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

Boas, Franz. Race, Language and Culture – T31


Review of Graebner,
“Methode der Ethnologie” 11

Mr. Graebner is one of the serious and broad-minded students
who are not satisfied with an accumulation of facts, but who are
carrying through their own investigations according to a well-considered
plan, and who try to contribute to science in a certain well-defined line
of research and look for results that have a definite bearing upon the
whole field of their inquiries. In the present book Mr. Graebner gives
us a statement of the method that he is following and which will interest
all ethnologists. If, however, Mr. Graebner calls his method the
method of ethnology, we cannot agree with him. He must not expect
that all ethnologists will limit the field of their researches in the way set
forth in these “Methods.” It appears from Mr. Foy's, the editor's, preface,
that in this respect his own views and Graebner's coincide; in fact,
in outlining the program of the whole series, Mr. Foy excludes expressly
“alle geschichtsphilosophischen und volkerpsychologischen Betrachtungen”
(p. V). This exclusion of the psychological field seems to me
to give to the whole “Method” a mechanical character, and to be the
essential cause of differences of opinion between the author and myself
which I shall briefly characterize in the following pages.

The book is divided into three chapters: critique of sources, interpretation
of data and combination of data. I do not quite share Mr. Graebner's
unfavorable view in regard to the lack of critique of all writers on
ethnological subjects, and in regard to the feeling that we are confronted
by an appalling lack of all method; a feeling that, according to the
author, the historian experiences who takes up the study of ethnology.
It is true that much that has been written is based on inadequate evidence,
and that particularly the so-called “comparative” ethnologists
do not weigh their evidence well. Spencer, Frazer and Westermarck,
not to mention others, have been criticized again and again by experts
from this point of view. However, the whole modern method of ethnology,
at least as developed in the United States, is a continuous struggle
295for gaining a critical viewpoint in regard to data collected by earlier
authors who did not understand the objects and problems of modern
anthropology. We believe that a safe interpretation of the older observed
data must be based on careful archeological, ethnological and
somatological field work. While I see a perfectly sound tendency in
these studies, sounder than Mr. Graebner believes it to be, I still recognize
the usefulness of the first chapter in which the author expresses the
experiences of the historian in a form interesting and important to the
unexperienced ethnologist. On the whole, the training given nowadays
to students in universities and museums will impress upon them the safeguards
on which the author insists, and which are too often forgotten
by the amateur.

Our interest centers in the following two chapters: Interpretation and
Combination of Data. The fundamental difference of opinion between
the author and myself appears in the chapter on Interpretation. He
defines interpretation as the determination of the purpose, meaning and
significance of ethnic phenomena (p. 55); but he does not devote a
single word to the question how these are to be discovered. He accepts,
without any attempt at a methodical investigation, myths as interpretations
of celestial phenomena (pp. 56, 57), as, for instance, the Jona
theme as signifying the temporary disappearance of a heavenly body;
a conclusion which I for one am not by any means ready to accept. At
this place the complete omission of all psychological considerations
makes itself keenly felt. The significance of an ethnic phenomenon is
not by any means identical with its distribution in space and time, and
with its more or less regular associations with other ethnic phenomena.
Its historical source may perhaps be determined by geographic-historical
considerations, but its gradual development and ethnic significance
in a psychological sense, as it occurs in each area, must be
studied by means of psychological investigations in which the different
interpretations and attitudes of the people themselves toward the phenomenon
present the principal material. In the case of mythology, by
means of which Mr. Graebner exemplifies his considerations, I should
demand first of all an investigation of the question: why, and in how
far are tales explanatory or related to ritualistic forms? The very existence
of these questions and the possibility of approaching them has been
entirely overlooked by the author. On the whole, he seems to assume
that the psychological interpretation is self-evident in most cases, but
that by migrations and by dissemination combinations may be brought
296about which may lead to misinterpretations in so far as several groups
that were originally distinct may be considered as one by origin (p. 64).

Related to this disregard of the psychological problem is Mr. Graebner's
claim, that no objective criteria have been found that can prove
relations other than those due to historical connection; that the evolutionary
investigation can do no more than answer the question: “How
can I best and with the least number of contradictions imagine the
course of human development in accordance with my general, fundamental
views?” (p. 82). Against this method he claims that transfer
has been proved to exist everywhere, while the presence of parallel
development cannot be proved by objective criteria (p. 107). I think,
we must say, that certain types of changes due to internal forces have
been observed everywhere, and that, therefore, the question of similar
or dissimilar evolution through internal forces does not rest on a more
hypothetical basis than changes due to transmission.

Another fundamental difference of opinion between Graebner and
myself relates to the phenomenon of “convergence,” and here again
the conclusions reached by the author seem to me due to a narrow,
mechanical definition of the term “convergence.” He ascribes this idea
to Thilenius and Ehrenreich. I may, perhaps, point out that I have
raised the essential point in an essay “The Limitations of the Comparative
Method of Anthropology,” 12 and again in my essay “The Mind of
Primitive Man.” 23 Graebner's first error in regard to this phenomenon
is one which he shares with almost all other students of anthropogeography.
I quote from p. 94: “Gleichartige Erscheinungen konnen auch
durch Angleichung urspriinglich verschiedener Erscheinungen unter
dem Einfluss gleicher Natur- oder Kulturumgebung zustande kommen.
Da eine spezifisch gleiche Kulturumgebung ausser durch Kulturverwandtschaft
aber ihrerseits nur als durch gleiche Naturumgebung
hervorgerufen denkbar ist, bleibt diese allein als primare Ursache von
Konvergenzen übrig.” This presupposes an existence of a mankind
without any individual differences, or an absolute identity of the psychical
conditions that are affected by geographical environment. As soon
as the cultural basis is distinct, even the most absolute identity of environment
cannot be assumed to lead to the same result. It is a curious
view that is so often held, that when we speak of the influence of environment
upon the human mind, only the enrivonment need be considered.
297Is not in every problem of interaction the character of each of the
interacting phenomena of equal importance? In the particular case here
discussed we may say that our whole experience does not exhibit a single
case in which two distinct tribal groups are so much alike in their mental
characteristics that, when they are subjected to the same modifying
causes, these mental differences could be disregarded, and it is an entirely
hypothetical and improbable assumption that in earlier periods absolute
mental uniformity as expressed in culture ever existed in distinct groups.

The idea that in cases of independent origin of the same cultural
phenomena identity of environment can give the only satisfactory explanation
is deeply rooted in Mr. Graebner's mind, for he repeats, on
p. 112: “Gleiche Kulturbedingungen bei selbstandiger Entstehung
konnen ihrerseits wieder nur auf die Naturbedingungen zurückgehen.”

The phenomenon of convergence is next considered as non-existent
for two reasons: a theoretical one and an empirical one. The former
is based on the consideration that convergence can occur only under
identical cultural conditions, and that, therefore, heterogeneous cultural
conditions such as are found in cultures not genetically related, cannot
possibly lead to the same result. The empirical argument is based on a
consideration of conditions found in Europe (pp. 113-114). A consideration
of the same data leads me to results diametrically opposed to
those observed by Graebner. The very fact that in modern civilization
a new idea is frequently discovered independently by several individuals
seems to me a proof of parallel lines of thought; and Mr. Graebner's
statement that the thought of only one man becomes socially active, i.e.,
is adopted, seems to me to demonstrate just the reverse of what he
claims. For an idea expressed at a time that is not ready for it remains
barren of results; pronounced at a period when many think on similar,
convergent lines, it is fruitful and may revolutionize human thought.
May I point out that Graebner's own book may be taken as an example
of this tendency? For it expresses the same fundamental idea that is
so potent at present in all lines of biological research, that of the permanence
of unit characters. An idea may become effective whenever
the ethnic conditions are favorable to its adoption and development, no
matter what the historical origin of the present general status may have

The questions of independent origin and convergence cannot be
entirely separated, and some of the previous remarks may perhaps
rather relate to the probability of independent origin which Graebner
298practically denies. One aspect of the theory of convergence relates more
specifically to the question whether two ethnic groups that are genetically
distinct, which are confronted by the same problem, will solve it in a
similar manner. The theory of convergence claims that similar ways
may (not must) be found. This would be a truism, if there existed
only one way of solving this problem, and convergence is obviously the
more probable the fewer the possible solutions of the problem. This,
however, is not what we ordinarily understand by convergence.
Ethnic phenomena are, on the whole, exceedingly complex, and apparently
similar ones may embrace quite distinct complexes of ideas and
may be due to distinct causes. To take a definite example: Taboos may
be arbitrarily forbidden actions; they may be actions that are not performed
because they are not customary, or those that are not performed
because associated with religious or other concepts. Thus a trail may
be forbidden because the owner does not allow trespassing, or it may
have a sacred character, or it may be feared. All ethnic units, separated
from their cultural setting, are artificial units, and we always omit in
our comparisons certain groups of distinctive characteristics — no matter
whether the comparisons are made from the point of view of cultural
transmission, or of evolutionary series. Thus, in our case, the forbidden
action stands out clearly as a unit, that of the taboo, although its psychological
sources are entirely distinct — and this is one of the essential
features of convergence. Nobody claims that convergence means an
absolute identity of phenomena derived from heterogeneous sources;
but we think we have ample proof to show that the most diverse ethnic
phenomena, when subject to similar psychical conditions, or when referring
to similar activities, will give similar results (not equal results),
which we group naturally under the same category when viewed not
from an historical standpoint, but from that of psychology, technology or
other similar standpoints. The problem of convergence lies in the correct
interpretation of the significance of ethnic phenomena that are
apparently identical, but in many respects distinct; and also in the
tendency of distinct phenomena to become psychologically similar, due
to the shifting of some of their concomitant elements — as when the
reason for a taboo shifts from the ground of religious avoidance to that
of mere custom.

In the foregoing remarks I have tried to show why Mr. Graebner's
negative critique of parallelism and convergence does not seem to me
conclusive. Just as little convincing appear to me the arguments on
299which he bases his method of determining cultural relationships. Here,
also, the fundamental error seems to me based on the complete disregard
of mental phenomena. Mr. Graebner lays down the following
methodological principle: “Two or more phenomena are comparable,
and the one may be used to interpret the other, if it can be shown that
they belong, if not to the same local cultural complex, at least to the
same cultural group” (p. 64). It seems to me an entirely arbitrary
hypothesis to assume a priori the homogeneity of similar phenomena
belonging to the same cultural group. Mr. Graebner explains his standpoint
by the example of the discussion of agricultural rites in Frazer's
“Golden Bough,” and accepts the discussion on account of the homogeneity
of the cultural groups of Europe and western Asia, from which
the examples have been taken. This part of Frazer's deductions seems
to me just as unmethodical as the others which are based on examples
taken from a wider series of cultural groups. The concepts of comparability
and homogeneity, as I understand them, have to deal not only
with historical relationship, but to a much higher degree with psychological
similarity, for only as elements of the mental make-up of society
do ideas or actions become potent and determining elements of further
development. To give an instance of what I mean: If the aged are
killed by one people for economic reasons, by another to insure them a
happy future life, then the two customs are not comparable, even if
they should have their origin in the same historical sources. Graebner's
idea appears clearly in the following statement: “If in different parts
of the earth peoples are found that are closely related in their ways of
thinking and feeling, evidently the same question arises, that has been
treated before in regard to cultural forms, viz., whether these similarities
are not based on community of descent or on early cultural contact”
(p. 112). Such a view can be maintained only if we disregard the
action of inner forces, that may lead two people of like cultural possessions
after their separation to entirely distinct conditions. In short it
is based on the view of a very limited action of internal forces.

Through the restriction of comparability and interpretation exclusively
to the phenomena of transmission and original unity — a definition
that I do not find given, but that is everywhere implied — and by the
hypothesis, that ethnic phenomena that occur in two areas due to transmission
or to original unity will always remain comparable and can be
mutually interpreted, the author is necessarily led to his conclusions,
which are merely a restatement of his incomplete definitions and of his
300hypothesis; for, if we call comparable exclusively phenomena that are
historically related, naturally then there can be no other kind of comparability,
and psychological ethnology does not exist.

Exactly the same criticism must be made against the sense in which
the term “causal connection” is used. Here also the psychological connections
are intentionally excluded, because the psychological argument,
its method and validity, are not congenial to the author; and
“causal connection” is simply identified with historical connection. On
this basis only can I understand the statement that in literary tradition
causal relations are directly given (p. 73). This is not meant to refer
to modern historical science, but to the literary sources of Asia and
Europe. Is not literary tradition on the whole proof of the misunderstanding
of causal relations, rather than the reverse — provided we understand
under causal nexus not the simple mechanical aspect of transmission,
but the complex social conditions that admit transmission and
that bring about internal changes.

A correlate of the assumption that ethnic elements that are genetically
related remain always comparable plays a most important part in Mr.
Graebner's method of proving cultural relations: “Whenever a phenomenon
appears as an inorganic element in its ethnic surroundings, its
presence is due to transmission.” This might be true if primitive cultures
were homogeneous units; which, however, is not the case. The more we
learn of primitive culture, the clearer it becomes that not only is the participation
of each individual in the culture of his tribe of an individual
character, or determined by the social grouping of the tribe, but that
also in the same mind the most heterogeneous complexes of habits,
thoughts and actions may lie side by side, without ever coming into
conflict. The opinion expressed by Mr. Graebner seems to me so little
true, that I rather incline toward the reverse opinion. It seems at least
plausible, although it has never been proved, that on the whole only
such ethnic features are transmitted that in some way conform to the
character of some feature of the life of the people that adopt them. The
criterion in question seems to me, therefore, not acceptable, until it can
be sustained by observed facts.

This idea is probably related to the author's conception of the transmission
of cultural elements in the form of complexes. He says: “A
migration of single cultural elements, also of tales, over wide distances,
without the spread of other cultural possessions at the same time, may
be designated without hesitation as a ‘kulturgeschichtliches’ nonsense”
301(p. 116). I should like to see the proof of this daring proposition. It is,
of course, not the question whether one cultural group owes much or
little to another one, but whether cultural elements are necessarily transmitted
in groups. To take only a few examples. Is not the gradual introduction
of cultivated plants and domesticated animals a case in kind?
Does not the irregular distribution of tales show that they are carried
from tribe to tribe without relation to other transmissions? It seems to
me that the more the problem of cultural contact is studied, the more
amazing becomes the independence of far-reaching influences in one
respect, from the spread of other cultural possessions. The example of
language used by Mr. Graebner (p. in) presents facts entirely different
from those which he imagines. Thus we find phonetic influences
without corresponding lexical or morphological influences and vice
. The serious defect of the “Method” is here clearly seen. Instead
of operating with the purely mechanical concepts of transmission and
conservatism relating to the most ancient types of culture, we must investigate
the innumerable cases of transmission that happen under our
very eyes and try to understand how transmission is brought about and
what are the conditions that favor the grouping of certain new elements
in an older culture.

I think I have shown that not only the psychological and evolutionary
standpoints contain hypothetical elements that must be subject to a rigid
criticism, but that the restriction of all ethnic happenings to mechanical
transmission or preservation contains many hypotheses the validity of
which is open to most serious doubt. Mr. Graebner has failed in his
attempt, because he does not apply the same rigorous standard to his
own favorite views, that he applies so successfully to a discussion of the
evolutionary theory (pp. 77 et seq.) Here he is at his best, and his
criticism of the many hypothetical assumptions contained in all theories
of the evolution of culture are well taken and should be read and minded
by all students of ethnology. In a few cases, particularly in the discussion
of correlated ethnic phenomena, he does not seem to do quite
justice to the force of the argument, because he prefers spatial interpretation
of these correlations to a sequential one; but both are certainly
equally possible and probable.

It is, however, curious to note that, notwithstanding his uncompromising
negative position, the author tacitly re-introduces some of the
most fundamental concepts of cultural evolution. Thus he speaks on p.
63 of the “well-known tendency of degeneration and disintegration, according
302to which myths become legends and fairy-tales, significant institutions
formal traits”; and again on p. 152: “Undoubtedly sound
points of view are, that the beginnings of every phenomenon must be
simple and in a way grow naturally, and that the development must be
intelligible by the most simple psychological process.” My criticism of
these assumptions would be much more far-reaching than that of Mr.

Thus it seems to me that the methods of Mr. Graebner are subject to
the same strictures as those of the other schools, and the “Ferninterpretation,”
“Kulturkreise” and “Kulturschichten” must be considered as no
less hypothetical than the “Stufenbau” of Breysig or the sequences of

In the development of science it is, however, useful to carry through
a hypothesis to its limits and to investigate the ultimate conclusions to
which it will lead. From this point of view pages 104 — 151, in which
the principle of conservatism and transmission are strained to the utmost
with an absolute disregard of all other possibilities, will be helpful for
a gradual clearing of our views. Perhaps even more helpful is the actual
application that Mr. Graebner has made of these principles in his
chosen field of Melanesia in its relations to the whole rest of the world.

My own opinions in regard to the value of a single evolutionary series,
the importance of very old cultural elements that survive in many parts
of the world, and the occurrence of transmission over enormous areas
coincide to a great extent with those of Mr. Graebner. I also hold the
opinion that the discovery of a really new idea is much more difficult
than is generally admitted, and therefore a manifold spontaneous origin
quite unlikely. Nevertheless, I cannot acknowledge that he has given
us any safe criterion that would enable us to tell that in any given case
transmission can be definitely proved against independent origin, and
I am just as skeptical as before reading his book in regard to the advisability
of accepting Ratzel's “Ferninterpretation.” I rather repeat once
more the warning that I have given again and again for twenty years:
rather to be overcautious in admitting transmission as the cause of
analogies in cases of the sporadic occurrence of similar phenomena,
than to operate with the concept of lost links of a chain of cultural

That through the exaggerated application of a single principle, when
several must be admitted as acting, new viewpoints may be discovered
— that much I willingly admit, and I enjoy to follow the daring generalizations
303to which Mr. Graebner is led. I may, however, be pardoned
if I cannot accept this as the method of ethnology. I see safe progress
essentially in the patient unraveling of the mental processes that may
be observed among primitive and civilized peoples, and that express
the actual conditions under which cultural forms develop. When we
begin to know these, we shall also be able to proceed gradually to the
more difficult problems of the cultural relations between isolated areas
that exhibit peculiar similarities.304

11 Science, N.S., vol. 34 (1911), pp. 804-810.

21 Science, N.S., vol. 4 (1896), pp. 901-908; pp. 270 et seq. of this volume.

32 Journal of American Folk-Lore, vol. 14 (1901), pp. 1-11.