On Population Structure in Generalized Hunting and Collecting Populations

Author(s): Joseph B. Birdsell

Source: Evolution, Vol. 12, No. 2 (Jun., 1958), pp. 189-205
Published by: Society for the Study of Evolution
Stable URL: http://www.jstor.org/stable/2406029
Accessed: 14-01-2016 16:37 UTC

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/
info/about/policies/terms.jsp

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content
in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship.
For more information about JSTOR, please contact [email protected].

Wiley and Society for the Study of Evolution are collaborating with JSTOR to digitize, preserve and extend access to
Evolution.

http://www.jstor.org

This content downloaded from 132.239.1.231 on Thu, 14 Jan 2016 16:37:42 UTC
All use subject to JSTOR Terms and Conditions
 ON POPULATION STRUCTURE IN GENERALIZED
HUNTING AND COLLECTING POPULATIONS 1

JOSEPH B. BIRDSELL

University of Californtia at Los A4tgeles

Received November 9, 1957

INTRODUCTION meat products through the winter months.

The same factors suggest the possible in-
For the purposes of this research gen-
eralized hunting and collecting popula- clusion of the Caribou Eskimo. In South
tions may be defined as those peoples de- America the Ona of Tierra del Fuego and
pendent for sustenance on a wide variety the adjacent coastal and more northerly
of foods obtained by hunting and collect- plains tribes belong in this classification.
ing methods in patterns which change Several groups of peoples technically
seasonally, and among whom food stor- classed at the hunting and collecting level
age and preservation are of little or no of economy must be explicitly excluded.
consequence. They have no domestic The basis for exclusion generally rests
plants, nor any domesticated animals other upon their access at least seasonally to
than the dog, which is primarily utilized rich and abundant food resources of lim-
as an adjunct to hunting and not as a ited variety combined with more effective
direct food source. More than a dozen methods of preservation, so that survival
regional clusters of peoples at this simple is less upon a day-to-day basis and some
economic level existed at the time of his- accumulation of wealth is possible. These
torical contact. The Australians and the richer sources of energy influence density,
Tasmanians in Australasia are well known population structure, and social and politi-
examples. Negritic peoples in eastern cal organization in such a way that the
Asia at this level include the Aeta of the generalizations derived from the more
Philippines, the Semang of the Malay stringent economies no longer apply. The
Peninsula, and the inhabitants of the acorn gathering Indians in California are
Andaman Islands. In Africa all of the excluded since the abundance of this pri-
so-called Bushmen tribes and the rain mary source of food and its easy preserva-
forest Negritoes of the Congo Basin are tion renders them more analogous in
in this economic group. A small group density to the primitive horticultural peo-
of jungle tribes in India falls into this ples of the world. The tribes in the
category. In the New World the Sho- Northwest coast area, for whom salmon
shoni tribes of the Great Basin belong provided the primary food source, are
here, despite the relative importance of likewise omitted from consideration. The
the pi-non nut in their economy and the horse nomads of the Great Plains may,
ease with which it lends itself to preserva- during the worst seasons of the year, ap-
tion during winter months. It is probable proximate our generalized conditions, but
that the inland Athabascan and Algonquin during the buffalo hunting season their
speaking peoples of Canada belong with culture took on more complex aspects and
the others in spite of assistance provided for this reason they will be disregarded.
by the climate for the preservation of The coastal Eskimo perhaps present a
borderline case, but since during the criti-
1 Presented at symposium on "The Evolu-
tionary Aspects of Human Behavior" at Twelfth cal winter months their survival depended
Annual Meeting of the Society for the Study of upon hunting of a limited variety of sea
Evolution, Stanford University, August 12, 1957. mammals they do not seem generalized
EVOLUTION 12: 189-205. June, 1958. 189

This content downloaded from 132.239.1.231 on Thu, 14 Jan 2016 16:37:42 UTC
All use subject to JSTOR Terms and Conditions
190 JOSEPH B. BIRDSELL

enough in their economy for inclusion The regularities to be explored in the

here. remaining sections of this paper cluster
The bearers of these simple generalized about three themes: (A) territoriality;
hunting and collecting economies include (B) population dynamics; (C) cultural
representatives, at least in hybridized cohesion. These formulations are to be
form, of all the major racial groups. They considered preliminary in nature since the
live through the near maximum range of quantified data upon which they depend
earthly environments with mean annual are extraordinarily rare in the literature
rainfall ranging from a minimum of 4 and opportunity has not yet allowed com-
inches to values in excess of 200 inches; prehensive search of all pertinent publica-
and in temperatures that range from des- tions. Enough sources have been covered
ert tropics, through tropical rain forests, to indicate the general range of phe-
to temperate zone and probably into the nomena, but detailed documentation must
continental Arctic. Culturally,their typol- await time for further elaboration. These
ogies range from the paleolithic of the generalizations are not offered in a dog-
Tasmanians, through epipaleolithic forms, matic spirit. It is to be expected that
into those which include contributions exceptions to them will occasionally occur
from obviously neolithic sources. At this among the peoples under consideration.
economic level, with its simple day-to-day But it is believed that they have a high
pursuit of food, there are certain broad probability of becoming uniting principles
similarities among all the peoples who be- for the analysis of this level of economy.
long to it. A number of these are well
recognized in the existing anthropological (A) Territoriality
literature. Such peoples are characterized
by having little or no accumulated wealth, Like most terrestial vertebrates, man
they lack class differentiation, they are at generalized hunting and collecting levels
united by no effective political authority, of economy shows clear-cut territoriality.
and organized warfare is lacking. Their It is expressed in hunting and collecting
densities in the best cases seldom range rights in a given area and not of possession
above 1 to 2 persons per square mile and of realty. In most groups it is vested in
in the worst environments 100 square the local band, or horde, and boundaries
miles may be required to support each are comlmionlydefined by natural topo-
individual. graphic features. In flat, arid areas bound-
aries may be functionally defined in terms
DISCUSSION
of waterless stretches. The widespread
occurrence of territoriality among hu-
In addition to these readily visible gen- mans, as well as among infrahuman groups
eralizations, there exist others which do of terrestial vertebrates, seems to be a
not rise easily to the surface of unquan- consequence of the need to insure that
tified data. Ethnographers with few ex- for species survival the best breeding
ceptions have systematically particularized areas of a range should not be over-
the culture of each people. They have not populated. Among a few generalized
only resisted the search for underlying uni- hunting and collecting peoples, such as
formities, but have failed to ask the ques- the Shoshoni, territoriality in the con-
tions which would lead to their discovery. ventional sense does not seem to exist
The generally humanistic approach which (Steward, 1938). But since this same
characterizes the work of most anthro- group of tribes in an over-all sense show
pologists has failed to provide the quan- that population density is rather rigor-
tified data which are needed to explore ously determined by environmental varia-
relationships between variables within a bles (as indicated below), it must be as-
single culture, or across many cultures. sumed that some informal and amorphous

This content downloaded from 132.239.1.231 on Thu, 14 Jan 2016 16:37:42 UTC
All use subject to JSTOR Terms and Conditions
 HUNTING AND COLLECTING POPULATIONS 191

system controls population numbers with- for a much smaller series of tribes. As in
out rigidly defined territories being in- the earlier Australian analysis a single en-
volved. Even the best of ethnographers, vironmental variable, mean annual rain-
and Steward is certainly among these, fall, was utilized in the analysis. Since
provide little data for the solution of such the Shoshoni show a much narrower
questions, since they are not customarily range in this variable, from 4 to 14 inches
emphasized. annually, it will be seen that her sample
Anthropologists for a number of decades covers only a small sector of the com-
have tacitly maintained that human popu- parable Australian data. By reanalyzing
lations are free from all forms of environ- that portion of the Australian sample
mental determinism. The sweep of this covering the same limited rainfall range
assertion, which resulted as a reaction to she has shown that her coefficient of cor-
earlier claims that many aspects of culture relation falls but little below that for a
were determined by the environment, has comparable sector of the Australian data.
obscured the fact that at the simple levels It would seem that the Shoshoni materials
of generalized hunting and collecting econ- reveal a lower correlation coefficient pri-
omies population densities may be pri- marily as a result of the more limited
marily determined by environmental fac- range of the primary variable. It may be
tors. This is not to deny that at any concluded that the Shoshoni are as rigor-
human level population densities are also ously determined in their population den-
subject to the influence of cultural varia- sities as are the Australian aborigines.
bles, but these factors may be less im- This parallel is the more remarkable in
portant than those inherent in the en- the light of the contrast in the economies.
vironment in the determination of densi- The Shoshoni subsist almost entirely upon
ties considered here. The first evidence plant products and make large use of the
for the rigorous environmental determina- pi-non nut harvest, a product which is
tion of human density was presented easily stored and upon which they are
(Birdsell, 1953) for the Australian abo- dependent for their winter sustenance.
rigines. In that arid continent it was The Australians are to a much greater
found that mean annual rainfall taken degree hunters and have little or no food
alone constituted the most important en- preservation. The contrast between the
vironmental determinant of native den- two groups is further heightened by dif-
sities and that the curvilinear coefficient ferences in topography. Most Australian
of correlation, r = .8. Rainfall for these bands live in country of relatively low re-
tribal groups ranged from 5 to 151 inches lief and the altitudinal effects of changing
for annual average, and its relationship to flora and fauna are generally absent. On
tribal area with an inverse logarithmic the other hand, the Shoshoni families
function. This relationship was based customarily range from the desert floor at
upon 123 tribal groups which were con- the proper season of the year, to the high
sidered ecologically homogeneous in na- altitudes of the numerous mountain
ture. It seems clear that if other effective ranges which cross the Great Basin and
environmental variables had been intro- which provide the pi-nonnut harvest upon
duced into the analysis at that time, the which they are so dependent. The Sho-
coefficient of correlation would have been shoni show considerably higher densities
even higher. for a given value of rainfall than do the
In a recent but as yet unpublished Australians. Since the general levels of
study, Vorapich has taken Steward's data technology are similar it must be assumed
for the Great Basin Shoshoni and demon- that these altitudinal variations allow the
strated a similar environmental determina- greater population density found among
tion of human densities in that area. Her the Shoshoni.
coefficient of correlations approximates .5 These two analyses of the environ-

This content downloaded from 132.239.1.231 on Thu, 14 Jan 2016 16:37:42 UTC
All use subject to JSTOR Terms and Conditions
192 JOSEPH B. BIRDSELL

mental determinants of human densities corrective forces exist which tend to re-
at generalized level of hunting and col- turn densities to their proper values in
lecting populations are the by-products of terms of available resources. Such cor-
meticulous field work by two individual rective forces may be inherent in the
investigators. The Australian data are dynamics of the population, in the en-
based upon the extraordinarily detailed vironment itself, as in times of drought or
tribal maps for that continent published by failure of food resources for any other
Tindale (1940), whereas Vorapich's treat- reason, or they may involve cultural fac-
ment rests upon the very able field work tors which operate so as to help maintain
of Steward (1938). It may be classed an essentially stable population.
as one of the professional misfortunes of
anthropology that few other field workers (B) Population Dynamics
have realized the great potential of such
detailed studies of population distribu- One of the basic parameters in any
tions, so comparative studies for other evaluation of the role of population dy-
groups may not be possible now or in namics of a people involves the estima-
the future. But the qualitative descrip- tion of the intrinsic rate of increase, the
tions in the literature strongly suggest rate of increase in an unlimited environ-
that other populations at comparablelevels ment. While this may be determined with
of culture show densities also primarily some facility in experimental populations
determined by environmental variables. of animals, man seldom provides such a
It should be noted with some caution that situation. In a recent analysis it has been
whereas these two groups, living in arid shown (Birdsell, in press) that four re-
regions, show a high correlation with a corded instances in human history give
single environmental variable, and this in some basis for making an estimation.
both cases, mean annual rainfall, that in Three of these examples occurred in the
other climates and places a multivariant insular environments of Tristan da Cunha
analysis of the environmental factors con- in the South Atlantic, Pitcairn Island in
ditioning density would undoubtedly be the eastern Central Pacific, and Cape Bar-
necessary. The equation which relates ren Island in Bass Strait between the
primitive human densities to their en- mainland of Australia and Tasmania. In
vironment does not allow projection from each instance available data suggests that
one region to another.- For each local these simple horticultural populations fol-
group of populations the fuller array of low an intrinsic rate of increase in which
environmental determinants must be ana- the population doubled in each generation.
lyzed. The local flora and fauna available This rate continued essentially unchanged
to the human populations will vary ac- until the population reached a point at
cording to the phylogenetic history in which the food resources of the environ-
that region. Finally, the ability of each ment exert a marked depressant effect.
human group to utilize these sources of In fact, they overshoot the theoretical
energy will vary with their extractive ef- curve of approach to the equilibriumvalue.
ficiency as determined by their technology. Since it may be questioned whether the
The working assumption based upon data from these horticultural populations
the two previous examples that general- would apply to the generalized level of
ized hunting and collecting populations hunting and collecting peoples, it is for-
show densities primarily determined by tunate that a fragmentary record remains
environmental factors clearly implies that for a similar experimental population in
such populations are in stable equilibrium Australia. There, in an interesting late
with the carrying capacity of their coun- 19th century account, is recorded the fate
try. Equilibrium as used here allows for of a Marawra tribesman who, accom-
minor oscillations in time but implies that panied by two of his wives, disappeared

This content downloaded from 132.239.1.231 on Thu, 14 Jan 2016 16:37:42 UTC
All use subject to JSTOR Terms and Conditions
 HUNTING AND COLLECTING POPULATIONS 193

into the uninhabited mallee country to the that, where a high infant mortality fails
west of a frontier settlement on the Dar- to balance population numbers, infanticide
ling River. When found again thirty is generally resorted to. Quantitative
years after the original exodus, this little data on rate of infanticide are rare but
group of aborigines had increased to 28 several accounts suggest that it custom-
in number. Fragmentary as the data are, arily ranged to as high as 30 per cent of
they indicate that for this group the in- the births among some of the Australian
trinsic rate of increase was closer to that aborigines. Whereas infanticide is clearly
of tripling the population each generation in the realm of cultural variables which
than of doubling it. It lends some as- tend to control population at the equi-
surance that the data from the other three librium level, it is based upon the rigorous
examples do, in fact, apply to the in- requirements of the environment.
trinsic rate of increase of hunting and This protracted period of nursing of the
collecting people in an unlimited environ- human infant at generalized hunting and
ment. Until better data are provided, it collecting levels of economy results in a
is safe to assume that their numbers will low variance in the effective number of
double with the passage of each genera- offspring per mating and consequently
tion. influences some of the evolutionary proc-
This potential rate of increase which esses which operate on such populations.
seems to characterize even simple human At higher cultural levels, as among pas-
populations requires some reconciliation tural and agricultural people, the earlier
with the concept of stable populations as weaning of offspring allows for their
determined by environmental variables. much more frequent procreation. Among
In a broad sense, it is clear that the po- the simpler populations, the scanty evi-
tential for population expansion acts as a dence suggests that the range of effective
buffer against the threat of extinction offspring per mating ranges from 0 to
which lies inherent in the vagaries of life perhaps 6, whereas at higher cultural
at this technological level in an uncon- levels it may be more nearly from 0 to 12.
trolled environment. The forces which Thus variance of the offspring per mating
contain this potential rate of expansion among the simpler economic groups ef-
are both environmental and cultural in fectively approaches a constant value and
nature. It is characteristic of generalized it may be estimated to be equal to 2, or
hunting and collecting peoples that proper perhaps slightly less, regardless of race
foods for infants are lacking. Without and climate. This constant variance af-
domestic animals only mother's milk is fects the size of the effective breeding
available. In the absence of agriculture, populations and simplifies its calculations
starches of the proper type for infant feed- for such groups where primary and vital
ing are scarce or absent. A direct con- statistics are so scanty. In the equation
sequence of this environmental pressure given by Wright (1938)
is that the weaning of infants is much = 4N-2
delayed. Descriptive accounts generally
cTK2 + 2-
agree that such people nurse their infants
from 2 to 4 years after birth, and it may In essence, the equation means that the
reasonably be assumed that the average effective size of the population for evolu-
duration of breast feeding is close to three tionary processes is a direct function of
years. The importance for survival of the size of the breeding population and an
band and individual mobility combined inverse function of the variance of the ef-
with the hard living and variable nutri- fective number of offspring produced. It
tion available at these cultural levels pre- is here contended that for generalized
cludes the nursing of two children simul- hunting and collecting people the variance
taneously. It is not entirely surprising is essentially a constant owing to the pro-

This content downloaded from 132.239.1.231 on Thu, 14 Jan 2016 16:37:42 UTC
All use subject to JSTOR Terms and Conditions
194 JOSEPH B. BIRDSELL

tracted nursing period in such cultures, into the belief that such systems involve
and hence the effective size of the popula- actual marriages between biological cous-
tion becomes directly related to the size ins with a high degree of frequency. It
of the breeding population alone. This will be shown elsewhere that the kinship
regularity is of some importance in con- abstraction does not have biological valid-
sidering how random genetic drift will in- ity; that where quantitative data are avail-
fluence such groups for whom demo- able the frequency of such biological mat-
graphic data are exceedingly scarce. ings is low, that populations show no sig-
Another influence which is commonly nificant increase in genetic homozygosity.
considered as a potential factor in the It may safely be concluded that general-
structuring of human populations is known ized hunting and collecting peoples do not
as assortative mating. Positive assorta- in a genetical sense show any significant
tive mating involves likes marrying likes degree of inbreeding.
on a phenotypic basis. Its consequence is
to increase genetic homozygosity and to (C) Factors Affecting Cultural Cohesion
diminish heterozygosity. Negative as-
sortative mating produces opposite results Three basic social units seem to be con-
with an increase of heterozygosity. It stantly present among people at a gen-
has yet to be demonstrated that either of eralized hunting and collecting economy:
these forms of mating systems are of (1) the biological family; (2) the band;
evolutionary significance in any human (3) the tribe. The biological family oc-
population. Among the uncomplicated curs in one form or another among all
populations considered here, women are human groups. Among the groups under
universally married at an early age. The discussion it either takes the form of
number of permanently unmarried adults monogamy or goes further and allow-s a
in such groups is, as a rule, of the mag- minority of males to practice polygamous
nitude of 1 or 2 per cent. Under these marriage. It is frequently claimed that
circumstances, even though the detailed the human biological family is cultural
data are wanting, it is safe to assume that invention. While it is true that it occurs
assortative mating, whether in positive or in a number of forms, it is better con-
negative form, is not a factor in popula- sidered as a biological prerequisite for
tion structuring. species survival. Among hunting and
Another system of mating which may collecting peoples, the economic partner-
possibly influence populations is known ship between an adult male and one or
as inbreeding. This process involves mat- more adult females is a necessity for the
ing between biologically related individ- successful rearing of the next generation.
uals and in experimental populations it The next largest basic social unit among
does have the effect of markedly increas- the peoples under discussion is known as
ing genetic homozygosity. As applied to the band, horde, or local group. In the
natural human populations, however, the anthropological literature it is rather
term has been much abused. In part this loosely defined either as an extended
distortion has arisen from the fact that family or a group of several related fam-
many anthropologists have confused clas- ilies. While quantitative data are lacking
sificatory kinship systems of mating with for a proper analysis, it is clear that
the biological systems which actually pre- among most groups the band is larger
vail in populations. Thus a fair minority than would be reasonably expected from
of the world's people show a classificatory among the survivors of a single extended
preference for marriage between first family. Few data exist to indicate how
cousins, either cross or parallel. This the several families may, in fact, be bio-
social abstraction is all too frequently logically or socially related. Band size
translated among anthropological circles varies considerably among different groups

This content downloaded from 132.239.1.231 on Thu, 14 Jan 2016 16:37:42 UTC
All use subject to JSTOR Terms and Conditions
 HUNTING AND COLLECTING POPULATIONS 195

and the range may perhaps be estimated is characteristic of the southern Bushmen,
as averaging 20 to 100 individuals. An- the Andamanese, many Canadian Algon-
thropologists in general have not examined kins, and Athabascans. Steward has in-
the composition of the band among their dicated that theoretically a third type of
respective peoples and have had little in- band, which would be the matrilineal
terest in the social and environmental counterpart of the first, should exist
forces which tend to produce human among some primitive agriculturists, but
groups of fairly constant size which func- these are outside the scope of our interest
tion as a basic social unit for at least a here.
portion of the seasonal cycle. Some system of equilibrium forces, in-
An unusual awareness of the general volving both ecological and cultural fac-
problem has been shown by Steward who tors, tend to maintain appropriate popula-
says (1938, p. 258), "All peoples in an tion size for band structure for given local
area of low population density have some conditions. Since anthropologists have
form of politically autonomous, landown- not defined details of the nature of the
ing band, which is greater than the bi- cultural forces involved and have gathered
lateral family. The size of the band and no materials which allow a detailed ex-
the extent of the territory it utilizes are amination of the ecological factors, it is
determined by the number of persons who, impossible to construct models to explore
due largely to ecological factors, habitually the systems occurring among various
cooperate at least during part of the an- peoples. On the other hand, it does seem
nual round of economic and social ac- clear that band size is not primarily de-
tivities. Band unity is expressed in a termined by chance factors and that sys-
consciousness of common interest and sub- tematic forces do exist. One may hope
mission to some degree of central control that future generations of field workers
during the community enterprises, al- may collect enough demographic and eco-
though such control may be lacking dur- logical data to provide a basis for such
ing parts of the year. The authority of investigation.
the leader is consequently small and tem- The third basic unit, the tribe, is an-
porary and its position is seldom a fixed other loosely defined anthropological ab-
institution." He goes on to define the straction. In essence, it refers to a cluster
two types of bands which occur among of culturally and linguistically related
generalized hunting and collecting peo- groups of peoples but may range in size
ples. The commonest consists of the uni- from a few hundred to several millions of
lineal, exogamous, patrilineal, patrilocal persons. This range of variation occurs
band which consists of actual or fictitious as a continuum so that arbitrary distinc-
relations. It occurs among peoples with tions between various forms of tribes have
an emphasis on hunting and hence is male- not been defined. The factor most im-
dominant culturally. The ecology that portantly affecting tribal size seems to be
prevails among such peoples prevents the degree of centralized political authority.
size of the band from ranging much out- It is perhaps sufficient for this problem to
side of 50 to 100 people. It is character- state that tribes among generalized hunt-
istic among Australians, Tasmanians, Se- ing and collecting peoples are not political
mang, African negritoes, northern Bush- entities, and that central political authority
men, Fuegians, and Southern Californians. is largely or completely absent. Such
The second type of band is known as the tribes do show some cohesiveness but it is
composite band. Since it consists of un- based upon linguistic and cultural rela-
related families and is not unilineal, its tionships alone, as reinforced by spatial
members need not observe band exogamy. relationships. It seems to have been pre-
Ecological factors allow the size of these viously unnoted that there is a broad
bands to approximate 100 individuals. It tendency for the aggregation of related

This content downloaded from 132.239.1.231 on Thu, 14 Jan 2016 16:37:42 UTC
All use subject to JSTOR Terms and Conditions
196 JOSEPH B. BIRDSELL

bands which comprise such tribes to show economic groups, even though this paper
modal values approximating 500 indi- will attempt to outline in exploratory
viduals. This is true for the Australians fashion only two of the more important
(Birdsell, 1953) and the Tasmanians, opposing forces. It must not be assumed
both of which were characterized by patri- that a stable equilibrium involves either
lineal exogamous bands. It seems to also a single system of balance or that the re-
have been true of the Andamanese, with storing forces operate in a direct, un-
their composite band structure. Other complicated fashion. The empirical data
groups may reveal a tendency toward from Australia determining the statistical
somewhat higher values as a central sta- abstraction which describes the average
tistical tendency. It would be of interest size of tribal population approximating
to test whether these tendencies toward 500 individuals in fact represents a fairly
a constant size of tribal population are a wide range of values, and indicate that
direct function of the size of the bands serious disturbances of the balanced condi-
interacting as seems most likely or whether tion may not be compensated for until
the limiting factors lie in the limitations tribal population actually falls below 200
upon interactions among a given number individuals or exceeds 800 individuals.
of individuals. In the former case, one The process of restoration may well in-
would expect average tribal size to vary volve oscillations rather than a single di-
directly with the mean number of persons rect return to an ideal balance.
making up the bands. In the latter in- One of the primary factors in the sys-
stance total tribal populations would tend tem of equilibrium involves the number
to approximate a constant size regardless of local groups or bands which cohere to
of band size. Unfortunately the data to form a single relative homogeneous lin-
make such tests do not seem to be pub- guistic and cultural unit, or tribe. An-
lished. In any case, the constancy of other and opposing force depends on dis-
population size among the tribes under tance as an isolating mechanism in cul-
consideration imply a system of equi- tural behavior. Both of these forces are
librium forces. The remainder of this based in a spatial geometry and it is pro-
paper will be devoted to the tentative posed to explore some of their idealized
identification of some of these. Since characteristics through simple geometric
empirical data are lacking, the stabilizing models. In the exogamous patrilineal
forces postulated will have to be explored band structure characterizing most of
by means of model constructs. The fol- Australia, cultural impact and change
lowing models are intended to apply only primarily operate through band contact.
to exogamous patrilineal band organiza- This is not to say that interband and in-
tion. tertribal contacts between families do not
Social equilibrium, like mechanical equi- occur, but rather that normally they seem
librium, results when a system of forces to be conducted by bands. At the first
act in such a way that deviations from sight it would seem that intertribal mar-
the normative position set up restoring riages might be the mechanism of cultural
forces which tend to re-establish the bal- diffusion in the continent of Australia,
anced condition existing prior to the dis- but these are universally patrilocal, since
turbance. Such a stable condition may be the wife joins the band of the husband.
obtained in the most elemental case with The woman's position there is such that
but two opposing forces, but it would be she has little chance to influence language,
unrealistic to assume that the forces act ceremony, or the male functions in the
so simply. It will be better to consider total culture. These imported wives are
that an unknown number of forces are in- silent nonentities until they learn the lan-
volved in the suspension system of bal- guage of their husband's group, and then
ancing tribal numbers in these simple they are quickly absorbed into the hus-

This content downloaded from 132.239.1.231 on Thu, 14 Jan 2016 16:37:42 UTC
All use subject to JSTOR Terms and Conditions
 HUNTING AND COLLECTING POPULATIONS 197

INTRATRIBAL
CONTACTS =0 =12 =40
EXTRATRIBALCONTACTS =4 = 2 =20
TOTALCONTACTS=4 =24 =60

CONTACTS =0
INTRATRIBAL = 12 =42 /
CONTACTS- =6
EXTRATRIBAL =18 =30
TOTAL =6
CONTACTS =30 =72

FIG. 1. Geometrical arrangements to show contacts available to local groups.

band's band and tribe. On the other hand ber of extra-band contacts physically, that
intertribal marriages may be treated as is geographically, available to each local
systematic expressions of the rates and group. The upper 3 diagrams are based
directions of intertribal and interband in- upon the assumption that each band will
teractions. For it cannot be doubted that have 4 neighboring local groups with
wives are obtained from groups with which it has effective contacts along com-
whom social relations are at least oc- mon boundaries, with apical contacts ig-
casionally established. Thus rates of in- nored. The lower series of 3 diagrams
tertribal marriages may be taken as a are based upon the assumption that each
secondary index expressing in indirect local group will have contiguous bound-
fashion the frequencies and directions of aries with 6 other adjacent bands. The
intertribal contacts which have never been five-sided configuration must be ignored
systematically observed. since it cannot be graphically shown in
two dimensional space. The left hand
MODEL OF SPATIAL EFFECTS UPON BAND
diagram in both rows represents the as-
SOCIAL INTERACTIONS
sumption that the single band is the unit
In figure 1 a series of simple geometri- of cultural cohesiveness and that culture
cal arrangements are shown involving two will change uniformly if gradually in
idealized assumptions concerning the num- crossing each band boundary. The inte-

This content downloaded from 132.239.1.231 on Thu, 14 Jan 2016 16:37:42 UTC
All use subject to JSTOR Terms and Conditions
198 JOSEPH B. BIRDSELL

10oe 100

90 -90
-0--0--- HORDES SQUARE
HORDESHEXAGONAL
80- ' 80
80) MEAN RATE OF AUSTRALIAN
W EXTRA-TRIBALMARRIAGES
~70 -70

E60 60

M 50 Q 50

0~~~~~~~~~~~~~~~~~~~~~4
? -0----- _o
I 20 20

z 10~~~~~~~~~~~~~~~~~~~~~~~~1
1L

0 5 tO
t0 5 20' 25 3 35 40 45 50 55 60 e65
NUMBEROF HORDES
FIG. 2. Extension and summary of trend in diagrams of figure 1.
For explanation see text, p. 199.

gration of Australian spatial relations in as in fact the evidence indicates. If it is

either geometry would result in a con- further assumed that the rate of interband
tinent in which there were no tribes and exchange of women, and of other forms
for which cultural change could best be of social contact, is the same on all of its
shown by a series of clines. Under such boundaries, whether these involve extra-
a system, with each band equivalent to tribal or intratribal bands, then an im-
the tribe and with patrilocal exogamic portant geometrical principle in the cul-
marriage uniformly practiced, all mar- tural relationships emerges. For the con-
riages would be intertribal in nature. If figuration involving four-sided bands the
frequency of social contacts followed the number of intertribal contacts or mar-
pattern of intertribal marriages as it is riages is reduced to 50 per cent of the
assumed here, then the spread of cultural total, whereas for the hexagonal shaped
changes would be a slow, gradual and bands the intertribal marriages are re-
uniform process and aboriginal cultural duced to 60 per cent.
space would be structured only at the Figure 1 carries the investigation one
band levels. step further in the right hand diagram,
In the middle diagrams of both rows it which involves 25 bands in each tribe for
has been assumed that on the average the square configuration and 19 bands in
tribes consist of 9 bands in a four-sided each tribe for the hexagonal pattern of
space and 7 bands in a hexagonal space. space. Utilizing the previous assumptions
Both of these concepts imply that clusters it is found that increasing the number of
of bands are the units of cultural cohesion bands per tribe has further reduced the

This content downloaded from 132.239.1.231 on Thu, 14 Jan 2016 16:37:42 UTC
All use subject to JSTOR Terms and Conditions
 HUNTING AND COLLECTING POPULATIONS 199

number of extra-tribal marriages and con- etry of the tribal boundaries. So that
tacts, even though no cultural barriers perhaps it can be considered for the
have been erected at tribal boundaries. present that these two factors tend to
For the 25 bands with the square shape, compensate for each other.
the extratribal marriages or contacts are Certain other points must be considered.
reduced to 3313 per cent of the total, and For example, if the sizes of tribal areas
for the 19 hexagonally shaped bands the are changing unduly rapidly, as may oc-
value is about 42 per cent. cur under abrupt changes in ecology, as
The trend established by these diagrams going from a desert to a riverine area,
has been carried further and summarized then the count will be distorted. In a
in figure 2. It will be seen that for both similar fashion coastal tribes will yield too
types of geometrical configurations an in- low a count in contact boundaries. For
crease in the number of bands per tribe, these reasons the basic series of tribes
which represents the culturally coherent published earlier (Birdsell, 1953) have
unit, is followed by a smoothly diminish- been utilized in the count, and 100 tribes
ing proportion of extra-tribal marriages from these data have been examined for
and presumably contacts. It will be noted the number of contacts along their bound-
that the hexagonal configuration shows a aries. The results indicate an average of
less abrupt reduction in such frequency 5.5 tribal contacts for each of the 100
of contact but that it follows the same gen- tribes. The results strongly suggest that
eral pattern of the simpler four-sided con- the hexagonal configurations in figure 1
figuration. This is a version of the simple more closely approximate empirical condi-
geometric principle that for similar figures tions in Australia (and not unlikely else-
the contained area increases more rapidly where) than does the simpler square con-
than does the bounding perimeter. figuration.
At this point it is of interest to em- Since the curves show that the geo-
pirically check the actual configuration of metrically determined number of extra-
Australian tribes as shown by Tindale's tribal marriages or contacts are a function
map (1940). There are no data giving of the number of bands per tribe, it is of
a detailed plotting of bands in a geo- some importance to try to estimate for
metrical sense so that it is necessary to Australia the average number of local
assume that their general form will ap- groups per tribe. A tabulation of the
proximate that of the tribe which con- band count for a small series of tribes
tains them. Thus an essentially circular derived from Tindale's study (1940) sug-
tribal area will be filled with bands which gests that when small corrections are made
are assumed to be roughly circular in for under-counting, that the Australian
their own shape; that is, not markedly tribe may average about 10 bands. As-
elongated in either major axis. On the suming the tribal population to actually
other hand, a tribal territory which is average 500 persons gives an average
narrow and long must be assumed to con- band population of 50. This value is ex-
tain bands with somewhat similar shapes. ceeded in some favorable areas, but it
This assumption cannot be tested and it certainly is considerably higher than the
may well err in the direction of sys- real figure in other lower density regions.
tematically giving a larger number of It can be used for this preliminary study
sides per band than in fact exists. This without too much distortion.
possible error, however, is in part cor- In both figures 1 and 2 it has been as-
rected for by ignoring in the count those sumed that tribal boundaries will inhibit
contacts which represent the apices of neither the interband exchange of women
tribes in contact. In practice the bands in or the frequency of interband social rela-
these apical territories certainly would tions and cultural contacts which are pre-
have contact despite the narrowing geom- sumed to follow the same pattern. For-

This content downloaded from 132.239.1.231 on Thu, 14 Jan 2016 16:37:42 UTC
All use subject to JSTOR Terms and Conditions
200 JOSEPH B. BIRDSELL

tunately the data exist for correcting this based relationships of the band. This
assumption. Tindale, to whom anthro- index, which measures the expected num-
pologists owe so much for detailed and ber of extra-tribal marriages, should fac-
quantitative knowledge of Australian abo- tors of cultural cohesion not disturb them,
riginal anthropology, published (1953) as compared to those which actually occur,
an exhaustive study of the rates of inter- might well be called the marital index of
tribal marriages among individuals prior cultural cohesion. Even though the aver-
to the distorting influence of white contact. age number of bands per Australian tribe
These data were based upon geneological is but scantly documented, it certainly is
materials and have been analyzed with contained within the range between 5 and
meticulous care. While there are some 15, and this corresponds to index values
slight regional variations in rate, the con- ranging roughly from 3 to 41/ times for
tinental average was 14 per cent inter- the cohesive influence. It is to be noted
tribal marriages of the total number re- that this index applies only to Steward's
corded. This figure not only provides the unilineal band structure, in which th-e
human microevolutionist with an invalu- local groups are both patrilineal and
able basis for estimating the rate of gene exogamous. Even so, it is applicable to
flow via migration but here gives an op- a large proportion of the peoples at this
portunity to estimate the intensity of the economic level of subsistence.
barriers against cultural exchanges which
in fact do exist at tribal boundaries. The MODEL FOR SOCIAL RATES OF BAND
existence of such barriers may be postu- INTERACTIONS
lated in an a priori sense from the very
existence of the clusters of homogeneous In the previous model, based upon the
bands known as tribes. Without an in- geometrical distribution of bands in space,
hibiting influence at the tribal boundaries the direction of the forces involved is evi-
these aggregations of local groups could dent. This geometry works so that an
not maintain their homogeneity but would increase in the number of bands in a tribe
in fact dissolve into an unstructured band further insulates it from extra-tribal cul-
space. tural influences and so would tend, work-
A mean figure of 14 per cent extra- ing by itself, to increase the number of
tribal marriage among precontact Aus- bands per tribe. In the reverse sense,
tralians allows an estimate to be made of tribes containing too few bands, unless
the cohesive factors which produced tribal isolated by topographical or ecological
entity. In terms of the idealized geometry features, tend to be exposed to an undue
of hordes of hexagonal configuration, a number of extra-tribal influences cul-
tribe with 10 hordes would be expected turally, and so would find it difficult to
to show about 52 per cent extra-tribal maintain a homogeneous core of cultural
marriages if the tribal boundaries exerted values. It may be postulated that below
no cultural inhibiting forces. The fact this threshold tribes would tend to dis-
that only 14 per cent do occur indicates appear through absorption by neighboring
that 3.7 times as many marriages might and more stable groups.
be expected without the interference of The force opposing this tendency to in-
these factors for internal cohesion. In crease the number of bands per tribe is
areas where tribes average 15 bands based in another aspect of the geometry
apiece, about 44 per cent of extra-tribal which involves distance. The influence
marriages would be expected, which gives of distance can perhaps be best examined
the value of 3.1 for this index. For tribes through models which synthesize the total
which, on the average, comprise of about social and cultural interactions of a band
5 local groups the index rises to 4.6, in the central position in a tribe in terms
owing to the nature of the geometrically of contiguous and non-contiguous bands

This content downloaded from 132.239.1.231 on Thu, 14 Jan 2016 16:37:42 UTC
All use subject to JSTOR Terms and Conditions
 HUNTING AND COLLECTING POPULATIONS 201

within the same tribe, and those extra- territories. Spaced about these contigu-
tribal bands which it may occasionally in- ous bands are 12 local groups once re-
teract. Again, it is unfortunate that field moved from the central one, and here
observers provided very little empirical the contacts will be more formal and less
data as to the nature and frequency of frequent. Thus in this configuration a
such contacts. For Australia, and these tribe is considered to consist of 19 bands
conditions presumably hold reasonably in all, a figure which is somewhat larger
elsewhere, a given band may be expected than that attained on the average among
to have most frequent contacts with those Australian tribes. Finally, outside of the
local groups which share its boundaries. tribal territory but contiguous to its
Less frequently it will meet with local boundary, are 18 bands which in this
groups of its own tribe but one band re- model belong to the 6 adjoining tribes.
moved from its territory, and upoln rare The central band will have infrequent
occasions it may be expected that it will contacts with these extra-tribal local
come into face-to-face contact with bands groups, and only upon such occasions has
from outside of its own tribal group. involved the performance of inter-tribal
Contacts between adjacent local groups no ceremonies. Thus, for model purposes
doubt frequently occur without the excuse it is considered that the central band of
of formal ceremonial occasions. But con- the tribe may come into contact with 36
tacts with more remote bands are pri- other bands, with varying rates of fre-
marily a consequence of invitations to quency.
participate in ceremonies of various sorts. The frequency of interaction between
This is not to say that other occasions, the central band and those once, twice
such as trading parties and revenge parties and thrice removed from it certainly occur
do not play a role in these contacts be- as an inverse function. Three sets of
tween bands, but it may be safely pre- assumptions have been utilized in figure 3,
sumed that they are less frequent and less based upon the following estimates of
important for the purposes of our model. frequency. In the first instance it is as-
Correspondingly they are ignored in these sumed that the central band will have in-
tentative versions. dividual contacts with each of the con-
The variables used in the construction tiguous bands once each year, that it will
of the following models fall under five have social contacts with the bands once
headings: (1) geometry of bands in space; removed on the average of every second
(2) frequency of interaction; (3) in- year, and it will contact extra-tribal bands
tensity of interaction; (4) duration of only every fourth year. Empirical data
interaction; and (5) facility of communi- on actual frequencies, as mentioned pre-
cation. Each of these factors will be dis- viously, are extremely scarce in the litera-
cussed separately below. ture, but they make it quite clear that this
The influence of the geometrical spatial is the most likely of the three assumptions
relations of bands has been explored pre- to approximate the true situation. The
viously in figures 1 and 2. For this second assumption involves more fre-
model the hexagonal configuration will quent contacts with more distant groups
be utilized since this seems to correspond in the following way: contacts with con-
more closely to actual conditions. In tiguous bands remain one per year for
such an arrangement of bands the one cen- each band; contacts with tribal bands once
trally located within the tribe, which is removed are increased to one each year;
here being considered, will have direct and extra-tribal band contacts are pre-
territorial contacts with the 6 contiguous sumed to occur every second year. This
bands about it. These contacts will be assumption seems less likely since it tends
frequent as the local groups go through to over-emphasize the frequency of social
the seasonal rounds within their own contacts with bands geographically re-

This content downloaded from 132.239.1.231 on Thu, 14 Jan 2016 16:37:42 UTC
All use subject to JSTOR Terms and Conditions
202 JOSEPH B. BIRDSELL

100 100

9 -O---0---O- INTERTRIBAL
CEREMONIES
ONCEIN 4 YEARS ( 90
LL __\_ .. .. U 2 /
Cn 80 - " " a I 80

70 / 70
0 *

0~~~~~~~~~~~~~~~~~~6

C~ /ETATIA)B(RBL (OTGOS

0 0~~~~~~~~~~~~~~
050 (XTATRBA)B/TIBL)A(CNIGOSO
z /YEOFETR-ODA ONAT/
FI.3 igamo eainsi ftp/adprcn /fetahra otcs
For xplaatio se tex, pp 201204
.
3 FIG 3.Darmo eainhi3ft0 n e etofetahra otcs

~ ~~~
F_~ exla0to ~o se ex,pp20124

mote. The third and least likely assump- which the 36 bands are assumed to be
tion considers that contacts will be estab- present, the intensity of interaction would
lished once each year with bands in all be 136 with each of the bands. This is
three of the previous categories. This is conservative in that field data do not sug-
not a realistic estimate and it is included gest that in intertribal ceremonies one-
only to provide an extreme limit for the half of the participants are extra-tribal in
phenomenon investigated. origin.
In these preliminary models the inten- The duration of interaction, while a
sity of interaction is defined simply as an theoretical component of the model, must
inverse function of the number of other be ignored in its actual construction. Im-
bands present. For the purposes of this portant ceremonial events occasionally are
construct, it is assumed that contiguous recorded as continuing through 3 or 4
bands are in effect all seen together, and months, but there is no evidence to indi-
that the rate of intensity of interaction for cate how frequently such ceremonies were
any given band is %. Where the whole held. In a general way it is clear that at
tribe gathers, an event probably rare in a generalized hunting and collecting level
actuality, there are 18 bands present, and of economy such an event must be very
it assumes that the central band interacts rare for the food resources available to
with the intensity of V18 with each of support such a large gathering for such a
these. For an intertribal ceremony in protracted period of time would limit the

This content downloaded from 132.239.1.231 on Thu, 14 Jan 2016 16:37:42 UTC
All use subject to JSTOR Terms and Conditions
 HUNTING AND COLLECTING POPULATIONS 203

number of possible occasions. This fac- earlier in this section. Thus (fa) includes
tor is omitted, but with the understanding ga, fa, ia, da, ca, which represent respec-
that it would tend to be inversely related tively the geometry, frequency of inter-
to the factor of frequency and so to level action, intensity of interaction, duration
out the influence of that variable. of interaction, and facility of communica-
Facility of communication is a measure tion. Thus the equation fully expanded
of linguistic similarity or difference. There stands as follows:
seem to be no data to indicate proper
values for this factor, but since tribes owe K = ga fa ia da Ca A + gb fb ib db Cb B
their existence to linguistic and cultural + gc f, i d, ec C
homogeneity, the differences between the The substitution of empirical values in
bands of the tribe cannot be great. For the equation requires a further explana-
the purposes of the present model, using tion. Contacts with contiguous bands oc-
unity as a measure of the linguistic facility cur locally, at tribal gatherings and at the
of communication within a band, it is great intertribal ceremonies. Interactions
considered that this facility might dimin- with local groups once removed occur
ish to .9 for communication with adjacent only in the latter two occasions. Extra-
bands. For bands once removed the fact tribal band contacts require intertribal
has been reduced to .7 and for extra-tribal ceremonies to come into being. Using the
bands, those twice removed, the factor is first assumption, which concludes with
reduced to .5. The last factor is based extra-tribal interactions occurring only
upon the common native description of the once in four years, the equation becomes:
language of adjacent tribes that "I can
hear him but I can't talk to him." It is K = 89.05 per cent A+ 10.66 per cent B
probable that the first two values should +0.29 per cent C
be somewhat higher, but since this factor
has relatively little influence upon the Using the second assumption in which
curve derived from the total series of frequency of extra-tribal interaction is in-
factors, the values can stand at these creased to every second year the equation
levels. stands as:
If it is assumed that the total social in- K = 80.23 per cent A+ 19.20 per cent B
teractions of one horde with all other
hordes with whom it makes occasional +0.57 per cent C
contacts remains essentially constant when Using the third and most extreme as-
assayed over long periods of years, then sumption in which intertribal ceremonies
it is possible to establish a measure of are presumed held every year with regard
the relative frequency of contacts of the to the bands in question the equation
three different kinds discussed above. reaches the values of:
The equation in its simplest form is:
K = 73.13 per cent A+ 17.50 per cent B
K = (fa)A + (fb)B + (fc)C +9.37 per cent C
The constant A designates the proportion These three calculationshave been graphed
of interactions with immediately contigu- in figure 3. Since type A and B interac-
ous bands, the constant B shows interac- tions combined represent intratribal con-
tions with other bands in the tribe but tacts and only type C represents extra-
once removed from the central band, and tribal social interaction on a relative scale
C represents the relative rate of social in- of 100 per cent, it is highly significant
teraction with local groups outside of the that under the most likely set of assump-
tribe. The variables f are composite and tions, that is the first type, these extra-
each includes the five factors discussed tribal contacts amount to 0.3 per cent.

This content downloaded from 132.239.1.231 on Thu, 14 Jan 2016 16:37:42 UTC
All use subject to JSTOR Terms and Conditions
204 JOSEPH B. BIRDSELL

Even under the second set of conditions it might prove possible to express the
they only rise to 0.6 per cent. The totally two forces in terms of numbers of bands
unrealistic assumptions of the third type and equate them for the equilibrium con-
yield a value of less than 10 per cent ditions, and so calculate the numbers of
-extra-tribal interactions. Even without bands expected under the simplified two
the detailed empirical data of the kind force system for internal cohesion.
that would allow the equation to be cal-
CONCLUSIONS
culated accurately, it becomes evident that
without introducing inhibiting factors of Despite the appalling lack of detailed
tribal boundaries to reduce the interaction quantitative data upon which generaliza-
rates with extra-tribal bands, that distance tions can safely be based, it is to be
itself reduces the rate to almost unim- concluded that generalized hunting and
portant values. A proper correction for collecting populations, as defined in this
the actual inhibiting factors that do exist paper, show further regularities than
at tribal boundaries would even further re- those previously observed in anthropo-
duce the relative rates of social and cul- logical literature. Under the general
tural intercourse with extra-tribal groups. theme of territoriality two detailed stud-
Thus the model for rates of band inter- ies, one in Australia and the other among
action, which are based upon a geometri- the Great Basin Shoshoni indicate that
cal function of distance, demonstrates the human densities at this economic level are
nature if not the actual values of the force primarily determined by environmental
which acts in primary opposition to that variables which act in rigorous fashion.
previously discussed under the model on Ecological considerations for other such
geometry of band relationships. It would populations suggest that though the vari-
be an oversimplification to claim that ables may become more complex, they
the tendency for tribes among generalized operate just as stringently. The corol-
hunting and collecting peoples to approach lary of this finding is that populations of
a stable population number is the result hunting and collecting peoples are essen-
of these two forces alone. Certainly other tially stable (but not to the extent of
kinds of factors operate to keep the equi- barring minor oscillations) and are in
librium stable. But it is perhaps not equilibrium with the carrying capacity of
claiming too much to say that the two dis- the country, as determined by the extrac-
cussed here are perhaps of primary im- tive efficiency of the population in obtain-
portance in the network of forces which ing energy.
maintain the general equilibrium values.2 Under the subject of population dy-
If refined empirical data were available, namics it would appear that such eco-
nomically simple populations may be
2 The regions of tribal fragmentation in expected to show an intrinsic rate of in-
Australia which seem primarily associated with crease which involves doubling their
the recent acceptance of the iniatory rites of
circumcision and subincision, and secondarily
numbers in each generation, when placed
with low reliability of rainfall (Birdsell, 1953), in an unlimited environment. Considera-
may ultimately be explained in terms of these tion of the prolonged period of nursing
two vectors. If these rites are diffused (by and generally high infant mortality lead
formal and positive introduction) to portions of to the conclusions that the effective num-
tribes, then the phenomenon of transient frag-
mentation can be interpreted as a consequence ber of offspring of such populations
of a disturbing force which is corrected with shows a low and essentially constant vari-
the passage of time, by the ever present major ance so that the equation which relates
forces producing stability in the size of the effective size of the population to the size
tribal population. The general pattern of the
of the breeding population is simplified
spatial distribution of fragmentation corresponds
to an age-area theory based upon the above into a direct relationship between these
considerations. two values.

This content downloaded from 132.239.1.231 on Thu, 14 Jan 2016 16:37:42 UTC
All use subject to JSTOR Terms and Conditions
 HUNTING AND COLLECTING POPULATIONS 205

A preliminary exploration of some of 1952-53 has been supported by a John

the factors making for social cohesion Simon Guggenheim Memorial Founda-
indicates that the stability observed in the tion grant, a Fulbright Fellowship, the
size of tribal populations among such Wenner-Gren Foundation Grant for
peoples may be primarily related to two Anthropological Research, Inc., and the
opposing forces which have their basis in Committee on Research, University of
spatial geometry. A simple model based California at Los Angeles. I acknowl-
upon a hexagonal configuration for band edge my gratitude in receiving these insti-
geometry shows that tribal cohesiveness tutional sources of support. The success
is increased as the number of bands in- of the field work of both expeditions was
crease. At the same time it is possible in a large degree due to the collaboration
to measure for Australia the cultural in- of Mr. N. B. Tindale, ethnologist, South
hibiting forces exerted by tribal bound- Australian Museum, Adelaide. In an
aries in terms of extra-tribal marriages. age in which anthropologists have tended
On the assumption that Australian tribes to forego the effort of providing quantita-
approximate 10 bands on the average, it tive data, Tindale has made the greatest
was found that 3.7 times as many extra- permanent contribution to the anthro-
tribal marriages should occur according pology of the Australian aborigines by
to geometry as do occur, if no barriers to providing a wealth of such material.
them existed at the tribal boundaries. In
another type of model, which represents REFERENCES
a crude assay of the character of rates of BARTHOLOMEW, GEORGE A., JR., AND JOSEPHB.
interaction between bands it is possible BIRDSELL. 1953. Ecology and the proto-
to show that under any set of reasonable hominids. American Anthropologist, Vol.
55, No. 4, pp. 481-498.
assumptions a centrally located band in a BIRDSELL, JOSEPHB. 1953. Some environmen-
tribe has a negligible portion of its total tal and cultural factors influencing the struc-
interaction with extra-tribal local groups. turing of Australian aboriginal populations.
This model demonstrates the great im- American Naturalist, Vol. 87, no. 834 (Supp.
to May-June, 1953), pp. 171-207.
portance of distance in opposing the [In press.] Some population problems
enlargement of the number of bands with- involving Pleistocene man. Cold Spring
in a tribe. It is hoped that with more Harbor Symposia on Quantitative Biology,
empirical data it will be possible in the Vol. 22.
STEWARD,JULIAN H. 1938. Basin-Plateau
future to develop series of quantitative
aboriginal sociopolitical groups. Smithson-
studies which not only more accurately ian Institution, Bureau of American Eth-
define the stabilizing factors discussed in nology, Bulletin 120, pp. 1-346.
this paper, but perhaps to explore other TINDALE, NORMAN B. 1940. Distribution of
less important ones which certainly exist. Australian aboriginal tribes: a field survey.
Trans. Roy. Soc. So. Australia, Vol. 64, pp.
140-231.
ACKNOWLEDGMENT 1953. Tribal and intertribal marriage
among the Australian aborigines. Human
This paper is one of a series resulting Biology, Vol. 25, No. 3, pp. 169-190.
from the Harvard-Adelaide anthropo- VORKAPICH,MYRA. [in manuscript]. Ecology
logical expedition of 1938-39 which was as a factor in Great Basin aboriginal popula-
tion density.
generously sponsored by the Carnegie WRIGHT, SEWALL. 1938. Size of population
Corporation. Subsequent research, in- and breeding structure in relation to evolu-
cluding two years of further field work in tion. Science, Vol. 87, pp. 430-431.

This content downloaded from 132.239.1.231 on Thu, 14 Jan 2016 16:37:42 UTC
All use subject to JSTOR Terms and Conditions