CTLF Corpus de textes linguistiques fondamentaux • IMPRIMER • RETOUR ÉCRAN
CTLF - Menu général - Textes

Boas, Franz. Race, Language and Culture – T30


Evolution or Diffusion? 11

In a paper on Tewa kin, clan, and moiety by Elsie Clews Parsons 22 and
another on the social organizations of the tribes of the North Pacific
Coast by myself 33 the distribution of clans and related social phenomena
in two regions has been discussed. The inference must be drawn that in
geographically extreme areas in these districts distinctive types of social
organization occur, the intermediate regions showing transitional types.

This phenomenon is by no means confined to these regions or to
social organization, but may be observed to a greater or less extent in
all other cultural phenomena and in other parts of the world. The
component elements of folktales common to two areas decrease in number
the greater the distance, and while in intermediate regions we may
find much that reminds us of the extreme types, that are being compared,
the extremes themselves may be fundamentally distinct. This
condition may be observed in the folklore of the North Pacific Coast
when comparing Alaskan tribes with those of Oregon, or the Coast
tribes with those of the interior, or when comparing the folklore of the
Plateau tribes with that of the Pueblos. The same condition may be observed
also in material culture and is found when we compare the tribes
of the Plateaus with those of the Plains, or the Eskimo and the Northwest
Coast tribes. It may be seen in the distribution of art styles. All this
does not preclude the possibility of a unified stylistic pattern originating
in the intermediate areas, and it does not imply necessarily a greater
purity of the extreme, and a more mixed character of the intermediate

It does, however, prove, in our opinion, that all special cultural forms
are the products of historical growth, and that unless considerations
entirely foreign to the observed distribution are introduced, no proof
can be given that one of the extreme forms is more ancient than the

If we adopt the theory that matrilineal clans must be older than
patrilineal or bilateral organization, we might be tempted to say that
in the southern part of British Columbia and the eastern Pueblo district
the clan organization has broken down, the more so the farther we move
away from the centers in which this type of organization is still flourishing.
The distribution itself does not lead to such an assumption. On the
contrary, we see merely the intermingling of two distinctive types, the
combination of which leads to new forms and new ideas.

The importance of diffusion has been so firmly established by the
investigation of American material culture, ceremonies, art and mythology,
as well as by the study of African cultural forms and by that
of the prehistory of Europe, that we cannot deny its existence in the
development of any local cultural type. It has not only been proved
objectively by comparative studies, but the field student has also ample
evidence showing the ways in which diffusion works. We know of cases
in which a single individual has introduced a whole set of important
myths. As an instance we might mention the tale of the origin of the
Raven which is found in one single tribe on the northern part of Vancouver
Island. It is still known to a few individuals that this tale was
introduced by a man who had for many years been a slave in Alaska,
and who was ultimately ransomed by his friends. Nevertheless, the
myth is regularly told as part of the Raven cycle, although it is repudiated
by all the neighboring tribes. Another example is the introduction
of the Badger clan in Laguna by a Zuñi woman. Her
husband, also from Zuñi, introduced to Laguna Zuñi kachina rituals
and Zuñi stories which are now flourishing in their new environment. In
earlier times the carrying away of women after raids, adoptions of
foreigners, and other similar phenomena must have been a fruitful
source of introduction of foreign ideas, the more so the smaller in
numbers the tribe, and the more efficacious the influence of a single
person. The introduction of new ideas must by no means be considered
as resulting purely mechanically in additions to the cultural pattern,
but also as an important stimulus to new inner developments.

A purely inductive study of ethnic phenomena leads to the conclusion
that mixed cultural types that are geographically or historically intermediate
between two extremes, give evidence of diffusion.

The question then arises as to how the extreme and most divergent
forms must be considered. In our particular examples, the North
Pacific clan organization with a small number of clans and many local
291groups possessing definite privileges must be compared with the bilateral
organization of the south with numerous independent local units practically
without privileges. In the Southwest, the matrilineal clan organization
of the western Pueblos, almost entirely devoid of moieties,
must be compared with the paternal moieties of the east without

If it can be shown inductively that one of these types is the older one
and that there are inherent dynamic conditions that tend to bring about
transition from the older condition to the newer one, and that these
conditions work in such a way that their potency decreases from the
center to the periphery, the theory of a uniform development might be
maintained. We require, therefore, in this case proof of three historical
conditions: First, proof that one type is older than the other; second,
that the younger type develops necessarily from the older one — in other
words, that the dynamic conditions for a change in this direction are
ever present; and thirdly that these conditions act with increasing intensity
from the periphery towards the center.

As against these hypotheses the theory of diffusion takes the two
distinctive types as given, and accepts as proven the presence of

It should be borne in mind that the assumption of the antiquity of
one particular type is essentially due to a classification in which the form
that appears as the simplest from any one point of view is considered at
the same time as historically the oldest. Nobody has felt the weakness
of this assumption more clearly than Tylor who tried to support the
general thesis by the study of survivals which indicate the character of
earlier developmental stages. It cannot be claimed that a systematic
attempt has ever been made to substantiate the theory of a definite
evolutionary sequence on the basis of the study of survivals. All that
can be said is that fragments of earlier historical stages are bound to
exist and are found. We can, perhaps, best illustrate this by the example
of matrilineal institutions. Whenever these are connected with
the holding of social prerogatives in the hands of men, and where, nevertheless,
the family in our sense is an important social feature, there is a
constant cause of conflict because the matrilineal descent requires that
property or position must pass out of the family into another family
group. This entails an element of weakness, because the allegiance of
the individual is divided between two conflicting groups. It is, therefore,
plausible, that, in this case, matrilineal society contains elements
292of instability, and may, owing to inner dynamic conditions, develop
into a patrilineal or bilateral system. Then we may find examples of
the survival of matrilineal forms in patrilineal society. This, however,
does not by any means prove that everywhere matrilineal society must
have been the earlier form. It merely proves the instability of matrilineal
society of a certain type.

To us the assumption of a unique form of cultural beginnings does
not seem plausible. Setting aside the question of what form of social
life may have existed at the time when our ancestors first developed
speech and the use of tools, we find everywhere phenomena that point
to very early differentiations from which even the simplest cultural
forms developed. Language and art are perhaps the best proof of this
contention. Even if we should accept with Trombetti the unity of the
origin of human speech, or with Marty, the conscious invention of language
for the purpose of communication, we must concede that in the
early development of language fundamental categories of grammar and
lexicography have arisen that cannot be reduced to common principles,
excepting those general forms that are determined logically or by the
fact that language is a means of communication. The same is true in
regard to stylistic forms of art which cannot be reduced to a single
source. What is true of language and art, which do not become a subject
of retrospective reasoning, seems to us no less true of those aspects
of life which are subject to remodeling by rationalizing processes. To
this class belong the forms of social organization. The theory of the
priority of maternal organization implies necessarily that the original
economic and social unit consisted of a first generation of mothers and
their brothers and of a second generation of children, and that the
fathers of the children and the grandchildren were only temporary
visitors to the family unit. It implies, therefore, a cohesion of this group
long after the children had become independent adults, and a group
consciousness in which no relations between father and children existed.
The continued cohesion between mother and adult children is, to say
the least, doubtful. According to the usual division of labor, such an
organization rigidly carried through in a sparsely occupied territory and
among a tribe dependent on hunting, would have doomed to extinction
all groups without brothers and adult sons. While groups of this type
may result from nonmarital sexual relations, we do not know of any
cases where relations between men and women remain temporary
throughout life, but marital relations continuing over a more or less
293extended period are the norm, and the social group includes the father. 14
It is, therefore, to us equally likely that primary units existed which consisted
of families in our sense, and that adult children separated from
the original groups and formed new family groups. Unless it can be
proved that in an overwhelming number of cases the bilateral family
retains evidences of a prior maternal stage, we have no right to assume
that all the ancient types of groups of kin would conform to the same
pattern, without any regard to the economic and other conditions that
determine the size and character of the social unit.

It seems to us that the uniformity of early patterns cannot be proved.
By analogy to the phenomena recently mentioned, we may rather infer
diversity of early patterns.

We believe, therefore, that the great mass of observed facts bears out
the theory that in the regions under consideration two fundamentally
distinct forms came into contact, that the one is not derived from the
other, but that through the mingling of the two forms new types arose
in the intermediate districts.294

11 American Anthropologist, N.S., vol. 26 (1924), pp. 340-344.

22 Ibid., pp. 333-339.

33 Ibid., pp. 323-332; pp. 370 et seq. of this volume.

41 For a full discussion of this matter see R. H. Lowie, Primitive Society (New
York, 1920), pp. 63 et seq.