerable warfare captives taken as slaves
; ;
and as sacrificial victims; children of slaves, however, probably free;

human sacrifice; cannibalism, probably ritual; human trophies, espe-
cially skulls and flayed skins ritual blood letting celestial and animal
; ;
deities, and incipient agricultural ritual; construction of cause-

ways, aqueducts, canals, defensive works, and stone buildings; bark
cloth loom weaving of domesticated cotton and use of batik and tie-
;
dyeing painting, negative painting, incised, and plastic treatment of

;
pottery; featherwork and feather mosaics; metates; armor; coiled

and woven ( ? ) baskets cloth garments sandals and many other cul-
; ; ;
ture elements of a very wide inter-American distribution. This

Formative Period culture contained the essential features of the

Circum-Caribbean culture.^^
In many respects, however, the Circum-Caribbean culture is more
closely linked with the Andes than with Mexico or Yucatan. Central
America and the Andes share such archeological traits as stone-cist
and deep-shaft graves, stone stools, jaguar-form stools, Manabi-type
stone slabs, and mace-head clubs, and such ethnographic features as
the platform bed, burial of subjects with a chief, mummification, and
tumbaga (the last north to Panama) Colombia was even more closely
.
linked with the Central Andes (p. 731) though it also shared very
fundamental features with Central America. Thus, the Chibchan lan-
guage extended northward to the Ulua-Stmio-M osquito in Nicaragua,
and both Colombia and Central America seem to have had a basically
matrilineal society, in contrast to both Mexico and the Central Andes.
There is also a number of traits which link Central America and the
Tropical Forest tribes: manioc, Muscovy duck, babracot, pole-and-
thatch house, palisaded villages, communal house, hammock, blowgun,
and other elements especially adapted to tropical rain forests (p. 698).
A significant number of these Andean and Tropical Forest traits
also reached the Antilles.^
Mexican influence is not wanting in Central America, but it appears
to be relatively late, it is definitely concentrated among the Meso-
American tribes who were scattered along the Pacific Coast, and it did
not reach the Antilles (Handbook, 4: 199).
In the northern Andes and in Central America, it seems certain that
a Formative Period culture preceded the Circum-Caribbean culture.
The areal differentiation and interareal linkage of this postulated
early culture are still to be determined. Possibly all of it came from
South America present data, however, cannot support such a conten-
;
tion. The data of both archeology and ethnology, on the other hand,
show that apart from the general inter-American ,or Formative Period
features. Central America is related to South America rather than to
Mexico or Yucatan; i. e., the flow of more specific features has been
predominantly northward for a long time. For precisely how long
this has been soremains to be determined. It is suggestive that litters
were known in the Andean Early Periods and that deep-shaft graves
in Colombia may have comparable antiquity.
There is reason to suspect that at some early period or periods,
Andean culture may have spread with considerable vigor into tropical
rain forest areas, where it subsequently diminished or disappeared.
8"For modification of the following analysis, see Steward, 1947, 1949.

' interesting features of the Circum-Caribbean culture were found among certain
Some
North American tribes, especially the Natchez, bordering the Gulf States. No historical
connection of these cultures is postulated, but a comparison of them would be interesting.
738931—49 50
This expansion was, perhaps, comparable to that of the Pueblo II

Period in the Southwestern United States. A
similar, though more
restricted expansion, occurred during the Tiahuanaco and Inca
Periods. Evidence of this early spread consists of such archeological
remains as stone construction, irrigation works, terraces, roads and
causeways, which are found in various places in eastern Bolivia, in
the llanos of eastern Colombia, in the Venezuelan Andes, and in the
Highlands and North Coast of Colombia. Though some of these
remains may eventually be identified with the historic tribes, many
of them seem to have fallen into disuse long before the Spanish Con-
quest. Central America, too, may well be found to have archeological
remains evidencing a more complex culture than that found at the
Conquest.
A highly speculative explanation of these remains is that an An-
dean-type culture expanded at an early period into sparsely popu-
lated areas well beyond the Andes. These far-flung remains, how-
ever, do not have a uniform style, comparable to that of the Pueblo
II, Tiahuanaco, or Inca cultures, and the expansion must have been
gradual. In the course of time, population pressure and warfare
perhaps led to the development of tightly nucleated, palisaded vil-
lages in place of dispersed settlements, and to the abandonment of
hilltop roadways, stone construction, and the building of
forts,
mounds, except for burial purposes. At the Conquest, the only great
civilization outside Highland areas were those of the Maya^ and these
were in the dispersed settlement pattern. This unique Maya adapta-
tion and survival presents an interesting problem.
—
The Tropical Forests. A class-structured society was not char-
acteristic of the Tropical Forest tribes, and, though the large villages
found in many regions of exceptionally dense population would seem
to have afforded a basis for it, social cleavage was predominantly in
kinship groups. In many tribes, however, rank was accorded to war-
riors, who formed something of a class, and captives were kept as
slaves. Among some tribes which acquired the horse after the Con-
quest, classes definitely emerged to form a pattern like that of the
Sub-Andean tribes.
It is difficult to know whether the status accorded warriors was the
result of purely local development or of diffusion, because stimulus-
diffusion from the Andean and Sub-Andean cultures may have been
far-reaching. An opportunity for such diffusion was provided by the
proximity of some of the class-structured tribes to the Andes. The
definite class-tendency of the Atacameilo, Diaguita, and Araucanians
can certainly be traced to the Andes. Even the GomecMngon and
Huarpe of the neighboring Pampas, despite their Marginal ecology,
showed slight influence of this tendency. The Ahipon, Mocovi, Pay a-
gud^ and Mhayd of the eastern Chaco, though definitely Tropical Forest
in their technologies and material culture and separated from the
Andes by the Marginal peoples of the western Cliaco, may have been
affected by the Andes at some time in the past. It is interesting that the
Mhayd were one of the few non- Andean tribes to carry on wars of con-
quest. The Tupi and Carib, though extraordinarily warlike, were
geographically remote from the Andes and the Caribbean area. They
lacked true warrior classes, and captives were bracketed into kin groups
rather than into a slave class. The Arawak^ especially of the Guianas,
possessed incipient social strata in that sons-in-law became a patri-
arch's retainers, though here, too, society was still on a kinship basis.
It may be significant that some Guiana tribes, especially the ArawaJc^
were matrilineal, like the Circum-Caribbean peoples, whereas most
Tropical Forest tribes were strongly patrilineal.
The Tropical Forest tribes lacked a temple-idol cult. A few of
them had public harvest ceremonies, and among many, the shaman
performed oracular functions in the men's club. As a rule, however,
the shaman apparently conferred with his own spirit-helper rather
than with a tribal god. The gods, especially celestial ones, which
became objects of tribal worship among the Circum-Caribbean and
Andean peoples, were usually little more than mythological characters
in the Tropical Forests. If there is a connection between Tropical
Forest and Circum-Caribbean religion, it is probable that the latter,
under Andean influence, built on a Tropical Forest pattern, assigning
the shaman priestly functions, especially as an oracle for the tribal
gods, which were represented by idols and kept in a special temple.
The Tropical Forest men's house was the functional equivalent of the
temple in being the scene of various rites, but it also had purely social
purposes.
In technological and material culture, the Tropical Forest tribes
had all Circum-Caribbean features except metallurgy they used loom
;
weaving, domesticated cotton, ceramics, and woven basketry. The

resemblance of these material traits to those of the Circum-Caribbean
area, however, is greatest in the Guianas and Amazon. In eastern
Brazil and around the periphery of the Amazon, many of them
disappear.
In addition to the technologies, which were perhaps ultimately
Andean-derived, the Troi)ical Forests had many traits which they
shared with the Circum-Caribbean peoples but hardly at all with
the Andes. Many, but not all, of these are adapted to tropical rain
forests: the tropical root crops; thatched, frame houses; scant gar-
ments; hammocks; dugout canoes; bark cloth; blowguns; arrow
poisons; fish poisons; rubber-ball games; hollow-log drums; and use
of tobacco.
The present thesis postulates that the Tropical Forest culture de-
rived its essential technology from the Circum-Caribbean culture and
that it also acquired certain rain forest traits, but it failed to borrow
the Circum-Caribbean sociopolitical and religious patterns. In be-
coming adapted to fluvial, littoral, and rain forest areas, the tech-
nological complex spread via the main waterways. Specifically, it
seems to have spread from Venezuela down the Atlantic Coast and up
the Amazon and its tributaries (perhaps secondarily it spread up the
Orinoco and down the Rio Negro), suffering successive losses as it
reached the headwaters, where many of the tribes remained Marginal or
Semi-Marginal. It is clearly evident, however, in the Peruvian Mon-
tana (Handbook, 3 535 5 697) and it may have broken through the
: ; : ,
watershed of the Amazon in eastern Bolivia to reach the upper Para-

guay River, while another stream flowed down the coast of Brazil to
the Parana Delta. It followed the rain forest and major waterways
but scarcely penetrated the plains and savannas, where the tribes
remained Marginal.
Certain major linguistic groups with a wide distribution must have
played an important role in the diffusion of the Tropical Forest
culture. In the north, the Arawak^ some of them with a Circum-
Caribbean culture, and the Garib are outstanding. South of the
Amazon, the Tupi were among the major culture carriers. On the
other hand, the many small, linguistically isolated Marginal and Semi-
Marginal groups clustered around the periphery of the Amazon Basin
(map 18) are probably to be regarded as predecessors of the far-flung
language groups. Pushed to the back country, they have long re-
mained in isolation.
Some authors have ascribed to the Arawah the most important role
in developing and spreading South American cultures. To these peo-
ple has been attributed the diffusion of vertical looms, the cult complex
with secret trumpets, elaborate ceramics, and other traits throughout
the Tropical Forests. They have been accredited with the spread of
Circum-Caribbean traits to the Antilles and to Central America and
even the introduction of the early Chavin, Tiahuanaco, and other
cultures to the Andes. In view of the very great differences in the
—
cultures of local Arawahan tribes for example, the Circum-Caribbean
Island Arawak of the Antilles, the Sub- Andean Arawak of eastern
Bolivia, the Semi-Marginal Campa^ Piro^ and others of eastern Peru,
—
and the Tropical Forest Arcuwakans of the Guianas it is necessary
first to establish that there was such a thing as an Arawakan culture.
In terms of their sociopolitical patterns, it is obvious that these tribes
would have to be classed with their neighbors on a culture- area basis
rather than with one another on a linguistic basis. The more im-
portant traits and technologies, such as ceramics, loom weaving, and
basketry, occurred among the Quechuu and Aymara of the Andes,

the Chibcha^ Carib^ and other tribes of the Circum-Caribbean area,
and the Carib, Panomis^ and dozens of smaller linguistic groups in the
Tropical Forests they were by no means peculiar to the Arawakans.
;
The Arawakans certainly played an important role in diffusing these

features in parts of the Tropical Forest areas, but even here their role
has yet to be clarified. One has the impression that the importance
attached to Arawakan migrations resulted in part, at least, from a
certain methodological approach to the problem of culture classifica-
tion when cultures are described by linguistic groups rather than by
:
areas, there is a concomitant tendency toassume a culture type for each

language and then to reconstruct the history in terms of the migrations
of language groups.
POST-COLUMBIAN CULTURE HISTORY
Comparatively little attention has been accorded the acculturation

and assimilation of the Indian to European culture since the Spanish
Conquest. The literature is probably adequate to sketch the broad
outlines, but it has been utilized mainly to reconstruct aboriginal
cultures and the data of the 400-year post-Contact Period have usually
been compressed into a two-dimensional picture without historical
perspective. What are virtually pioneering essays in identifying
and describing some of the post-Columbian acculturational periods
are, however, included in several Handbook articles. (See especially
Kubler's on the Andes (2: 331), Murra's on Ecuador (2: 785), Stew-
ard's on the Montana (3: 507), and Tschopik's on the Aymara
(2:501).)
The Europeans brought to South America a comparatively homo-
geneous culture. Though their initial penetration was variously made
by the conquistador, missionary, or colonist, the prevailing purpose
was to exploit the native peoples through a system of tribute, which
soon amounted to mass labor. The success of this system and its effect
on the Indian depended upon various local factors the native popu-
:
lation density the incidence of disease on the natives the accessibility

; ;
of the area and its suitability for Iberian types of land use and the ;
predominating facet of European culture, such as proselytization,

mining, or conquest.
—
The Marginal and Tropical Forest areas. Along the coast and
navigable interior waterways, the Indian was, as we have seen (p.
664 f.) quickly exterminated or absorbed. In the less accessible areas,
,
he survived and retained a native culture, but he was strongly affected

by the Whites even before direct contact with them. Throughout the
Tropical Forest. and Marginal areas, the course of pre- and post-
Contact acculturation was similar in certain general respects.
Prior to the arrival of the Europeans in these areas, the Marginal

and Tropical Forest Indians received certain vital trade goods which
brought a temporary efflorescence of and even deep changes in their
culture. New crops, domesticated animals, and steel tools, and, later,
firearms and other items so improved exploitative activities and trans-
portational facilities that villages of unprecedented size developed,
tribal contacts were increased, and even aboriginal culture elements,
such as the blowgun and bow, were diffused to many tribes which had
previously lacked them. Communities became larger, requiring new
political controls and involving new kinds of interpersonal contacts,
and warfare was intensified by the pressures of dislocated tribes.
These changes were within the frame of the preexisting, native
culture, but certain striking new sociopolitical patterns emerged
from the new ecological patterns. In the case of the Pampean and
Patagonian tribes, the effect was very similar to that among the
Plains Indians of North America. The former were hunters of
guanacos and rheas in grass and brush country the latter, of bison
;
on the prairies. In pre-horse times, both had lived in small bands.

The Patagonian bands were probably patrilineal, like those of the
Ona; some of the Plains tribes, however, had been farmers. In both
areas, the horse and a limited amount of gear were introduced by the
Spaniards, even before there was close contact between Indian. and
White. In both, the advantage of the horse as an adjunct of hunting
and as a pack animal permitted the small aboriginal groups to amal-
gamate into bands, which, though unstable, were far larger than
anything possible among foot Indians. Some of the North American
farming tribes of the Plains gave up their fields to become horse
nomads. The territory claimed by each of these bands became larger,
though less well defined than previously, and the chief gained in
power and prestige. The tribes bordering these areas were dislocated,
causing intertribal strife and intensifying warfare. Cultural features
less closely connected with the ecology were different in the Plains
and the Pampas for example, dress, house type, and ritual elements,
;
but even some secondary traits, curiously, appeared in both. For

example, people artificially lengthened their tresses with horse hair.
(Regarding the influence of the horse on the Chaco, see volume 1 of
the Handbook, pp. 202-203, 265-267, 304-308. For the Goajiro, see
Volume 4, pp. 369-383.)
In some areas, the Indians were first contacted by missionaries.
Though the orders adopted different policies, all of them settled the
Indians in large villages and encouraged farming and various new
crafts. This served to introduce considerable European culture, in-
cluding Christianity, but disease rapidly reduced the populations in
the mission villages. In many cases, the natives rebelled and returned
to aboriginal modes of life. Those who remained continuously in

the missions have become very nearly extinct.
Where the European colonists established settlements, the Indians
became attached, voluntarily or through enslavement, to them. Con-
sumption of European food and goods, use of European dress, and
completely new economic and social habits quickly eradicated native
culture. The Indians who survived the frightful toll of epidemic
diseases were gradually absorbed into the White and Negro popula-
tions.
Along the coasts in the tropical areas and in the Antilles, the
population was comparatively great, and it might be thought that it
could have served the Spaniards for mass labor, as in the Andes.
Instead the Indians disappeared amazingly soon and their place was
taken by Negro slaves. One factor in this decline was the devastating
epidemics of European diseases, especially in the large settlements
established by the Spaniards. Another was probably the violence of
the European Contact on peoples far less prepared for virtual servi-
tude than the Central Andean tribes who had been under the Iwâ
Empire.
Each year, the surviving Marginaland Tropical Forest peoples
decline further in numbers and become more assimilated to national
culture.
—
The Sub-Andean and Circum-Caribbean tribes. These peoples
experienced the same general post-Contact trends as those of the
Tropical Forests, except for a difference brought about by the distinc-
tive features of their aboriginal culture. All the Indians of the An-
tillesand most of those elsewhere on the coasts and lowlands became
extinct or were absorbed. In a few inaccessible regions, especially in
the mountains, Indian cultures still survive, but they have experienced
a phase of deculturation since the Conquest.
The older chronicles describe the class system, temple-idol cult,
organized warfare, and esthetic craft production of Circum-Caribbean
types. Modern ethnographies disclose only an unstratified society, a
shamanistic religion, little or no warfare, and crude crafts. (See, for
example, the Cimu^ vol. 4, p. 257.) This step down to a lower level,
very similar to that of the Tropical Forest tribes (4:2), was the direct
result of the Conquest, which destroyed the higher levels of organiza-
tion rather than the folk culture. The military basis for a class struc-
ture was removed when warfare, cannibalism, human sacrifice, trophy
taking, and slavery were stopped, and the hereditary basis was lost
when the chiefs and nobles were required to give over power to Euro-
peans. In addition, the tribes were forced into submarginal lands
which could not maintain the earlier sociopolitical patterns, and,
during the Conquest, warfare and disease greatly reduced the popula-
tion. Organized native cult religion yielded to the Catholic Church,

and only the more informal and private shamanistic and magical prac-
tices and folk superstitions survived. Technology and material cul-
ture deteriorated, partly because their finest expressions had been dedi-
cated to the nobility, which was now gone, and partly because the new
economy sank to lower levels of efficiency.
Stripped of their Circum-Caribbean and Sub-Andean features, these
tribes were degraded to a generalized Tropical Forest level, and their
patterns came to resemble those of the Amazon considerably more than
those of their own This phase of deculturation has consid-
ancestors.
erable theoretical interest for it suggests that the Circum-Caribbean
cultures were in reality Tropical Forest type cultures with an Andean

overlay. This thesis, however, is antithetical to that suggested previ-
ously that the Circum-Caribbean cultures might have represented a
:
somewhat deculturated form of an earlier, fairly widespread Andean

or Formative Period culture, which, by successive losses of social,
religious, and esthetic features during both prehistoric and historic
periods gradually converged toward the Tropical Forest pattern.
These observations are made to point up a problem rather than to
offer hypotheses. To ascertain the prehistoric levels as well as the
directions of cultural flow, archeology is badly needed, especially in
Colombia, Ecuador, and Central America. To ascertain the precise

nature of the native cultures at the time of the Conquest, the early
chroniclers must be exploited to the fullest, for the ethnographies of
the few tribes surviving today give a very incorrect picture of
aboriginal cultures.
—
The Central Andes. The Central Andes have differed from the
remainder of South America in surviving the shock of Conquest and
the attrition of European Contact without serious loss of population
or culture. Epidemic diseases were not so devastating during the
Conquest and Colonial Periods, and, though in the beginning the pop-
ulation may have been halved, the Indians have always outnumbered
the Whites. The sheer numbers not only served to retard assimilation
they invalidated many European policies used elsewhere in South
America. The minority of Europeans found it easier to exploit the
vast labor supply through adapting their methods to preexisting native
sociopolitical patterns than to completely uproot, recast, and finally
replace the people. In the Inca Empire, they found a pattern more
or less adaptable to their ends. They blanketed the higher officials
into their class system and required the already subservient and much-
exploited masses to supply tribute, taxes, labor (mita), and servants
(yanaconas).
The effect of the Conquest on the Central Andes as a whole was
similar to that on the Grcum-Caribbean area. The more organized
and distinctive forms of tlie Central Andean pattern were replaced

by European forms while the folk culture remained more intact.
The Inca governmental system was replaced by the Spanish, with
its fountainhead at first in Spain. The native social castes were lev-
eled, except as the curacas found a place in the European scheme. The
priest-temple-idol cult and the rites pertaining to it were formally
replaced by Christianity, but shamanism, sorcery, and the more covert
beliefs and practices organized around local shrines and deities, es-
pecially the mountain spirits, were able to survive. Craft production
became more or less restricted to the plain, utility goods which had
always been made for home consumption. The nobility and the Inca
church, which had received most of the art products, were now gone.
The Spaniards were more interested in tribute, and they mined their
own gold, which they converted into money rather than into orna-
ments. Many European elements were adopted, for example, wheat,
barley, cattle, horses, mechanical looms, methods of house construc-
tion, and types of dress, but these merely augmented an already rich
native material culture. At first, the new produce enabled the Indian
to pay his tribute but was not adopted into his culture.
There have been two main acculturational trends: (1) individuals
and classes of individuals have been detached from the context of
native life and assimilated quite rapidly into the national, Hispanic-
American culture; (2) comparatively stable native communities have
experienced more gradual acculturation.
Since the Conquest, large numbers of individuals have left their
native communities to become a somewhat mobile, national prole-
tariat that is integrated with the Europeans in a condition of economic
and social interdependence. Their dislodgement from the native
communities has several causes; work as servants to the Whites (the
yanaconate) mine labor, service in the army, the rewards of various
,
new types of production in the urban centers, and the loss of their
own lands. The last, and perhaps the most important, factor was
brought about by the early wars new types of land use, especially
: ;
cattle raising, which crowded out part of the dense farming popula-
tion ; complication of legal titles to land increase of the native popu-
;
lation; and the introduction of individualized land ownership and

cash crops, which frequently led to land sales. Some of the detached
persons became farm workers (colonos) on the haciendas and either
were rapidly Hispanicized through the efforts of the landlords, or
simply lost their Indian features without commensurate gain. In-
creasing numbers, however, have become urbanized and, though retain-
ing some Indian culture, they wear European dress, work for wages,
can read and write, and are geared to the national socioeconomic
system through a national consciousness and a sense of solidarity.
The main reservoir of Indian life in the modern Andes is the com-
munity (comunidad or ayllu), which has remained relatively in-
tact, except that during the Colonial Period its size was increased
through the system of reductions (reducciones) to facilitate adminis-
trative controland religious proselytization of the Indian. In Peru
and Bolivia, there are 3,000 such communities today, each with about
500 persons. Their patterns, despite a considerable content of Span-
ish elements, retain many of the principal characteristics of the native
folk culture (Handbook, 2: 441) production mainly for local use,
:
though always some surplus crops for sale; native methods of culti-
vation; exchange of labor with community members (aine) Indian ;
garments, houses, and other material items; Quechua or Aymara

speech, though much bilingualism inability to read; local control of
;
matters not of national concern ; the extended family as the principal

sociological unit; community solidarity expressed through village
endogamy, local ceremonies, shrine (huaca) worship of local deities,
especially those of mountains and practice of shamanism and magic.
;
Goods are exchanged at periodic markets, which may represent the

resurgence of a pattern that was somewhat suppressed under the Inca
control of production and distribution.
As the national economy expands, as education is extended, and as
transportation brings the remoter rural areas into contact with the
more advanced centers, the Indian communities are being gradually
transformed. Almost all land is now individually owned, facilitat-
ing its loss through sale. The sale of land together with limited liv-
ing space is breaking down the patrilineal, extended family, because
new generations tend to bud
and become independent. This in
off
turn weakens the community. Cash crops are introducing a money
economy and greater use of manufactured goods. Special industries
also augment community income and promote urbanization. Many
non-Indian communities are incorporated and have producers' and
consumers' cooperatives. Bilingualism is increasing. Formal edu-
cation is introducing the national culture and a national conscious-
ness.
Year by year, individuals who have been dislodged from their
communities and even whole communities gradually pass from the
category of Indian to that of national Ecuadorian, Peruvian, and
Bolivian. The process of mestizaje, or assimilation to the national
culture, seems likely to accelerate.
HISTORICAL SUMMARY
The Indians reached America from Siberia some 15,000 to

first
20,000 years ago and eventually spread southward into South

America, arriving in Tierra del Fuego not less than 3,000 to 5,000
years ago. Their original culture included fire making; flint chip-
ping; spears; scrapers; skin dressing; birth, puberty, and death
observances; shamanism; and perhaps various ritual elements. In
the course of time, bows, harpoons, nets, traps, and other material
items and various ritual and social elements were acquired in different
areas. At no time, however, did the early hunters and gatherers have
a uniform sociopolitical pattern. The extremely great local differ-
ences in environments, natural resources, and subsistence techniques
imposed upon the population of each region the necessity of grouping
itselfand behaving in very different ways. Some were dispersed in
conjugal family groups; others were nucleated in extended families,
which were patrilineal or matrilineal, sedentary or nomadic, and
permanent or temporary according to the socioeconomic activities
required in each area, A few, where local abundance of food per-
mitted unusually large population aggregates, were grouped into
loose bands consisting of several conjugal families or of extended
matrilineal or patrilineal families. The Marginal tribes encountered
at the Spanish Conquest probably give an idea of the general variety
of early ecological adaptations and sociopolitical types, but it is
certain that all of these have changed over the years as new weapons
and transportational facilities introduced new subsistence patterns
and as the peoples migrated or were pushed into new environments.
The domestication of a considerable number of native American
plants was begun several thousand years ago, certainly before trans-
Pacific voyages could have taken place. By 500 A. D., and probably
1,000 years earlier, more than 30 of these were grown in the Central
Andes, and some of them had reached the limits of their genetic
variability. They became the basis of a large, stable population in
the Andes, which, in its earliest known manifestations, the Chavin
Periods,"^ had already attained mature esthetic, social, and political
patterns, and, by the Early Periods possessed all essential technolo-
gies and refinements of material culture. It had a class-structured
society, organized warfare and human-trophy taking, a priest-temple-
idol cult, and excellent loom weaving, ceramics, metallurgy, basketry,
architecture, and transportation.
It is postulated that a culture, called the Formative Period culture
and possessing the general features of the Central Andean Early
Periods, extended northward to Mexico. The region of its ultimate
origin is not known, but it might have been in South America. The
archeological and ethnographic cultures of the Circum-Caribbean area
represent a more specific formulation of the Formative Period cul-
ture, but the specific features link them most closely with the Andes
and indicate a predominant northward flow of culture. It is pos-
'" See footnote 6, p. 753.
The general cultural tradition represented by the Circum-Caribbean

tribes spread southward from Venezuela, probably down the Coast of
the Guianas to the Amazon, and up the main tributaries of the Amazon
(map 22 compare with maps 2 and 3)
; The sociopolitical and reli-
.
gious patterns and metallurgy, however, were largely lost, and only
ceramics, weaving, and basketry survived. Even some of the tech-
nological traits dropped out toward the headwaters of the Amazon.
The tribes of the Northwest Amazon and some of those in the Mon-
tana and the southern part of the Amazon Basin remained Semi-Mar-
ginal. In the more remote localities on the Amazon headwaters and
between the major rivers, beyond the effects of this cultural diffusion,
the tribes remained Marginal. The Marginal and Semi-Marginal peo-
ples form an almost continuous area that extends like a great U
around the Amazon Basin, from the Amazon-Orinoco watershed south
through the Montaiia and east across parts of eastern Bolivia and
Mato Grosso to the Highlands of eastern Brazil. The U is broken
mainly in the upper Madeira River region in eastern Bolivia, where
Tropical Forest tribes pushed southward, meeting other Tropical
Forest peoples who had followed the rain forests down the coast of
Brazil and inland to Paraguay and beyond.
Thus, it may
be postulated that a single great historical tradition
originating from the Andes in an early period carried sociopolitical
and religious patterns and a developed technology from the northern
Andes into Central America and almost entirely around the Carib-
bean Sea where it became somewhat adapted to sea coasts. It almost
completely failed to penetrate the Tropical Forests directly from the
Andes. Instead, the technological traits were carried by seacoast and
riparian peoples down the Guiana coast and up the Amazon, and the}^
are absent beyond the areas of easy navigation.
In the southern Andes, the technological, material, and ritual pat-
terns spread south to the Araucanians. Traces of Andean influence,
perhaps even of sociopolitical patterns, may have reached the eastern
Chaco, despite intervening Marginal tribes, and there blended with
Tropical Forest influence from the coast of Brazil.
In its main outlines and element content, this development and flow
of culture was indigenous to America. It is not impossible that a
number of isolated elements reached America from the Pacific. It
does not really matter, however, whether blowguns, lime chewing with
a narcotic, bark cloth, a few domesticated plants, and even such things
as frame, thatched houses were native to America. They were incor-
porated into various local patterns which they affected little if at all.
The basic patterns, in fact, seem to have been established, perhaps in
Peru, though quite possibly elsewhere, well before any trans-Pacific
voyages could have been made.
After the Spanish Conquest, the Marginal tribes rapidly disap-

peared, except in the more inaccessible areas. Lacking a vigorous cul-
ture before the Conquest and vulnerable to European contacts after-
ward, the tribes of the Brazilian coast, of southern Argentina, of the
Chilean Archipelago are virtually extinct, and they have left little
cultural or racial imprint on the modern population. The greatest
survival of Indians with a native, but not precisely pre-Columbian,
culture is around the periphery of the Amazon Basin. It is solely a
question of time, however, before these, like the other Marginal peoples,
lose their Indian characteristics.
The Tropical Forest tribes disappeared as cultural entities along
the coasts and navigable rivers, partly because they were absorbed or
overwhelmed by Europeans, partly because they were not adaptable
to slave labor under the European pattern, so that Negro slaves were
imported to replace them. They have, however, left a strong racial
strain and even many culture elements in the modern population.
The Circum-Caribbean tribes have become virtually extinct in the
Antilles, but many survive in the portions of Venezuela, Colombia,
and Central America which are unfavorable for White settlement.
The loss of the class system, the war complex, the religious cult, and
the technological refinements has deculturated them to a Tropical
Forest level.
GLOSSARY
CERAMICS
The following list was compiled by Dr. Gordon R. Willey and Dr.
Irving Rouse. The Spanish equivalents, where such are in common
usage, were supplied by Dr. George Howard. The list includes only
the more general terms for construction, forms, and decoration. The
dozens of terms applying to wares of only local interest are not
included.
Adorno. See Lug.
Annular base (base anular). A raised, approximately cylindrical rim on which
the vessel rests. When of extreme height, it is called a pedestal base.
Applique. See Filleting.
Basket construction. Building up the clay vessel vîthin a vpoven basketry
container.
Beaker. A vessel vphich has straight, parallel sides and a height that exceeds
the diameter.
Bowl (bol, escudilla, puco). A vessel which is wider than tall and which has a
large orifice.
Brushing or combing. Decoration applied before firing by combing or brushing
the wet surface of the vessel with a comblike or brushlike instrument.
Champleve. Decoration applied before or after firing by cutting the vessel sur-
face and carving out portions of it.
Coiling (procedimiento de construccion del rodete). Forming the clay ves-
sel by spiraling wet strands of clay to build up the shape desired. Sometimes
other annular methods were used instead of continuous coiling.
Coil obliteration (alisamiento de rodete). Obliteration of the lines of coils
by smoothing, scraping, or annealing them together before the vessel is fired
in order to obtain a smooth surface.
Collar (cuello). A convex section of the vessel wall surmounting a shoulder
or neck.
Combing. See Brushing.
Corrugations. Lines of coiling which have been intentionally left unobliterated.
Direct molded or shaped (modelado). A clay vessel formed by manipulating
the clay mass or lump with the hands coils are not used.
;
Dish or plate (plato). A shallow, open bowl.

Double-vessel (vaso doble). Usually, a vessel having two compartments which
have an internal connection.
EflSgy adornos. Life or idea forms affixed to the vessel.
EflSgy vessels. Vessels which have been shaped to represent some life or idea
form.
Engraving. Decoration after drying or after firing by scoring the vessel surface.
Fabric-, basket-, or net-impressed (impresiones de textiles, de canasterias, de
redes). Decoration made before firing by pressing a fabric, basket, or net
on the wet surface of the clay.
Filleting or applique. S'oft strands or strips of clay affixed to the vessel surface
before firing.
773
Fingernail impression and finger-impression (presiones digito-unguiculares).

Decoration before firing by impressing the wet surface of the vessel with the
fingernails or fingertips.
Firing (cocimiento). The baking process which makes the pottery object hard
and durable.
Fugitive painting. Painting which is applied after the vessel has been fired and
which is not fixed by refiring. Such painting tends to rub off.
Grit temper. Sand or crushed rock used as a temper or aplastic.
Handle (asa). A loop-shaped strip of clay attached by each end to the vessel
wall.
[ncision (incision, decoracion incisa, linea grabada). Decoration applied before
drying by scoring the wet surface of the clay.
Inclusions. Particles in the clay, added, whether accidentally or intentionally,
as a tempering material.
Jar. A vessel whose height is greater than its diameter and whose orifice is
constricted to less than the maximum diameter of the vessel.
Keel. The angle formed by the body and shoulder of a vessel.
Linear-punctation (surco inciso con escalonamiento interior punteado). Some-
times called punctation-incision or drag-and-jab punctation. Made by alter-
nately pulling and jabbing a sharp-pointed instrument over the soft, unfired
vessel surface.
Lug, adorno, or nubbin. A handle which is solid, x'ather than looped or pro-
jecting.
Modeling. Manipulation of the clay either with the hands or by means of pre-
pared molds to give the vessel or object decorative forms.
Monochrome painting (decoracion pintada monocroma). The use of a single
color for decorative purposes.
Neck. An outcurving section above the body or shoulder of a bowl or jar.
Negative design. A negative design achieved by a positive painting technique,
in contrast to the negative, lost-color, or resist-painting technique.
Negative or lost-color painting (decoracion pintada negativa). A lost-color,
resist-dye, or batik process, whereby the vessel surface is partially covered
with a resistant, such as wax, and immersed in the dye, paint, or slip.
Firing fixes the color but burns off the wax or resistant, leaving the areas
where it had been uncolored.
011a. Usually, a large, globular vessel, either of bowl or jar proportions, which
has a constricted orifice.
Oxidizing atmosphere (cocimiento en medio oxidante). Firing or baking the
vessel so that it receives an ample supply of oxygen and the clay turns
a reddish color. (See Reducing atmosphere.)
Paste (pasta). The matrix, or mixture of clay and water, from which the
pottery vessel or object is fashioned.
Pedestal base. See Annular base.
Plate. See Dish.
Polished black ware. (Sometimes called "bucchero" in Peru.) It is reduced
rather than oxidized in firing, the vessel surface being blackened by smoke
smudging and then polished.
Polishing (pulimiento). Smoothing and burnishing the dry, unfired, soft vessel
by rubbing with a small, smooth stone or a stick.
Polychrome painting (decoracion pintada policroma). The use of two or more
colors for forming designs.
Positive painting. Direct application of pigments to the surface of the vessel,
which is subsequently fired.
:
Vol.5] GLOSSARY 775

Punctation (decoracion punteada, punto grabado). Decoration applied before
firing by jabbing the wet surface of the clay.
Prepared molds (tecnica del moldeado). The use of negative molds or forms,
usually made of pottery, to impart the desired form or shape to a pottery
vessel or object. These molds were usually employed in pairs, front and
back.
Reducing atmosphere (cocimiento en medio reductor). A condition of firing
under which the vessel or object receives an insufficient supply of oxygen,
resulting in a greyish or blackish ware.
Relief-modeling. Life or idea forms modeled in relief on the vessel surface
before firing.
Rim (borde). The lip, or finished top edge, of the vessel wall.
Sherd (tiesto). A fragment of a clay vessel.
Slierd temper. The use of ground pottery fragments as temper.
Shoulder. The incurving upper section of the wall of a bowl or jar.
Slip (engobe). A wash or dip to which a pottery vessel or object is subjected
either before or after the first firing and before the final firing. The wash
is usually of a different color or shade from the natural color of the clay, and
it serves as a background for any painted designs that may be applied.

Simple or open bowl. A bowl which has its maximum diameter at the orifice.
Spout (gollete). An elongated, tubular orifice.
Subglobular bowl (vase subglobular). Bowl which is a flattened sphere. The
maximum diameter is greater than the orifice and is somewhere near the
mid-point of the height.
Temper (degraissant, Fr.) The aplastic, or hard particles, added to the paste to
increase or decrease its cohesive qualities.
Tripod base (base tripode). Three feet, nodes, or projections to support the
vessel.
Three-color negative painting. This refers to vessels decorated in three colors
the natural color of the clay the paint or slip, which brings out the design
;
in negative and the third or final color, which is applied positively.

; It is
really a combination of negative and positive painting.
Tripod base (base tripode). Three feet, nodes, or projections to support the
vessel.
Urn (urna). A large, deep vessel, generally jar-shaped, which is used for pur-
poses of burial.
BASKETRY AND WEAVING
Amazon loom.See Loom.
Andean spindle.See Spindle.
Applique. Related to embroidery techniques. Shaped pieces of material con-
trasting with the ground material, usually fastened to it with decorative
stitchery.
"Arawak" loom. See Loom.
Backstrap loom. See Loom.
Bacai'ri spindle. See Spindle.
Batik. A form of reserve technique involving a resist material such as wax,
which is applied over the surface to be protected against absorbing the dye.
When removed, the reserved areas are the original color of the fabric or the
color given it by a previous dyeing. Several colors in the same piece are
possible.
Batten. Synonym for Sword. A shaped length of hardwood, often wide and
heavy. As verb, batten means to press or beat down to the working edge the
weft element put through the space created in the warp plane.
—
738931 49 51
Bobbin. A used in weaving on small two-bar looms. Long, very slender

tool
Weft is wound around the bobbin.
stick or other object.
Bororo spindle. See Spindle.
Braiding. See Plaiting.
Braidwork. See Plaiting.
Brocade. Superstructural form of patterning. While the material is still on the
loom, the weaver introduces supplementary decorative yarns to build up
designs. Brocaded motifs often are confused with similar effects gained
through embroidery with needle. Single-face brocading is procedure by
means of which decorative yarns show in much greater amounts on the
surface than on the reverse side of the fabric.
Carding. A combining process preliminary to spinning, during which loose fibers
are laid more or less parallel to each other. Hand tools employed are called
cards.
Chain twists. Roth's term for Twining, q. v.
Checker weave. In some basketry and matwork the warp and weft elements are
equally wide. When these are systematically interlaced in over-one-under-
one order, result is similar in appearance to a checkerboard.
Coiling. Many varieties used in basketry. The simplest consists of single
foundation elements spiralling from bottom to mouth of the basket. En-
circling this element are vertical stitches often concealing the foundation.
Each stitch on the working row interlocks in some way with the foundation
element below it. Coiled basketry is also known as sewn in contradistinction
to woven. The Fuegian half-hitch coil is identified by a half-hitch taken
around the coil. Each successive spiral of the foundation element is lashed
to the one underneath it by a row of half-hitches inserted through the lags
of those on the preceding row with the aid of an awl. A si)ecial Fuegian
form is known as Wrapped or knotted half-hitch. Coiling (or half-hitch)
without foundation consists of a series of half -hitches or buttonhole stitches
those in the first row are taken over a foundation, those in each successive
row interlocking into the stitches on the row immediately above (figs,
15,16).
Cotton bow. A device similar in appearance to the ordinary bow. Loose fibers
in a lap of cotton or wool are fluffed by repeatedly snapping the bowstring
on them (fig. 26).
Crocheting. A single-element technique by means of which a looped fabric is
made from a thread of indefinite length with one or more hooks.
Damask weave. A single-face or double-face fabric with design developed by
lengths of warp or weft elements on the surface of the fabric.
In contrast
to brocade, no decorative yarns are introduced, but basic elements are so
manipulated as to form both the structure of the cloth and the patterning.
DistafiF. A wooden object in various shapes over or around which prepared fibers
are placed to be within convenient reach of the spinner (pi. 24). One
Quechua distaff is similar to an inverted horseshoe mounted on a short
handle.
Double-ply cord. See Single-ply.
Drop spindle. See Spindle.
European loom. See Loom.
Felting. Synonymous with fulling or shrinking. Moisture and pressure applied
to a mat of loose wool fibers condenses it to form a fabric.
Finger looping. A form of single-element technique in which an indefinite length
of yarn is so manipulated as to construct meshes without knots (fig. 39).
Vol. 5] GLOSSARY 777
Finger weaving. —
Also called "freehand" and "pick-in" weaving. The elements
are interlaced with no aids other than the fingers or small pointed sticks.
Float. A skip of the warp or weft element over two or more of the opposite
system of yarns. This leaves certain predetermined yarns free of binders
for a longer or shorter distance on the surface of the cloth in order to build
up the pattern motif.
Frame. A contrivance for holding the warp skein at tension. A frame, unlike
the true loom is not equipped with a heddle or other warp-lifting device.
Foundation element. See Warp element.
Fuegian half-hitch coil. See Coiling.
Ghiordes knot. See Pile weave.
Half-hitch. Same With one end of the cord
as buttonhole stitch in embroidery.
at the left, the otherend (or threaded needle) is made to encircle the object
(or edge of material in sewing) and come up through the loop thus formed.
Half-hitch coil. See Coiling.
Half-hitch without foundation. See Coiling.
Hand cards. Two flat rectangular pieces of wood that are skin covered and
studded with short lengths of steel or wire. They are presumably European
in origin. When the cards are pulled over each other, the loose fibers placed
between them are cleaned and straightened.
Heddle. On a two-bar loom, the slender rod or stick from which depend string
loops to encircle alternate warp elements. All heddle-controlled warps can
be raised together.
Hexagonal weave. In the simplest form, the weaving elements progress in three
directions : horizontally, obliquely upward to the right, and obliquely upward
to the left (fig. 14; pi. 20).
Ikat. A Malay word meaning to knot, to bind, to tie around; applied to the
weaving of yarns patterned by reserving the original or first-dyed color.
This is done by tying protective bindings around groups of yarns prior to ar-
ranging them on the loom bars.
Interlocking tapestry technique. See Tapestry.
Interwoven patterns. Accomplished by the weaver during construction of the
cloth, while still on the loom.
Kilim. See Tapestry.
Knitting. A single-element technique by means of which a looped fabric is made
from a strand of indefinite length with the aid of two, four, or more wooden
or metal needles (fig. 39).
Knotted half-hitch. See Coiling.
Knotless netting. A single-element technique. A variety of patterns are devel-
oped by means of making row on row of loops. The half -hitch, or coil with-
out foundation, is fundamental to most types (fig. 40).
Lattice weave. See Hexagonal weave.
Loom. A contrivance for holding the warp skein at tension. It is equipped with
a heddle of some type. The presence of a heddle distinguishes the true loom
from the Frame. The so-called Arawak or Amazon loom is a two-bar
loom supported by upright poles; warp elements are wound around the
bars in the form of a ring. The finished web may be a seamless band of
cloth or a rectangular piece twice as long as the distance between the bars
(figs. 27, 30; pi. 26). The Backstrap loom is a two-bar loom, the lower end
of which it attached to the weaver's waist, the upper end to a tree or other
support. This is also called a Peruvian belt, girdle-back, or stick loom (figs.
28, 31 pi. 25).
;
The European loom is a foot or treadle loom with from two
to four harnesses controlling as many divisions of the warp elements. The
harness in treadle-loom weaving serves the same purpose as the heddle on

a backstrap loom (pi. 25). The Rio Ucayali loom has the shape of a lyr^
oval, V etc. Two end bars are lashed in position across the opening. Fab-
rics from Ucayali looms may be woven, twined, plaited, or otherwise fab-
ricated (pi. 26).
Loom These are the proximal and distal end bars of a backstrap loom.
bars.
The warp skein is bound or slipped over them.
Loop plaiting. A term used by Roth (1924, figs. 19, 20) to distinguish the
technique used for making three-dimensional cords.
Matwork. Closely woven, flat, four-selvage rectangles in checker weave or
twilling of the over-one-under-three type.
Negative patterning. The design motif in the original color of the cloth on
a dyed or painted background of contrasting color, sometimes achieved by
scraping away the paint or other colored substance to form the pattern.
The methods differ. See Batik and Ikat.
Netting. A single-element technique in which a succession of loops are kept in
position by knots. Tools include a mesh gage and a device known as a net-
ting needle, which holds an indefinite length of cord.
Open weave. See Hexagonal weave.
Pattern stick. Term used to designate small round stick (or several) inserted
under certain warps or series of warps in order to segx'egate them for the
pattern. Each stick acts as a warp-lifter, making an opening through the
warp plane for the line of weft.
Peruvian loom. See Loom.
Pile weave. The form recorded for the Araucanians consists of short lengths
of yarn knotted around pairs of warp elements, as in Turkish carpets. The
knot is called the Giordes knot (fig. 33).
Plain weave- Simple cloth weave over-one-under-one interlacing.
;
Plaiting. Synonymfor Braiding; a multiple-element technique. It requires at

least three strands, the number being limited only by the ability to interlace
one or several as a unit with the remainder of the group. Each element
isactive and passive at different times. In basketry and matwork, plaiting
usually connotes weaving; in cloth, the term is applied to a narrow band
fabricated by a method other than weaving.
Reef knot. A square knot.
Rio Ucayali loom. See Loom.
Selvage. The finished edge at the side of a breadth of cloth. It is formed by
the turns of the weft around the outside warp.
Set-up. See Warp element.
Sewn basketry. See Coiling.
Shed-stick. A
round slender or heavy bar. For weaving simple cloth, the shed-
stick serves to divide the entire plane of warp elements into two equal parts
for the passage of a line of weft between them. Warps passing beneath the
shed-stick are raised by means of the heddle rod with its pendant loops.
Single- face brocading. See Brocade.
Single-ply. The basal twisted strand as it comes from the spindle or other
spinning device. Double- or two-ply cord represents a combination of two
singles or the folding in half of one single to form a heavier element.
Single-twine technique. See Twining.
Slit tapestry. See Kilin tapestry under Tapestry.
Spindle. A slender, round, pointed stick from 8 to 18 inches in length. Near
one end is the whorl, a disk, or a ball, which steadies the motion when the
spindle is twirled. The Drop spindle (so-called Andean or Bacairi spindle)
;
Vol. 5] GLOSSARY 779

may have a hook, notch, small slit at one end. The supported spindle
oi'
(so-called Bororo spindle) smaller in diameter and has a ball-like whorl.

is
To set this spindle in motion, the spinner rolls the end of it under the palm
of her hand on the block forming one of the supports.
Spiraling weft. See Wickerwork.
Structural decoration. Design motifs or patterns, like stripes, which are built
up during the actual weaving of the cloth on the loom or, in basketry, during
the making of the basket or mat.
Surface decoration. Patterning or objects applied after the basketry or cloth
has been completed. Paint, knotted strings, feathers are among the forms
occurring. (Plate 18.)
Sword. See Batten.
Tapestry. Variation of simple plain or twill weaves to be identified by pattern
motifs of weft yarns battened together so closely as to conceal the warps.
In Interlocking tapestry technique wefts of adjoining color areas turn around
a common warp or they turn around each other in the space between two
warps. In Kilim tapestry the form is distinguished by slits at the edges of
pattern motifs. Colored wefts of one motif turn around its edge warps
colored wefts of adjoining motifs turn around their own edge warps. The
slits left between the motifs may be short or long.
Tenter. A stick placed just behind the working edge of a fabric. Cloth edges
are affixed to the stick in order to maintain uniform width of the fabric.
Thigh spinning. Yarn twisted without a tool. The spinner rolls a small flat
strand of fibers down his bare thigh with the palm of his hand.
Thread count. The number of warp and weft elements per unit of measurement,
usually per square centimeter or inch.
Three-element weaving. See Hexagonal weave.
Twill or twilling. Simple weave characterized by diagonal lines formed by the
intersection of warp and weft floats. Ordinary hand-woven twills have
four warps and four wefts in the unit. These divide as one-and-three
(uneven) and two-and-two (even) twills. In basketry, twilling produces
geometric motifs, frets, and other patterns built up of straight lines (pis.
18, 19, 20, 23; fig. 13).
Twining. A finger technique in which two elements are twisted about each
other frequently enclosing a third element within each twist. Also called
single twine and chain twist (fig. 38 pi. 17).
;
Two-element interlacing. Wickerwork, checkerwork, and twilling are three

simple techniques, each requiring only two elements comparable to the warp
and weft systems in cloth weaving.
Warp element. In woven basketry terminology, synonymous with Foundation
element. In cloth weaving, a single thread or cord extending from one end
bar of a loom to the other. The complete series of warps is called the Warp
skein, the Warp plane, or the Setup.
Warp floats. See Float.
Warp patterning. Motifs produced by raising certain warps to form unbroken
lengths on the surface of the fabric. Weavers on primitive looms use only
the fingers or small sticks as warp lifters.
Warp plane. See Warp element.
Warp skein. The continuous length of warp yarn wound around stakes or bars
preliminary to weaving.
Weaving. The interlacing of two systems of elements, the warp and the weft,
at right angles to each other. Cloth weaving requires the use of a frame
or a loom to hold the warp system taut.
Web. A
completed textile fabric or one under construction on the loom.
Whorl. A
discoidal or small spherical object mounted on the spindle stick to
give steadiness to its whirling motion.
Wicker. Plain over-one-under-one weaving in basketry. The warp element is
usually heavy and rigid in contrast to slender, flexible weft element. Wicker-
work for special uses, such as fishing gear, may be woven with the weft
element spiraling upward from bottom to mouth of the trap.
Wrapped or knotted half-hitch. See Coiling.
METALLURGY
The following list was prepared by Dr. William Eoot

Alloy. A homogenous mixture of two or more metals, usually made by melting
them together.
Amalgam. An alloy of mercury and some other metal.
Anneal. The process of heating a metal to a high temperature for some time
and then slowly cooling it. This makes the metal softer and less brittle.
See Temper.
Brass. An alloy of copper and zinc.
Bronze. An alloy of copper and tin.
Casting. The process by which an object is made by pouring molten metal into
a mold and allowing it to solidify.
Cire perdue process. This is a special method of preparing a mold for casting.
A wax model of the object to be cast is made. This is coated with moist
clay which is allowed to dry. The wax is then melted and poured out leaving
a mold of the desired form.
Coloring ( of a metal ) The process by which a metal surface is made to assume
.
a different color than the original metal. The color comes from some con-
stituent of the metal itself. See Mise en couleur. Not to be confused with
Gilding.
Density. The weight of a unit volume of a material. It is usually expressed
as grams per cubic centimeter.
Ductility.The quality of being capable of being permanently drawn out into
wire or of being hammered into a thin sheet without cracking. See
Malleability.
Electrum. An alloy of gold and silver.
Emboss. To decorate a metal surface by raised designs. See Repousse.
Engrave. To decorate an object with a design scratched into the surface.
Eutectic. The alloy that has the lowest possible melting point with the given
components.
Forging. To hammer heavy metal.
Gilding. To apply gold leaf, etc., to a surface. See Coloring.
Guanin. See Tumbaga.
Hardness. The resistance of a substance to denting or scratching. It is usually
measured on an arbitrary scale. The Brinell hardness number is the com-
monest, and is measured by the indentation effect of a hard ball pressed into
the surface of the metal to be tested.
Malleability. The quality of being capable of being hammered into thin sheets
without cracking. See Ductility.
Metal. A substance with metallic properties. It may be a native metal or an
alloy.
Mineral. A naturally occurring compound, usually of a metal. See Ore.
Vol.5] GLOSSARY 781
Mise en couleur. On treatment of a copper-gold alloy with acid, the copper is

removed from the surface and the gold is left. On burnishing, the surface
appears to be of gold. This is an example of Coloring.
Native metal. Metal that occurs in nature in the free state. The common
examples are gold, silver, and copper.
Ore. A naturally occurring metal or mineral usually mixed with impurities.
Refining. The purification of a metal.
Repousse. A design in relief on a metal sheet beaten or pressed up from the
reverse side.
Smelting. The process of winning a metal from its ores by heating the ore with
a substance like charcoal to a temperature above the melting point of the
metal. Sometimes used in the sense of melting or fusing.
Solder. To join together two pieces of metal by means of a third substance (the
solder). The solder is usually a lower melting metal or alloy. When the
solder melts, it alloys with the metal along the two edges and on cooling
forms a firm joint.
Tempering. To bring a metal to a proper degree of hardness and toughness,
usually by heating and cooling more or less quiclily. See Anneal.
Tensile strength. The force required to pull apart a substance. It is usually
expressed in pounds per square inch or in kilograms per square centimeter.
Tumbaga. An alloy of gold and copper. Also called Guanin.
Weld. To join together two pieces of metal either by (1) melting the edges
together (autogenous welding), or (2) hammering or pressing together the
overlapping edges while heated almost to their melting point (forge welding)
SOCIAL AND RELIGIOUS TERMS
Ethnology has a growing vocabulary of special terms, and each

is gradually acquiring a finer shade of meaning. Some day, it will
be worth while to compile a dictionary of such terms. For the
present, there is too much variation in their use by individual
scientists to permit standardization.
The following list selects only a few of the more common terms,
especially those which, being adopted from one language into an-
other, have taken on new meaning. The very word "anthropology"
is an example of changed meaning. In South America, anthropology,
following European usage, signifies what "physical anthropology"
means in North America. "Anthropology" in North America has
come to be an inclusive term, covering archeology, ethnology, and
physical anthropology. The definitions given below are not attempts
to fix the meanings of these terms, but, in the interest of international
intelligibility, to indicate their meanings in the Handbook, and in
North America.
Clan. Used both as a synonym for sib (see Sib) and as a term to distinguish a
matrilineal sib from a patrilineal sib.
Couvade. Literally, this means that the father goes
to bed and observes other
restrictions as if he rather than the mother had borne the child. In
usage, it has come to mean tliat the father no less than the mother is
restricted in his behavior after childbirth because native belief holds that
his behavior will affect the child. Thus, terms such as semicouvade,
meaning that the father is only somewhat restricted in his behavior, have
come into use.
Endogamy. The custom that one marry inside the group, for example, com-
munity, moiety, etc. The clan, however, is usually though not always,
exogamous by definition.
Exogamy. The rule that one must marry outside the group, for example, the
clan, moiety, band, community, etc.
Folklore. This term tends to have different, though not mutually exclusive^
meanings in North and South America. In North America, it is used princi-
pally for folk tales, though it also includes folk beliefs and customs, and
mythology. In South America, it signifies the folk beliefs and customs,
especially of contemporary, backwai'd peoples, and its meaning is often so
broad as to be almost synonymous with "ethnography" in the North American
sense.
Gens. A patrilineal sib.
Guardian spirit or spirit helper. A supernatural being, such as a plant, animal,
or other spirit, which is associated with and assists a particular person.
By contrast, a god is a supernatural being which is associated with and
assists a community, tribe, or nation.
Lineage. The descendants of one pair of ancestors, through either the male or
female line.
Medicine man. See Shaman.
Moiety. One The term has
of dual divisions of a tribe or sociopolitical unit.
two definitions: a division of the unit into two groups for games,
(1)
reciprocal burial rites, ceremonies, and other such purposes; (2) one of two
exogamous divisions. Exogamous moieties may be subdivided into clans.
Phratry. One of three or more tribal divisions which are usually exogamous
and which are subdivided into clans or sibs. If there were only two such
divisions, they would be called moieties.
Priest. The intermediary between the people and their gods. In contrast to
the shaman, the priest lacks a personal spirit helper and conducts rites and
group ceremonialism for the tribal god. An individual may, however, be
both priest and shaman.
Shaman. Medicine man (curandero). A person whose power is based on his
personal control of supernatural forces, usually guardian spirits. He diag-
noses and cures disease, works magic, prognosticates, and often has super-
natural control over other phenomena.
Sib. An exogamous group. (See also Clan and Gens.) This term may be
applied to: (1) a lineage, i. e., a group of relatives in the male or female
line, who live together in one place and take their spouses from other groups
(2) a group of persons descended through the male or female line, who may
be scattered and cannot trace their relationship to one another but who
believe that they are descended from the same ancestor and marry outside
the group.
Spirit helper. The supernatural being which the shaman or medicine man uses
as his agent in causing and curing disease and working other magic. The
spirit is usually acquired through a dream or vision, which is sometimes pro-
duced by narcotics or other means.
Totem. A spirit, as of a plant, animal, or other phenomenon, believed to have
a special relationship to a group, such as a clan or moiety. It is contrasted
to the individual's guardian spirit and the tribe's god. The relationship of
the totem to the group is extremely varied, and many shades of meaning have
been attached to "totemism."

Handbook Steward

Uploaded by

Copyright:

Available Formats

Handbook Steward

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Handbook Steward

Uploaded by

Copyright:

Available Formats

SMITHSONIAN INSTITUTION

BUREAU OF AMERICAN ETHNOLOGY

Prepared in Cooperation With the United States Department of State as a Project

Digitalizado pelo Internet Archive.

It is the purpose of the present article to provide a basis for classi-

The fourfold classification used herein (map 18) corresponds in

1 1 am grateful to Drs. A. L. Kroeber, Robert H. Lowie, W. D. Strong, John M. Cooper,

670 SOUTH AMERICAN INDIANS [B. A. E. Bull. 143

( Solid black, Central Andes

tribes were improperly classified. In the preceding volumes of the

est peoples were identified by their agriculture and various material

672 SOUTH AMERICAN INDIANS [B. A. E. Bull. 143

loom weaving, twilled and woven

association with one another, participate in the same economic, social,

to a group of units which, though culturally and linguistically similar,

often, it is the extended family or lineage ; frequently, it is the multi-

community state, federation, or empire.

(1) The Marginal peoples had sociopolitical units which consisted

cult with a hierarchy of gods and priests public ceremonies forming

a ritual cycle and regimentation of a considerable portion of the lives

of the commoners through political controls enforced by the state more

674 SOUTH AMERICAN INDIANS [B. A. E. Bull. 143

metallurgy and the building arts, but it is outstanding in the quantity,

TheYahgan, for example, quite obviously had to live in small, widely

Vol.5] SOUTH AMERICAN CULTURES STEWARD 675

Population densities suggest potentialities for community develop-

Vol.5] SOUTH AMERICAN CULTURES —STEWARD 677

a lineage, were clustered inside a palisade. In some areas, however,

These general, comparative data on population densities and com-

THE MARGINAL TRIBES

The Marginal tribes were essentially hunters and gatherers, and

and arrow types.

mocks (some Ge^ Garari, MasJiacali^ Carajd), rubber balls (ApinayS,

Gamacdn, '^Tapuya''')^ penis thread {^''Tainiya^'' Patasho), feather-

Stripped of these borrowings, the Marginal cultures would be ex-

THE ECOLOGICAL BASIS

Subsistence. —The subsistence potentialities of the different Mar-

Andes, with its potatoes and cereal-crop staples, agriculture was

The main crop among most tribes was maize

Yaruro, presumably a practice of incipient farmers. The tropical

The canoe was of major importance to the Tropical Forest peoples,

{Guaja, /Siriono, Mt(/ra, Yaruro). Where construction is described,

Guato, Puri-Goroado, recent Gaingang). The large, frame multi-

SOCIOPOLITICAL AND RELIGIOUS PATTERNS

Vol.5] SOUTH AMERICAN CULTURES —STEWARD 683

to their husband's family. In this case, it consisted of husband, wife,

ous gradations from the independent conjugal family to the patri-

what institutionalized through the initiation of pubescent boys and

The Siriono (Handbook, 3: 458) had extended matrilineal families

Map 20.—Distribution of the types of kinship basis of aboriginal sociopolitical

traveled together and the constituent conjugal families often lived

Marginal level, though sociopolitically they probably resembled their

they obtained early in the 18th century from the Spaniards.

wise, the main sexual differences were in division of labor.

death {Gharrua^ Querandi), skull trophies {Gharrua, Querandi), and

by group rather than by locality, whereas a lineage occurs only in one

688 SOUTH AMERICAN INDIANS [B. A. E. Bull. 143

such multilineage groups, and there is perhaps some significance in the

(4) Sociopolitical units of uncertain composition. The Boto- —

Vol.5] SOUTH AMERICAN CULTURES —STEWARD 691