4 4
ETHNOSCIENCE VS. CULTURAL MATERIALISM:
A STUDY IN FALSE OPPOSITIONS'
Katherine Newman
Department of Anthropology
University of California
Berkeley, California
The complex and sometimes bitter debate be-
tween proponents of "idealism" and advocates of
"materialism" has long been with us in Anthropol-
ogy. The conflict between these two perspectives
centers on theoretical and political differences too
numerous to detail. However, an examination of a
particular instance of this general debate may shed
some light on the sorts of issues involved. Specifi-
cally, this paper attempts to analyze the basis of
the dispute between the two subdisciplines of
Anthropology known as ethnoscience and cultural
materialism, which represent the idealist and
materialist traditions respectively. At varying
points in time, both ethnoscientists and cultural
materialists proclaimed their theoretical and meth-
odological contributions as the foundation of
a "new" anthropology, thus rejecting as inadequate
what had come before (Sturtevant 1964; Harris
1968, 1975; Goodenough 1970; Kottak 1975).
However, a closer examination of the differences
between cultural materialism and ethnoscience
suggests that to view them as competing paradigms
is to create an unjustified and false dichotomy.
It may be useful to approach the dispute be-
tween cultural materialists and ethnoscientists
from the vantage point of the following three ques-
tions: 1) What is the proper domain of study for
Anthropology? 2) In what way does the analysis
of this domain contribute to an understanding of
causality in social evolution? and 3) What are the
appropriate methods for the investigation of this
domain? Cultural materialists and ethnoscientists
offer very different answers to each of these ques-
tions. However, I shall argue that their differences,
while profound, are theoretically compatible.
Classical ethnoscience, as it was developed by
individuals such as Frake, Goodenough, Conklin,
Sturtevant and others, had as its major focus the
discovery and description of culture. This is not
a particularly earth-shattering proposal since the
study of culture has always been a major concern
within anthropology. However, the early ethno-
scientists had developed a new, though certainly
not unanticipated, sense of what culture was. They
proposed that culture be regarded as a shared
grammar of behavior for particular social systems.
Just as linguistic grammars provide rules for the
production of acceptable utterances, so a cultural
grammar guides the social actor in the production
and interpretation of socially appropriate behavior.
The cultural grammar itself, the cognitive rules
which were thought to underlie social interaction,
was to be the focus of the so-called New Ethnog-
raphy. Moreover, this new orientation toward cul-
ture was intended to replace, rather than merely
complement, what the early ethnoscientists felt
were the inadequate and unscientific definitions
used by more traditional anthropologists.
"Classical" ethnoscience had a distinctively
progammatic quality about it. "Transcultural
Studies in Cognition," the special publication of
the American Anthropologist dedicated entirely to
cognitive anthropology, announced to the field
that the new approach was the wave of the future:
Ethnoscience shows promise as the New Eth-
nography required to advance the whole of
cultural anthropology. (Sturtevant 1964: 101)
The insistence that culture be studied as a cognitive
grammar evoked criticisms from all sides. Anthro-
pologists schooled in the tradition of participant-
observation argued that people seldom follow the
"cultural rules" imparted to anthropologists during
interviews. The whole point of the fieldwork em-
phasis was to reveal "what was really going on,"
something which the native of the culture was
thought not to be fully aware of (Rappaport 1967;
Vayda and Rappaport 1968). Thus, some individ-
uals attacked the major premise of ethnoscience,
namely, that the native is the most reliable author-
ity where underlying rules of behavior are con-
cerned.
Adherents of the humanistic, wholistic view of
anthropology complained that the new view of
culture was devoid of the essential spirit of the
discipline, which is its concern with the human ex-
perience and all of its problems:
. . . I think . . . that much of the current
effort . . . has the effect of sacrificing the in-
sight into the nature of human behavior
which is the ultimate aim of all ethnography.
4 5
It results in astoundingly pallid, sterile, and
fragmentary ethnography. It is effectively a
retreat to method without sufficient reference
to goals. [Berreman 1966]
However, the most vehement critics of early
ethnoscience were the cultural materialists, led by
the well-known Columbia anthropologist, Marvin
Harris. The cultural materalists' critique focused on
the issue of the proper domain of study for anthro-
pology. Harris did not deny that cognitive systems
existed, nor did he suggest that these systems were
not involved in the production of social behavior.
What the cultural materialists did balk at was the
apparent exclusion of non-cognitive, and especially
non-linguistic, realities: technological development,
the organization of production, and other material
constraints on the content of cognitive codes. They
argued that whether or not such basic economic
facts were part of "culture," these facts certainly
should be included in any ethnography. Perhaps
more importantly, cultural materialists insisted
that the examination of the "substructural" com-
ponents of social systems had to take priority in
any causal understanding of social evolution.
Here I come to the crux of the matter insofar as
the initial question is concerned. According to the
classical ethnoscientists, the subject matter of
Anthropology was particular native semantic sys-
tems. Cultural materialists, on the other hand, pro-
posed that anthropology return to its original 19th
century focus on evolutionism, the "science of
history" (Harris 1968). This renaissance of interest
in the topic of social evolution was to be under-
taken from a particular point of view, namely, a
neo-Marxian perspective which strongly empha-
sized economic imperatives as the driving force
behind social change. Thus, at the very outset,
ethnoscience and cultural materialism took entirely
different stands on the question of what anthropol-
ogy was all about. Each was equally adamant about
the importance of reorienting the field around its
own particular interests. This insistence on a uni-
lateral "program" for anthropology won both
subdisciplines substantial criticism from colleagues
outside either orientation. However, for the pres-
ent purposes it will be argued that these two pro-
posals were only contradictory in the sense that
the total amount of resources expended in the
pursuit of anthropological knowledge is finite, and
therefore, choosing one strategy over another may
result in downplaying the value of other theoretical
alternatives. However, this point aside, I would
argue that the programs of ethnoscience and cul-
tural materialism easily could be taken as comple-
mentary-
If we proceed to the second question, namely,
the relevance of the subject matter of anthro-
pology to the question of causality in social evolu-
tion, we may find the root of the cultural mater-
ialists' misinterpretation of the ethnoscientific
quest. The key word to consider is causality. We
generally expect theories of social change and
evolution to provide more than just a sequence of
historical stages in the development of societies or
culture areas. Evolutionary theories also must
attempt to supply an explanation for the under-
lying causal mechanisms which "push" social
evolution in a certain direction.
Harris maintained that ethnoscience, as an out-
growth of idealism, was particularly incapable of
providing an adequate theory of techno-economic
and social evolution. In fact, in The Rise of An-
thropological Theory, Harris suggested that enthno-
science was an anthropological descendant of the
classical idealism of Hegel, who proposed a theory
of evolution based upon the unfolding of reason
(Harris: 1968; Hegel: 1956). On Harris' first point
I would certainly have to agree; ethnoscience, like
all other subdisciplines interested in the "native
worldview," is not well suited for (or, for that
matter, particularly interested in) providing evi-
dence for causal mechanisms of social evolution.
In fact, during the early days of ethnoscience,
scarcely any attention was paid to the whole issue
of evolution.
On the second point, however, I must disagree
with Harris. Hegel and his intellectual descendants
in the idealist tradition did have a theory of social
evolution which saw the development of reason as
the guiding force behind change. Ethnoscience
does not now, and never did to my knowledge, par-
take of this theory of evolution. Nowhere in the
literature of cognitive anthropology do we see a
theory of causality in social change, and certainly
not a theory which could be construed as being in
opposition to a materialist one. Harris, and other
critics of the ethnoscientific movement, assumed
that because ethnoscience and the idealist tradition
have certain common interests, they overlap in all
of their theoretical assumptions. This simply is not
the case. As I pointed out earlier, the classical
ethnoscientists were not concerned with the topic
of evolutionism.
Precisely the reverse is true of contemporary
ethnoscience. Native systems of classification rep-
resent one of the primary source materials for the
study of cognitive culture. As the data base on
these systems of classification grew, it became clear
that there were identifiable regularities in the
formal structure of many semantic domains in
unrelated language families. The work done by
individuals such as Brent Berlin, Paul Kay, and
others, in such areas as folk biology and color lexi-
cons, marked the end of the classical period in eth-
4 6
noscience and the beginning of the comparative/
evolutionary era (Berlin 1972); Berlin and Kay
1969; Kay 1975). The orientation of these cogni-
tive anthropologists is almost exclusively evolution-
ary in nature. However, it is crucial to the argument
presented here that we understand the present
status of this universalistic-evolutionary trend in
ethnoscience. What Berlin, Kay and their associates
have uncovered are sequences in the encoding of
lexical items. They have been able to establish that
these developments in the lexicons of unrelated
languages are followed with surprising regularity
across a very large sample of the world's languages.
Clearly, this work is still at the descriptive level-no
definitive theory is offered which would provide
the causal mechanisms underlying language evolu-
tion in the domains investigated by these individu-
als. Where tentative observations regarding theories
of change have been advanced, they have been
unquestionably materialist in character:
Consonant with our suggestion that the color
lexicon evolves in a specifiable order is the
additional observation that languages which
possess few basic color terms . . . are invari-
ably spoken by peoples which exhibit relative-
ly primitive levels of economic and technolog-
ical development . . . On the other hand,
languages possessing rather full color lexicons
are characteristically spoken by the more
civilized nations of the world. [Berlin 1970:
14]
Perhaps the most telling evidence of a materialist
theory of evolution in ethnoscience is to be found
in the work of Berlin, Breedlove, and Raven on the
effects of agricultural technology on native systems
of biological classification (Berlin 1972; Berlin,
Breedlove, and Raven 1970). This research demon-
strates that the lexicon for plant names undergoes
certain identifiable linguistic changes when cultiva-
tion is present. Specifically, Berlin and his col-
leagues show that the more significant (i.e., culti-
vated or protected) a plant is, the more resistant
the lexical items representing the plant are to lin-
guistic change. What we have here is not so much a
completely developed theory of lexical change in
ethnobotany, but rather, a recognition of the im-
portant role played by technology in lexical devel-
opment and some empirical support for Berlin's
view that causal explanations for language evolu-
tion will probably be found "with some more
general, technologically based theory of cultural
evolution." (Berlin 1972)
In order for the materialist critique of ethno-
science to stand, it would have to demonstrate that
the focus on the realm of ideas (e.g., semantic
classification, linguistic categories, etc.) has led to
an idealist theory of social evolution. It is true that
ethnoscientists spend little time concentrating on
the sorts of topics of primary concern to Harris,
Kottak and others in the materialist "school."
However, ethnoscience has not developed an alter-
native to the materialist theories of social change,
nor does it show any signs of moving in that direc-
tion.
What cognitive anthropology can contribute to
the study of social change is an understanding of
the "emic" face of "etic" processes. Having accept-
ed materialist explanations of social evolution,
ethnoscience proceeds to discover what the conse-
quences of that process have been for semantic
systems. Surely both aspects of the human experi-
ence are essential to any comprehensive theory of
change.
While these theoretical issues have been raised
repeatedly since the publication of The Rise of
Anthropological Theory (Harris 1968), perhaps the
most bitter argument between materialists and
ethnoscientists has revolved around the issue of re-
search methodology. Clearly, the methods employ-
ed by the two subdisciplines are different, but
again, I suggest, they are complementary. This
should come as no surprise, for as I have argued in
the preceding pages, proponents of the two "sides"
are interested in different, but certainly comple-
mentary issues.
The emphasis of the early ethnoscientists was at
least as much methodological as it was substantive
or theoretical. In systematizing their theory of
culture, they found it imperative to develop re-
search approaches which could be replicated and
formalized into predictive statements in the tradi-
tion of classical scientific methodology. Lack of a
reliable research methodology within traditional
anthropology was one of the key concerns of the
founders of ethnoscience. Consequently, they
placed a very strong emphasis on strict research
designs which could be duplicated by other work-
ers in the field for comparative pruposes. This pre-
occupation with rigorous data collection and analy-
sis served to place the issue methodology in the
foreground (which is rarely the case in anthropol-
ogy), where it was attacked rather vehemently by
Harris among others (Harris 1968).
We have already noted that cognitive anthro-
pology, as a science of the "superstructure," took
as its major concern the structure and content of
cultural rules as they exist in the minds of native
informants. One of the primary sources of infor-
mation about these cultural rules is the semantic
categories present in the native language, for one
of the prerequisites to "acting appropriately" by
the native definition, is a grasp of the meanings of
words, and the capacity to structure those mean-
ings into ordered categories. This shared knowledge
4 7
was thought to underly "grammatical" activities,
be they linguistic or interactive in some other
sense. As the investigation of semantic properties
of native languages assumed prominence in ethno-
science, the methodology naturally focused upon
the most efficient and rigorous ways of elucidating
this type of information. Various methods of elici-
tation, modeled loosely after techniques in field
linguistics, were developed. They ranged from
highly controlled elicitation techniques which
minimized the influence of external environmental
/social influences on the informant ("White
Room" ethnography), to the more naturalistic,
contextualized strategies which utilized the exter-
nal environment as a source of question frames
("Grass Hut" ethnography). However, the general
thrust of all these efforts was the same: to uncover
the basic cultural grammar, beginning with the
semantic properties of the native language.
The methodology of cultural materialism could
not have been farther removed from that of ethno-
science. The materialists' interests in aggregate
phenomena (e.g., food production, economic
input/output, resource distribution and consump-
tion, etc.) led to the use of research techniques of
"objective" measurement. By objective I mean, in
this case, the use of measurement systems which
were not necessarily familiar to the native culture
under investigation. Utilizing measurements of
kilocalories, crop yields, and other indicators of
aggregate economic/ecological activity, the cultural
materialists have attempted to provide information
on the productive forces which are seen to underlie
social organization, political alliance, ritual activ-
ity, and a host of other sub-systems (Rappaport
1967; Vayda and Rappaport 1968).
It seems only fitting that the two subfields,
with their divergent interests, would have develop-
ed different methodological approaches. It is in-
conceivable that the methods of either side would
have been applicable to the research topics of the
other. However, both methodological approaches
can coexist; as with all other research endeavors,
the selection of a methodology should be governed
by the type of information required. Proclaiming
the superiority of one strategy over another makes
sense only where the same data base is the object
of concern. Insofar as ethnoscience and cultural
materialism are concerned, it is rarely (if ever) the
case that the same kind of data would be required.
In this paper I have argued that to view ethno-
science and cultural materialism as conflicting ap-
proaches is to create an unnecessary and unjustified
opposition. The contrasts between the two sub-
disciplines of Anthropology are more a matter of
complementary view of the basic concerns of the
field. These views both contribute to our under-
standing of social evolution, the oldest of anthro-
pological problems, without contradicting each
other. Finally, as we would expect, each sub-
discipline adopts a different research strategy in
an attempt to uncover fundamentally different
bodies of information.
At the outset, it was suggested that the debate
between ethnoscience and cultural materialism is a
subset of a larger, longstanding controversy in
Anthropology. I think it is helpful to view the con-
flict in this context, and in this sense I would agree
with Harris that much of anthropological theory
since the mid-19th century has involved a certain
amount of shadow boxing with the work of Marx
and his followers (Harris: 1969). At the same time,
however, it is important to recognize that in the
case of ethnoscience vs. cultural materialism, the
difficulty is more a matter of two fields talking past
each other than it is one of fundamentally conflict-
ing views on the driving forces behind social
change. Both subdisciplines accept the materialist-
oriented view of the primacy of -technology and
economy.
While we may wish to claim priority of one sub-
discipline over the other, given some sense of
urgency for discovery or some recognition of the
limited resources available for research in general,
we must admit that neither perspective invalidates
or contradicts the other.
NOTES
'This paper was originally presented at the annual
meeting of the Kroeber Anthropological Society in
May 1976.
interests and emphases than of irreconcilable dif-
ferences. Each subdiscipline has a distinctive but
4 8
REFERENCES CITED
Berlin, B.
1970 A Universalistic-Evolutionary Approach
to Ethnographic Semantics. Bulletins of
the American Anthropological Association
3: 19-31.
Berlin, B. and P. Kay
1969 Basic Color Terms: Their Evolution and
Universality. Berkeley: University of Cali-
fornia Press.
Berlin, B., D. Breedlove, and P. Raven
1970 Cultural Significance and Lexical Reten-
tion in Tzeltal-Tzotzil Ethnobotany. In
Meaning in Mayan Languages, M. Ed-
mundson, Ed. The Hague: Mouton Press.
Berreman, G.
1966 Anemic and Emetic Analyses in Social
Anthropology. American Anthropologist
68: 346-354.
Goodenough, W.
1970 Description and Comparison in Cultural
Anthropology (Chapter 4). Chicago:
Aldine Publishing Co.
Harris, M.
1968 The Rise of Anthropological Theory. New
York: Thomas Crowell.
1975 "Cultural Materialism," a paper presented
at the annual meetings of the American
Anthropological Association, San Fran-
cisco.
Hegel, G.
1956 The Philosophy of History. New York:
Dover.
Kay, P.
1975 Synchronic Variation  and Diachronic
Change in Basic Color Terms. Language in
Society 4: 257-270.
Kottak, C.
1975 "Ethnoscience and Structuralism," a
paper presented at the annual meetings of
the American Anthropological Associa-
tion, San Francisco.
Sturtevant, W.
1964 Studies in Ethnoscience. American An-
thropologist 66: 99-124.
Rappaport, R.
1967 Pigs for the Ancestors. New Haven: Yale
University Press.
Vayda, A. and R. Rappaport
1968 Ecology, Cultural and Noncultural. In In-
troduction to Cultural Anthropology:
Essays in the Scope and Method of the
Science of Man, J. Clifton, Ed. Boston:
Houghton-Mifflin Publishing Co.